diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,73920 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5000236910684672,
+  "eval_steps": 10553,
+  "global_step": 10553,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.738213693437574e-05,
+      "grad_norm": 0.265625,
+      "learning_rate": 2e-05,
+      "loss": 1.5032,
+      "step": 1
+    },
+    {
+      "epoch": 4.738213693437574e-05,
+      "eval_loss": 1.6555964946746826,
+      "eval_runtime": 1297.243,
+      "eval_samples_per_second": 1.808,
+      "eval_steps_per_second": 1.808,
+      "step": 1
+    },
+    {
+      "epoch": 9.476427386875148e-05,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4e-05,
+      "loss": 1.7934,
+      "step": 2
+    },
+    {
+      "epoch": 0.00014214641080312722,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6e-05,
+      "loss": 1.5472,
+      "step": 3
+    },
+    {
+      "epoch": 0.00018952854773750296,
+      "grad_norm": 0.35546875,
+      "learning_rate": 8e-05,
+      "loss": 1.7256,
+      "step": 4
+    },
+    {
+      "epoch": 0.00023691068467187872,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001,
+      "loss": 1.2578,
+      "step": 5
+    },
+    {
+      "epoch": 0.00028429282160625445,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00012,
+      "loss": 1.3301,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003316749585406302,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00014,
+      "loss": 1.5469,
+      "step": 7
+    },
+    {
+      "epoch": 0.0003790570954750059,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016,
+      "loss": 1.4694,
+      "step": 8
+    },
+    {
+      "epoch": 0.00042643923240938164,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00018,
+      "loss": 0.9671,
+      "step": 9
+    },
+    {
+      "epoch": 0.00047382136934375743,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002,
+      "loss": 1.6308,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005212035062781332,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019999999889105325,
+      "loss": 1.5963,
+      "step": 11
+    },
+    {
+      "epoch": 0.0005685856432125089,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019999999556421307,
+      "loss": 1.694,
+      "step": 12
+    },
+    {
+      "epoch": 0.0006159677801468846,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019999999001947948,
+      "loss": 2.0886,
+      "step": 13
+    },
+    {
+      "epoch": 0.0006633499170812604,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019999998225685262,
+      "loss": 1.9375,
+      "step": 14
+    },
+    {
+      "epoch": 0.0007107320540156361,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019999997227633268,
+      "loss": 1.429,
+      "step": 15
+    },
+    {
+      "epoch": 0.0007581141909500118,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019999996007791988,
+      "loss": 1.2525,
+      "step": 16
+    },
+    {
+      "epoch": 0.0008054963278843876,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019999994566161444,
+      "loss": 0.9333,
+      "step": 17
+    },
+    {
+      "epoch": 0.0008528784648187633,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019999992902741678,
+      "loss": 1.07,
+      "step": 18
+    },
+    {
+      "epoch": 0.000900260601753139,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019999991017532716,
+      "loss": 1.0739,
+      "step": 19
+    },
+    {
+      "epoch": 0.0009476427386875149,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019999988910534606,
+      "loss": 0.9974,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009950248756218905,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019999986581747394,
+      "loss": 1.147,
+      "step": 21
+    },
+    {
+      "epoch": 0.0010424070125562663,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001999998403117113,
+      "loss": 1.2521,
+      "step": 22
+    },
+    {
+      "epoch": 0.001089789149490642,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019999981258805874,
+      "loss": 0.9572,
+      "step": 23
+    },
+    {
+      "epoch": 0.0011371712864250178,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019999978264651684,
+      "loss": 1.0038,
+      "step": 24
+    },
+    {
+      "epoch": 0.0011845534233593934,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999975048708626,
+      "loss": 1.1757,
+      "step": 25
+    },
+    {
+      "epoch": 0.0012319355602937693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019999971610976775,
+      "loss": 0.8609,
+      "step": 26
+    },
+    {
+      "epoch": 0.001279317697228145,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019999967951456204,
+      "loss": 1.7934,
+      "step": 27
+    },
+    {
+      "epoch": 0.0013266998341625207,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999964070146998,
+      "loss": 1.3382,
+      "step": 28
+    },
+    {
+      "epoch": 0.0013740819710968966,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019999959967049237,
+      "loss": 1.0583,
+      "step": 29
+    },
+    {
+      "epoch": 0.0014214641080312722,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019999955642163015,
+      "loss": 1.0192,
+      "step": 30
+    },
+    {
+      "epoch": 0.001468846244965648,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019999951095488433,
+      "loss": 1.5542,
+      "step": 31
+    },
+    {
+      "epoch": 0.0015162283819000236,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019999946327025584,
+      "loss": 1.3448,
+      "step": 32
+    },
+    {
+      "epoch": 0.0015636105188343995,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019999941336774576,
+      "loss": 1.6881,
+      "step": 33
+    },
+    {
+      "epoch": 0.0016109926557687751,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019999936124735524,
+      "loss": 0.7713,
+      "step": 34
+    },
+    {
+      "epoch": 0.001658374792703151,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019999930690908535,
+      "loss": 1.1188,
+      "step": 35
+    },
+    {
+      "epoch": 0.0017057569296375266,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019999925035293738,
+      "loss": 1.0596,
+      "step": 36
+    },
+    {
+      "epoch": 0.0017531390665719024,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019999919157891256,
+      "loss": 1.3195,
+      "step": 37
+    },
+    {
+      "epoch": 0.001800521203506278,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019999913058701217,
+      "loss": 0.9574,
+      "step": 38
+    },
+    {
+      "epoch": 0.0018479033404406539,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019999906737723757,
+      "loss": 0.7413,
+      "step": 39
+    },
+    {
+      "epoch": 0.0018952854773750297,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019999900194959017,
+      "loss": 1.1054,
+      "step": 40
+    },
+    {
+      "epoch": 0.0019426676143094053,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999893430407145,
+      "loss": 0.8824,
+      "step": 41
+    },
+    {
+      "epoch": 0.001990049751243781,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019999886444068286,
+      "loss": 1.0459,
+      "step": 42
+    },
+    {
+      "epoch": 0.002037431888178157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000199998792359426,
+      "loss": 0.7952,
+      "step": 43
+    },
+    {
+      "epoch": 0.0020848140251125327,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019999871806030239,
+      "loss": 1.1249,
+      "step": 44
+    },
+    {
+      "epoch": 0.0021321961620469083,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019999864154331376,
+      "loss": 1.4,
+      "step": 45
+    },
+    {
+      "epoch": 0.002179578298981284,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019999856280846173,
+      "loss": 1.5702,
+      "step": 46
+    },
+    {
+      "epoch": 0.00222696043591566,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019999848185574815,
+      "loss": 1.3921,
+      "step": 47
+    },
+    {
+      "epoch": 0.0022743425728500356,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019999839868517475,
+      "loss": 1.0725,
+      "step": 48
+    },
+    {
+      "epoch": 0.002321724709784411,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999831329674334,
+      "loss": 1.0056,
+      "step": 49
+    },
+    {
+      "epoch": 0.002369106846718787,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999822569045589,
+      "loss": 1.0241,
+      "step": 50
+    },
+    {
+      "epoch": 0.002416488983653163,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019999813586631427,
+      "loss": 1.0889,
+      "step": 51
+    },
+    {
+      "epoch": 0.0024638711205875385,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999804382432053,
+      "loss": 1.1529,
+      "step": 52
+    },
+    {
+      "epoch": 0.002511253257521914,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999794956447673,
+      "loss": 1.4482,
+      "step": 53
+    },
+    {
+      "epoch": 0.00255863539445629,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019999785308678488,
+      "loss": 0.7506,
+      "step": 54
+    },
+    {
+      "epoch": 0.002606017531390666,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019999775439124716,
+      "loss": 0.9041,
+      "step": 55
+    },
+    {
+      "epoch": 0.0026533996683250414,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019999765347786578,
+      "loss": 1.0598,
+      "step": 56
+    },
+    {
+      "epoch": 0.002700781805259417,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019999755034664295,
+      "loss": 0.9802,
+      "step": 57
+    },
+    {
+      "epoch": 0.002748163942193793,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019999744499758096,
+      "loss": 0.7913,
+      "step": 58
+    },
+    {
+      "epoch": 0.0027955460791281687,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019999733743068215,
+      "loss": 0.9532,
+      "step": 59
+    },
+    {
+      "epoch": 0.0028429282160625444,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001999972276459489,
+      "loss": 1.6101,
+      "step": 60
+    },
+    {
+      "epoch": 0.00289031035299692,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019999711564338367,
+      "loss": 1.2862,
+      "step": 61
+    },
+    {
+      "epoch": 0.002937692489931296,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001999970014229889,
+      "loss": 0.1047,
+      "step": 62
+    },
+    {
+      "epoch": 0.0029850746268656717,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001999968849847672,
+      "loss": 1.4375,
+      "step": 63
+    },
+    {
+      "epoch": 0.0030324567638000473,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019999676632872108,
+      "loss": 1.3146,
+      "step": 64
+    },
+    {
+      "epoch": 0.003079838900734423,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001999966454548532,
+      "loss": 0.2704,
+      "step": 65
+    },
+    {
+      "epoch": 0.003127221037668799,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001999965223631662,
+      "loss": 1.0367,
+      "step": 66
+    },
+    {
+      "epoch": 0.0031746031746031746,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001999963970536629,
+      "loss": 0.7727,
+      "step": 67
+    },
+    {
+      "epoch": 0.0032219853115375502,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019999626952634599,
+      "loss": 1.244,
+      "step": 68
+    },
+    {
+      "epoch": 0.0032693674484719263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019999613978121834,
+      "loss": 1.08,
+      "step": 69
+    },
+    {
+      "epoch": 0.003316749585406302,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999960078182828,
+      "loss": 1.1897,
+      "step": 70
+    },
+    {
+      "epoch": 0.0033641317223406775,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019999587363754234,
+      "loss": 1.0152,
+      "step": 71
+    },
+    {
+      "epoch": 0.003411513859275053,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019999573723899992,
+      "loss": 1.1051,
+      "step": 72
+    },
+    {
+      "epoch": 0.003458895996209429,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019999559862265856,
+      "loss": 0.8628,
+      "step": 73
+    },
+    {
+      "epoch": 0.003506278133143805,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019999545778852132,
+      "loss": 0.9969,
+      "step": 74
+    },
+    {
+      "epoch": 0.0035536602700781805,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019999531473659135,
+      "loss": 1.3045,
+      "step": 75
+    },
+    {
+      "epoch": 0.003601042407012556,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001999951694668718,
+      "loss": 0.7187,
+      "step": 76
+    },
+    {
+      "epoch": 0.003648424543946932,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001999950219793659,
+      "loss": 1.2484,
+      "step": 77
+    },
+    {
+      "epoch": 0.0036958066808813078,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001999948722740769,
+      "loss": 1.1648,
+      "step": 78
+    },
+    {
+      "epoch": 0.0037431888178156834,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001999947203510082,
+      "loss": 1.1182,
+      "step": 79
+    },
+    {
+      "epoch": 0.0037905709547500594,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999945662101631,
+      "loss": 0.9652,
+      "step": 80
+    },
+    {
+      "epoch": 0.003837953091684435,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019999440985154498,
+      "loss": 1.1329,
+      "step": 81
+    },
+    {
+      "epoch": 0.0038853352286188107,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001999942512751574,
+      "loss": 0.6673,
+      "step": 82
+    },
+    {
+      "epoch": 0.003932717365553186,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019999409048100382,
+      "loss": 0.743,
+      "step": 83
+    },
+    {
+      "epoch": 0.003980099502487562,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001999939274690878,
+      "loss": 0.8757,
+      "step": 84
+    },
+    {
+      "epoch": 0.0040274816394219376,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.000199993762239413,
+      "loss": 1.2065,
+      "step": 85
+    },
+    {
+      "epoch": 0.004074863776356314,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001999935947919831,
+      "loss": 1.5393,
+      "step": 86
+    },
+    {
+      "epoch": 0.00412224591329069,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019999342512680172,
+      "loss": 0.6055,
+      "step": 87
+    },
+    {
+      "epoch": 0.004169628050225065,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001999932532438727,
+      "loss": 0.963,
+      "step": 88
+    },
+    {
+      "epoch": 0.004217010187159441,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019999307914319981,
+      "loss": 0.5133,
+      "step": 89
+    },
+    {
+      "epoch": 0.0042643923240938165,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019999290282478698,
+      "loss": 1.1098,
+      "step": 90
+    },
+    {
+      "epoch": 0.004311774461028192,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019999272428863804,
+      "loss": 0.9357,
+      "step": 91
+    },
+    {
+      "epoch": 0.004359156597962568,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019999254353475702,
+      "loss": 0.9475,
+      "step": 92
+    },
+    {
+      "epoch": 0.004406538734896943,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019999236056314783,
+      "loss": 0.9209,
+      "step": 93
+    },
+    {
+      "epoch": 0.00445392087183132,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019999217537381464,
+      "loss": 1.2401,
+      "step": 94
+    },
+    {
+      "epoch": 0.0045013030087656955,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001999919879667615,
+      "loss": 1.2056,
+      "step": 95
+    },
+    {
+      "epoch": 0.004548685145700071,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019999179834199256,
+      "loss": 0.8911,
+      "step": 96
+    },
+    {
+      "epoch": 0.004596067282634447,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019999160649951202,
+      "loss": 1.5826,
+      "step": 97
+    },
+    {
+      "epoch": 0.004643449419568822,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019999141243932418,
+      "loss": 1.5075,
+      "step": 98
+    },
+    {
+      "epoch": 0.004690831556503198,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999121616143332,
+      "loss": 1.2497,
+      "step": 99
+    },
+    {
+      "epoch": 0.004738213693437574,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019999101766584378,
+      "loss": 1.0032,
+      "step": 100
+    },
+    {
+      "epoch": 0.00478559583037195,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019999081695255998,
+      "loss": 1.241,
+      "step": 101
+    },
+    {
+      "epoch": 0.004832977967306326,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019999061402158636,
+      "loss": 0.1521,
+      "step": 102
+    },
+    {
+      "epoch": 0.004880360104240701,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019999040887292745,
+      "loss": 1.5866,
+      "step": 103
+    },
+    {
+      "epoch": 0.004927742241175077,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019999020150658774,
+      "loss": 1.6,
+      "step": 104
+    },
+    {
+      "epoch": 0.004975124378109453,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019998999192257188,
+      "loss": 1.1149,
+      "step": 105
+    },
+    {
+      "epoch": 0.005022506515043828,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001999897801208845,
+      "loss": 0.9298,
+      "step": 106
+    },
+    {
+      "epoch": 0.005069888651978204,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001999895661015303,
+      "loss": 1.2095,
+      "step": 107
+    },
+    {
+      "epoch": 0.00511727078891258,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019998934986451404,
+      "loss": 0.8567,
+      "step": 108
+    },
+    {
+      "epoch": 0.005164652925846956,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001999891314098405,
+      "loss": 0.9423,
+      "step": 109
+    },
+    {
+      "epoch": 0.005212035062781332,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019998891073751452,
+      "loss": 0.8099,
+      "step": 110
+    },
+    {
+      "epoch": 0.005259417199715707,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019998868784754103,
+      "loss": 1.0169,
+      "step": 111
+    },
+    {
+      "epoch": 0.005306799336650083,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019998846273992492,
+      "loss": 1.4406,
+      "step": 112
+    },
+    {
+      "epoch": 0.0053541814735844585,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019998823541467122,
+      "loss": 0.9936,
+      "step": 113
+    },
+    {
+      "epoch": 0.005401563610518834,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019998800587178495,
+      "loss": 1.3192,
+      "step": 114
+    },
+    {
+      "epoch": 0.00544894574745321,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019998777411127123,
+      "loss": 0.8875,
+      "step": 115
+    },
+    {
+      "epoch": 0.005496327884387586,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019998754013313515,
+      "loss": 0.9446,
+      "step": 116
+    },
+    {
+      "epoch": 0.005543710021321962,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019998730393738198,
+      "loss": 1.1031,
+      "step": 117
+    },
+    {
+      "epoch": 0.0055910921582563375,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001999870655240169,
+      "loss": 1.1828,
+      "step": 118
+    },
+    {
+      "epoch": 0.005638474295190713,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019998682489304515,
+      "loss": 1.3674,
+      "step": 119
+    },
+    {
+      "epoch": 0.005685856432125089,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019998658204447217,
+      "loss": 0.8399,
+      "step": 120
+    },
+    {
+      "epoch": 0.005733238569059464,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001999863369783033,
+      "loss": 1.0854,
+      "step": 121
+    },
+    {
+      "epoch": 0.00578062070599384,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000199986089694544,
+      "loss": 0.4969,
+      "step": 122
+    },
+    {
+      "epoch": 0.0058280028429282165,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001999858401931997,
+      "loss": 0.0215,
+      "step": 123
+    },
+    {
+      "epoch": 0.005875384979862592,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019998558847427597,
+      "loss": 1.1571,
+      "step": 124
+    },
+    {
+      "epoch": 0.005922767116796968,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019998533453777838,
+      "loss": 1.3631,
+      "step": 125
+    },
+    {
+      "epoch": 0.005970149253731343,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001999850783837126,
+      "loss": 0.6508,
+      "step": 126
+    },
+    {
+      "epoch": 0.006017531390665719,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019998482001208425,
+      "loss": 0.8974,
+      "step": 127
+    },
+    {
+      "epoch": 0.006064913527600095,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019998455942289912,
+      "loss": 1.0397,
+      "step": 128
+    },
+    {
+      "epoch": 0.00611229566453447,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019998429661616292,
+      "loss": 0.8684,
+      "step": 129
+    },
+    {
+      "epoch": 0.006159677801468846,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019998403159188154,
+      "loss": 1.4368,
+      "step": 130
+    },
+    {
+      "epoch": 0.006207059938403222,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019998376435006082,
+      "loss": 0.6405,
+      "step": 131
+    },
+    {
+      "epoch": 0.006254442075337598,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019998349489070677,
+      "loss": 0.6771,
+      "step": 132
+    },
+    {
+      "epoch": 0.006301824212271974,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019998322321382523,
+      "loss": 1.0881,
+      "step": 133
+    },
+    {
+      "epoch": 0.006349206349206349,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019998294931942233,
+      "loss": 0.6219,
+      "step": 134
+    },
+    {
+      "epoch": 0.006396588486140725,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001999826732075041,
+      "loss": 0.9019,
+      "step": 135
+    },
+    {
+      "epoch": 0.0064439706230751004,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019998239487807666,
+      "loss": 1.1105,
+      "step": 136
+    },
+    {
+      "epoch": 0.006491352760009476,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019998211433114622,
+      "loss": 1.1331,
+      "step": 137
+    },
+    {
+      "epoch": 0.0065387348969438526,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019998183156671898,
+      "loss": 0.9425,
+      "step": 138
+    },
+    {
+      "epoch": 0.006586117033878228,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019998154658480122,
+      "loss": 0.9189,
+      "step": 139
+    },
+    {
+      "epoch": 0.006633499170812604,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019998125938539924,
+      "loss": 0.8435,
+      "step": 140
+    },
+    {
+      "epoch": 0.0066808813077469794,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001999809699685194,
+      "loss": 1.3486,
+      "step": 141
+    },
+    {
+      "epoch": 0.006728263444681355,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019998067833416815,
+      "loss": 1.1616,
+      "step": 142
+    },
+    {
+      "epoch": 0.006775645581615731,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019998038448235195,
+      "loss": 1.0803,
+      "step": 143
+    },
+    {
+      "epoch": 0.006823027718550106,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019998008841307736,
+      "loss": 0.9512,
+      "step": 144
+    },
+    {
+      "epoch": 0.006870409855484482,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019997979012635085,
+      "loss": 0.2863,
+      "step": 145
+    },
+    {
+      "epoch": 0.006917791992418858,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019997948962217912,
+      "loss": 0.7258,
+      "step": 146
+    },
+    {
+      "epoch": 0.006965174129353234,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001999791869005688,
+      "loss": 1.2192,
+      "step": 147
+    },
+    {
+      "epoch": 0.00701255626628761,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001999788819615266,
+      "loss": 1.2817,
+      "step": 148
+    },
+    {
+      "epoch": 0.007059938403221985,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019997857480505928,
+      "loss": 0.8469,
+      "step": 149
+    },
+    {
+      "epoch": 0.007107320540156361,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001999782654311737,
+      "loss": 0.6144,
+      "step": 150
+    },
+    {
+      "epoch": 0.0071547026770907365,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019997795383987663,
+      "loss": 1.7007,
+      "step": 151
+    },
+    {
+      "epoch": 0.007202084814025112,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019997764003117509,
+      "loss": 0.8697,
+      "step": 152
+    },
+    {
+      "epoch": 0.007249466950959489,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019997732400507597,
+      "loss": 1.0246,
+      "step": 153
+    },
+    {
+      "epoch": 0.007296849087893864,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019997700576158628,
+      "loss": 0.9222,
+      "step": 154
+    },
+    {
+      "epoch": 0.00734423122482824,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019997668530071308,
+      "loss": 1.0439,
+      "step": 155
+    },
+    {
+      "epoch": 0.0073916133617626155,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001999763626224635,
+      "loss": 1.2715,
+      "step": 156
+    },
+    {
+      "epoch": 0.007438995498696991,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001999760377268447,
+      "loss": 0.5978,
+      "step": 157
+    },
+    {
+      "epoch": 0.007486377635631367,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019997571061386386,
+      "loss": 1.2903,
+      "step": 158
+    },
+    {
+      "epoch": 0.007533759772565742,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019997538128352826,
+      "loss": 0.6958,
+      "step": 159
+    },
+    {
+      "epoch": 0.007581141909500119,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001999750497358452,
+      "loss": 1.2971,
+      "step": 160
+    },
+    {
+      "epoch": 0.0076285240464344945,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000199974715970822,
+      "loss": 0.7183,
+      "step": 161
+    },
+    {
+      "epoch": 0.00767590618336887,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001999743799884661,
+      "loss": 1.6736,
+      "step": 162
+    },
+    {
+      "epoch": 0.007723288320303246,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019997404178878495,
+      "loss": 0.7535,
+      "step": 163
+    },
+    {
+      "epoch": 0.007770670457237621,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019997370137178603,
+      "loss": 1.0898,
+      "step": 164
+    },
+    {
+      "epoch": 0.007818052594171997,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001999733587374769,
+      "loss": 0.5674,
+      "step": 165
+    },
+    {
+      "epoch": 0.007865434731106373,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019997301388586519,
+      "loss": 0.8733,
+      "step": 166
+    },
+    {
+      "epoch": 0.007912816868040748,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019997266681695845,
+      "loss": 0.6212,
+      "step": 167
+    },
+    {
+      "epoch": 0.007960199004975124,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019997231753076452,
+      "loss": 1.2933,
+      "step": 168
+    },
+    {
+      "epoch": 0.0080075811419095,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019997196602729102,
+      "loss": 0.9396,
+      "step": 169
+    },
+    {
+      "epoch": 0.008054963278843875,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001999716123065458,
+      "loss": 0.7431,
+      "step": 170
+    },
+    {
+      "epoch": 0.00810234541577825,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00019997125636853676,
+      "loss": 1.022,
+      "step": 171
+    },
+    {
+      "epoch": 0.008149727552712628,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019997089821327172,
+      "loss": 0.5743,
+      "step": 172
+    },
+    {
+      "epoch": 0.008197109689647004,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019997053784075858,
+      "loss": 0.4988,
+      "step": 173
+    },
+    {
+      "epoch": 0.00824449182658138,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019997017525100546,
+      "loss": 0.9723,
+      "step": 174
+    },
+    {
+      "epoch": 0.008291873963515755,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019996981044402033,
+      "loss": 1.2186,
+      "step": 175
+    },
+    {
+      "epoch": 0.00833925610045013,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019996944341981124,
+      "loss": 0.628,
+      "step": 176
+    },
+    {
+      "epoch": 0.008386638237384506,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019996907417838642,
+      "loss": 1.0701,
+      "step": 177
+    },
+    {
+      "epoch": 0.008434020374318882,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019996870271975402,
+      "loss": 0.8977,
+      "step": 178
+    },
+    {
+      "epoch": 0.008481402511253257,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019996832904392226,
+      "loss": 1.2282,
+      "step": 179
+    },
+    {
+      "epoch": 0.008528784648187633,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001999679531508994,
+      "loss": 1.3241,
+      "step": 180
+    },
+    {
+      "epoch": 0.008576166785122009,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001999675750406939,
+      "loss": 0.7242,
+      "step": 181
+    },
+    {
+      "epoch": 0.008623548922056384,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019996719471331403,
+      "loss": 0.5549,
+      "step": 182
+    },
+    {
+      "epoch": 0.00867093105899076,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019996681216876826,
+      "loss": 0.8259,
+      "step": 183
+    },
+    {
+      "epoch": 0.008718313195925136,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019996642740706508,
+      "loss": 1.0661,
+      "step": 184
+    },
+    {
+      "epoch": 0.008765695332859511,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000199966040428213,
+      "loss": 1.3544,
+      "step": 185
+    },
+    {
+      "epoch": 0.008813077469793887,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019996565123222066,
+      "loss": 1.1912,
+      "step": 186
+    },
+    {
+      "epoch": 0.008860459606728264,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019996525981909663,
+      "loss": 0.4306,
+      "step": 187
+    },
+    {
+      "epoch": 0.00890784174366264,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001999648661888496,
+      "loss": 0.7892,
+      "step": 188
+    },
+    {
+      "epoch": 0.008955223880597015,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019996447034148837,
+      "loss": 1.3539,
+      "step": 189
+    },
+    {
+      "epoch": 0.009002606017531391,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019996407227702162,
+      "loss": 1.5831,
+      "step": 190
+    },
+    {
+      "epoch": 0.009049988154465767,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019996367199545824,
+      "loss": 0.1309,
+      "step": 191
+    },
+    {
+      "epoch": 0.009097370291400142,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019996326949680708,
+      "loss": 1.0598,
+      "step": 192
+    },
+    {
+      "epoch": 0.009144752428334518,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019996286478107708,
+      "loss": 0.7607,
+      "step": 193
+    },
+    {
+      "epoch": 0.009192134565268894,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019996245784827723,
+      "loss": 0.9167,
+      "step": 194
+    },
+    {
+      "epoch": 0.00923951670220327,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019996204869841654,
+      "loss": 0.5669,
+      "step": 195
+    },
+    {
+      "epoch": 0.009286898839137645,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019996163733150408,
+      "loss": 0.2394,
+      "step": 196
+    },
+    {
+      "epoch": 0.00933428097607202,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019996122374754896,
+      "loss": 0.5832,
+      "step": 197
+    },
+    {
+      "epoch": 0.009381663113006396,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019996080794656038,
+      "loss": 0.8465,
+      "step": 198
+    },
+    {
+      "epoch": 0.009429045249940772,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019996038992854757,
+      "loss": 0.631,
+      "step": 199
+    },
+    {
+      "epoch": 0.009476427386875147,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019995996969351978,
+      "loss": 1.4833,
+      "step": 200
+    },
+    {
+      "epoch": 0.009523809523809525,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001999595472414863,
+      "loss": 1.349,
+      "step": 201
+    },
+    {
+      "epoch": 0.0095711916607439,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001999591225724566,
+      "loss": 0.9398,
+      "step": 202
+    },
+    {
+      "epoch": 0.009618573797678276,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019995869568643996,
+      "loss": 0.6837,
+      "step": 203
+    },
+    {
+      "epoch": 0.009665955934612652,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.000199958266583446,
+      "loss": 0.6042,
+      "step": 204
+    },
+    {
+      "epoch": 0.009713338071547027,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001999578352634841,
+      "loss": 1.0077,
+      "step": 205
+    },
+    {
+      "epoch": 0.009760720208481403,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019995740172656386,
+      "loss": 0.671,
+      "step": 206
+    },
+    {
+      "epoch": 0.009808102345415778,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019995696597269498,
+      "loss": 0.7752,
+      "step": 207
+    },
+    {
+      "epoch": 0.009855484482350154,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019995652800188705,
+      "loss": 1.3379,
+      "step": 208
+    },
+    {
+      "epoch": 0.00990286661928453,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019995608781414977,
+      "loss": 1.0481,
+      "step": 209
+    },
+    {
+      "epoch": 0.009950248756218905,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019995564540949297,
+      "loss": 1.1031,
+      "step": 210
+    },
+    {
+      "epoch": 0.009997630893153281,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001999552007879264,
+      "loss": 0.9984,
+      "step": 211
+    },
+    {
+      "epoch": 0.010045013030087657,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019995475394945996,
+      "loss": 1.442,
+      "step": 212
+    },
+    {
+      "epoch": 0.010092395167022032,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019995430489410353,
+      "loss": 1.1315,
+      "step": 213
+    },
+    {
+      "epoch": 0.010139777303956408,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019995385362186707,
+      "loss": 1.4601,
+      "step": 214
+    },
+    {
+      "epoch": 0.010187159440890783,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019995340013276064,
+      "loss": 1.091,
+      "step": 215
+    },
+    {
+      "epoch": 0.01023454157782516,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019995294442679421,
+      "loss": 0.94,
+      "step": 216
+    },
+    {
+      "epoch": 0.010281923714759536,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019995248650397799,
+      "loss": 1.067,
+      "step": 217
+    },
+    {
+      "epoch": 0.010329305851693912,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019995202636432203,
+      "loss": 1.0678,
+      "step": 218
+    },
+    {
+      "epoch": 0.010376687988628288,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019995156400783663,
+      "loss": 0.968,
+      "step": 219
+    },
+    {
+      "epoch": 0.010424070125562663,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019995109943453198,
+      "loss": 1.2878,
+      "step": 220
+    },
+    {
+      "epoch": 0.010471452262497039,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019995063264441844,
+      "loss": 1.0963,
+      "step": 221
+    },
+    {
+      "epoch": 0.010518834399431415,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999501636375063,
+      "loss": 1.3468,
+      "step": 222
+    },
+    {
+      "epoch": 0.01056621653636579,
+      "grad_norm": 0.375,
+      "learning_rate": 0.000199949692413806,
+      "loss": 0.6686,
+      "step": 223
+    },
+    {
+      "epoch": 0.010613598673300166,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000199949218973328,
+      "loss": 1.2847,
+      "step": 224
+    },
+    {
+      "epoch": 0.010660980810234541,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001999487433160828,
+      "loss": 1.4145,
+      "step": 225
+    },
+    {
+      "epoch": 0.010708362947168917,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019994826544208086,
+      "loss": 1.1932,
+      "step": 226
+    },
+    {
+      "epoch": 0.010755745084103293,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019994778535133292,
+      "loss": 1.2457,
+      "step": 227
+    },
+    {
+      "epoch": 0.010803127221037668,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019994730304384955,
+      "loss": 0.5284,
+      "step": 228
+    },
+    {
+      "epoch": 0.010850509357972044,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019994681851964144,
+      "loss": 1.2445,
+      "step": 229
+    },
+    {
+      "epoch": 0.01089789149490642,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019994633177871935,
+      "loss": 1.1842,
+      "step": 230
+    },
+    {
+      "epoch": 0.010945273631840797,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001999458428210941,
+      "loss": 1.0567,
+      "step": 231
+    },
+    {
+      "epoch": 0.010992655768775172,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019994535164677651,
+      "loss": 0.9081,
+      "step": 232
+    },
+    {
+      "epoch": 0.011040037905709548,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019994485825577748,
+      "loss": 0.788,
+      "step": 233
+    },
+    {
+      "epoch": 0.011087420042643924,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001999443626481079,
+      "loss": 1.2978,
+      "step": 234
+    },
+    {
+      "epoch": 0.0111348021795783,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019994386482377887,
+      "loss": 1.8736,
+      "step": 235
+    },
+    {
+      "epoch": 0.011182184316512675,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019994336478280138,
+      "loss": 0.6525,
+      "step": 236
+    },
+    {
+      "epoch": 0.01122956645344705,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019994286252518646,
+      "loss": 1.6496,
+      "step": 237
+    },
+    {
+      "epoch": 0.011276948590381426,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019994235805094536,
+      "loss": 0.8822,
+      "step": 238
+    },
+    {
+      "epoch": 0.011324330727315802,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019994185136008917,
+      "loss": 0.8822,
+      "step": 239
+    },
+    {
+      "epoch": 0.011371712864250177,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019994134245262922,
+      "loss": 1.0044,
+      "step": 240
+    },
+    {
+      "epoch": 0.011419095001184553,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001999408313285767,
+      "loss": 1.0126,
+      "step": 241
+    },
+    {
+      "epoch": 0.011466477138118929,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019994031798794298,
+      "loss": 1.2154,
+      "step": 242
+    },
+    {
+      "epoch": 0.011513859275053304,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001999398024307395,
+      "loss": 1.3496,
+      "step": 243
+    },
+    {
+      "epoch": 0.01156124141198768,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019993928465697765,
+      "loss": 1.4015,
+      "step": 244
+    },
+    {
+      "epoch": 0.011608623548922056,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001999387646666689,
+      "loss": 1.2544,
+      "step": 245
+    },
+    {
+      "epoch": 0.011656005685856433,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001999382424598248,
+      "loss": 0.6131,
+      "step": 246
+    },
+    {
+      "epoch": 0.011703387822790809,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019993771803645695,
+      "loss": 0.7921,
+      "step": 247
+    },
+    {
+      "epoch": 0.011750769959725184,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019993719139657694,
+      "loss": 1.1962,
+      "step": 248
+    },
+    {
+      "epoch": 0.01179815209665956,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019993666254019648,
+      "loss": 1.268,
+      "step": 249
+    },
+    {
+      "epoch": 0.011845534233593935,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001999361314673273,
+      "loss": 0.8002,
+      "step": 250
+    },
+    {
+      "epoch": 0.011892916370528311,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019993559817798118,
+      "loss": 0.6962,
+      "step": 251
+    },
+    {
+      "epoch": 0.011940298507462687,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019993506267216993,
+      "loss": 0.9077,
+      "step": 252
+    },
+    {
+      "epoch": 0.011987680644397062,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019993452494990543,
+      "loss": 0.9264,
+      "step": 253
+    },
+    {
+      "epoch": 0.012035062781331438,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001999339850111996,
+      "loss": 1.1098,
+      "step": 254
+    },
+    {
+      "epoch": 0.012082444918265814,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019993344285606447,
+      "loss": 1.0,
+      "step": 255
+    },
+    {
+      "epoch": 0.01212982705520019,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019993289848451197,
+      "loss": 1.2045,
+      "step": 256
+    },
+    {
+      "epoch": 0.012177209192134565,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019993235189655426,
+      "loss": 0.9974,
+      "step": 257
+    },
+    {
+      "epoch": 0.01222459132906894,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001999318030922034,
+      "loss": 0.9157,
+      "step": 258
+    },
+    {
+      "epoch": 0.012271973466003316,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001999312520714716,
+      "loss": 0.7541,
+      "step": 259
+    },
+    {
+      "epoch": 0.012319355602937692,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001999306988343711,
+      "loss": 0.8848,
+      "step": 260
+    },
+    {
+      "epoch": 0.012366737739872069,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019993014338091412,
+      "loss": 1.1219,
+      "step": 261
+    },
+    {
+      "epoch": 0.012414119876806445,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000199929585711113,
+      "loss": 1.0632,
+      "step": 262
+    },
+    {
+      "epoch": 0.01246150201374082,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001999290258249801,
+      "loss": 0.7072,
+      "step": 263
+    },
+    {
+      "epoch": 0.012508884150675196,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019992846372252787,
+      "loss": 1.2486,
+      "step": 264
+    },
+    {
+      "epoch": 0.012556266287609572,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019992789940376872,
+      "loss": 0.6055,
+      "step": 265
+    },
+    {
+      "epoch": 0.012603648424543947,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019992733286871523,
+      "loss": 0.7229,
+      "step": 266
+    },
+    {
+      "epoch": 0.012651030561478323,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019992676411737992,
+      "loss": 0.4119,
+      "step": 267
+    },
+    {
+      "epoch": 0.012698412698412698,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019992619314977543,
+      "loss": 1.1707,
+      "step": 268
+    },
+    {
+      "epoch": 0.012745794835347074,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001999256199659144,
+      "loss": 0.8128,
+      "step": 269
+    },
+    {
+      "epoch": 0.01279317697228145,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001999250445658096,
+      "loss": 1.109,
+      "step": 270
+    },
+    {
+      "epoch": 0.012840559109215825,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019992446694947367,
+      "loss": 1.1046,
+      "step": 271
+    },
+    {
+      "epoch": 0.012887941246150201,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019992388711691955,
+      "loss": 0.9036,
+      "step": 272
+    },
+    {
+      "epoch": 0.012935323383084577,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019992330506816001,
+      "loss": 1.2072,
+      "step": 273
+    },
+    {
+      "epoch": 0.012982705520018952,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019992272080320803,
+      "loss": 0.0806,
+      "step": 274
+    },
+    {
+      "epoch": 0.013030087656953328,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019992213432207655,
+      "loss": 1.1417,
+      "step": 275
+    },
+    {
+      "epoch": 0.013077469793887705,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019992154562477855,
+      "loss": 1.5429,
+      "step": 276
+    },
+    {
+      "epoch": 0.01312485193082208,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019992095471132705,
+      "loss": 0.7147,
+      "step": 277
+    },
+    {
+      "epoch": 0.013172234067756456,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019992036158173525,
+      "loss": 1.4001,
+      "step": 278
+    },
+    {
+      "epoch": 0.013219616204690832,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019991976623601624,
+      "loss": 0.8493,
+      "step": 279
+    },
+    {
+      "epoch": 0.013266998341625208,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019991916867418327,
+      "loss": 0.8411,
+      "step": 280
+    },
+    {
+      "epoch": 0.013314380478559583,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00019991856889624957,
+      "loss": 1.1472,
+      "step": 281
+    },
+    {
+      "epoch": 0.013361762615493959,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019991796690222843,
+      "loss": 1.1025,
+      "step": 282
+    },
+    {
+      "epoch": 0.013409144752428334,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019991736269213322,
+      "loss": 1.2855,
+      "step": 283
+    },
+    {
+      "epoch": 0.01345652688936271,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001999167562659773,
+      "loss": 0.8279,
+      "step": 284
+    },
+    {
+      "epoch": 0.013503909026297086,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019991614762377417,
+      "loss": 1.1669,
+      "step": 285
+    },
+    {
+      "epoch": 0.013551291163231461,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019991553676553734,
+      "loss": 0.5494,
+      "step": 286
+    },
+    {
+      "epoch": 0.013598673300165837,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001999149236912803,
+      "loss": 0.7617,
+      "step": 287
+    },
+    {
+      "epoch": 0.013646055437100213,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019991430840101668,
+      "loss": 0.9924,
+      "step": 288
+    },
+    {
+      "epoch": 0.013693437574034588,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019991369089476013,
+      "loss": 0.924,
+      "step": 289
+    },
+    {
+      "epoch": 0.013740819710968964,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019991307117252433,
+      "loss": 1.1997,
+      "step": 290
+    },
+    {
+      "epoch": 0.013788201847903341,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019991244923432303,
+      "loss": 0.4997,
+      "step": 291
+    },
+    {
+      "epoch": 0.013835583984837717,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019991182508017006,
+      "loss": 0.5573,
+      "step": 292
+    },
+    {
+      "epoch": 0.013882966121772092,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001999111987100792,
+      "loss": 1.2159,
+      "step": 293
+    },
+    {
+      "epoch": 0.013930348258706468,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001999105701240644,
+      "loss": 0.5836,
+      "step": 294
+    },
+    {
+      "epoch": 0.013977730395640844,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019990993932213956,
+      "loss": 0.5885,
+      "step": 295
+    },
+    {
+      "epoch": 0.01402511253257522,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00019990930630431865,
+      "loss": 0.0234,
+      "step": 296
+    },
+    {
+      "epoch": 0.014072494669509595,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001999086710706158,
+      "loss": 1.3111,
+      "step": 297
+    },
+    {
+      "epoch": 0.01411987680644397,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000199908033621045,
+      "loss": 0.678,
+      "step": 298
+    },
+    {
+      "epoch": 0.014167258943378346,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019990739395562047,
+      "loss": 1.1531,
+      "step": 299
+    },
+    {
+      "epoch": 0.014214641080312722,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019990675207435634,
+      "loss": 0.0211,
+      "step": 300
+    },
+    {
+      "epoch": 0.014262023217247097,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019990610797726688,
+      "loss": 0.8344,
+      "step": 301
+    },
+    {
+      "epoch": 0.014309405354181473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019990546166436635,
+      "loss": 1.1513,
+      "step": 302
+    },
+    {
+      "epoch": 0.014356787491115849,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019990481313566906,
+      "loss": 1.1517,
+      "step": 303
+    },
+    {
+      "epoch": 0.014404169628050224,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001999041623911895,
+      "loss": 1.2003,
+      "step": 304
+    },
+    {
+      "epoch": 0.0144515517649846,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019990350943094196,
+      "loss": 0.6667,
+      "step": 305
+    },
+    {
+      "epoch": 0.014498933901918977,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019990285425494105,
+      "loss": 1.274,
+      "step": 306
+    },
+    {
+      "epoch": 0.014546316038853353,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001999021968632012,
+      "loss": 1.4139,
+      "step": 307
+    },
+    {
+      "epoch": 0.014593698175787729,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019990153725573705,
+      "loss": 0.7986,
+      "step": 308
+    },
+    {
+      "epoch": 0.014641080312722104,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019990087543256323,
+      "loss": 0.6945,
+      "step": 309
+    },
+    {
+      "epoch": 0.01468846244965648,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019990021139369436,
+      "loss": 0.6369,
+      "step": 310
+    },
+    {
+      "epoch": 0.014735844586590855,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019989954513914527,
+      "loss": 0.8179,
+      "step": 311
+    },
+    {
+      "epoch": 0.014783226723525231,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019989887666893062,
+      "loss": 0.7585,
+      "step": 312
+    },
+    {
+      "epoch": 0.014830608860459607,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019989820598306532,
+      "loss": 1.2037,
+      "step": 313
+    },
+    {
+      "epoch": 0.014877990997393982,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019989753308156423,
+      "loss": 0.9347,
+      "step": 314
+    },
+    {
+      "epoch": 0.014925373134328358,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019989685796444225,
+      "loss": 1.1497,
+      "step": 315
+    },
+    {
+      "epoch": 0.014972755271262734,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019989618063171436,
+      "loss": 0.8473,
+      "step": 316
+    },
+    {
+      "epoch": 0.01502013740819711,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001998955010833956,
+      "loss": 0.8583,
+      "step": 317
+    },
+    {
+      "epoch": 0.015067519545131485,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019989481931950102,
+      "loss": 0.7101,
+      "step": 318
+    },
+    {
+      "epoch": 0.01511490168206586,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019989413534004575,
+      "loss": 0.9949,
+      "step": 319
+    },
+    {
+      "epoch": 0.015162283819000238,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019989344914504497,
+      "loss": 1.3381,
+      "step": 320
+    },
+    {
+      "epoch": 0.015209665955934613,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001998927607345139,
+      "loss": 0.9605,
+      "step": 321
+    },
+    {
+      "epoch": 0.015257048092868989,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00019989207010846777,
+      "loss": 0.8851,
+      "step": 322
+    },
+    {
+      "epoch": 0.015304430229803365,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019989137726692194,
+      "loss": 1.1357,
+      "step": 323
+    },
+    {
+      "epoch": 0.01535181236673774,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019989068220989175,
+      "loss": 0.9308,
+      "step": 324
+    },
+    {
+      "epoch": 0.015399194503672116,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019988998493739263,
+      "loss": 1.1891,
+      "step": 325
+    },
+    {
+      "epoch": 0.015446576640606492,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019988928544944007,
+      "loss": 0.6323,
+      "step": 326
+    },
+    {
+      "epoch": 0.015493958777540867,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019988858374604953,
+      "loss": 1.0981,
+      "step": 327
+    },
+    {
+      "epoch": 0.015541340914475243,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001998878798272366,
+      "loss": 1.325,
+      "step": 328
+    },
+    {
+      "epoch": 0.015588723051409618,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019988717369301688,
+      "loss": 0.0412,
+      "step": 329
+    },
+    {
+      "epoch": 0.015636105188343994,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019988646534340606,
+      "loss": 0.5804,
+      "step": 330
+    },
+    {
+      "epoch": 0.01568348732527837,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019988575477841985,
+      "loss": 1.3774,
+      "step": 331
+    },
+    {
+      "epoch": 0.015730869462212745,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019988504199807395,
+      "loss": 0.6024,
+      "step": 332
+    },
+    {
+      "epoch": 0.01577825159914712,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019988432700238424,
+      "loss": 1.3106,
+      "step": 333
+    },
+    {
+      "epoch": 0.015825633736081497,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019988360979136655,
+      "loss": 0.9213,
+      "step": 334
+    },
+    {
+      "epoch": 0.015873015873015872,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001998828903650368,
+      "loss": 1.7168,
+      "step": 335
+    },
+    {
+      "epoch": 0.015920398009950248,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001998821687234109,
+      "loss": 0.8373,
+      "step": 336
+    },
+    {
+      "epoch": 0.015967780146884623,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019988144486650491,
+      "loss": 1.3647,
+      "step": 337
+    },
+    {
+      "epoch": 0.016015162283819,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019988071879433483,
+      "loss": 0.8493,
+      "step": 338
+    },
+    {
+      "epoch": 0.016062544420753375,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001998799905069168,
+      "loss": 0.9083,
+      "step": 339
+    },
+    {
+      "epoch": 0.01610992655768775,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019987926000426703,
+      "loss": 1.4788,
+      "step": 340
+    },
+    {
+      "epoch": 0.016157308694622126,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001998785272864016,
+      "loss": 0.0531,
+      "step": 341
+    },
+    {
+      "epoch": 0.0162046908315565,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019987779235333683,
+      "loss": 0.5946,
+      "step": 342
+    },
+    {
+      "epoch": 0.01625207296849088,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000199877055205089,
+      "loss": 1.3108,
+      "step": 343
+    },
+    {
+      "epoch": 0.016299455105425256,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001998763158416745,
+      "loss": 0.6112,
+      "step": 344
+    },
+    {
+      "epoch": 0.016346837242359632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001998755742631097,
+      "loss": 0.7658,
+      "step": 345
+    },
+    {
+      "epoch": 0.016394219379294007,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000199874830469411,
+      "loss": 0.6044,
+      "step": 346
+    },
+    {
+      "epoch": 0.016441601516228383,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000199874084460595,
+      "loss": 1.118,
+      "step": 347
+    },
+    {
+      "epoch": 0.01648898365316276,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019987333623667814,
+      "loss": 0.4469,
+      "step": 348
+    },
+    {
+      "epoch": 0.016536365790097134,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001998725857976771,
+      "loss": 0.9925,
+      "step": 349
+    },
+    {
+      "epoch": 0.01658374792703151,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019987183314360848,
+      "loss": 1.8437,
+      "step": 350
+    },
+    {
+      "epoch": 0.016631130063965886,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019987107827448895,
+      "loss": 1.2108,
+      "step": 351
+    },
+    {
+      "epoch": 0.01667851220090026,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019987032119033528,
+      "loss": 1.193,
+      "step": 352
+    },
+    {
+      "epoch": 0.016725894337834637,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001998695618911643,
+      "loss": 1.4704,
+      "step": 353
+    },
+    {
+      "epoch": 0.016773276474769012,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019986880037699278,
+      "loss": 1.2658,
+      "step": 354
+    },
+    {
+      "epoch": 0.016820658611703388,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019986803664783767,
+      "loss": 0.5875,
+      "step": 355
+    },
+    {
+      "epoch": 0.016868040748637764,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019986727070371587,
+      "loss": 1.1199,
+      "step": 356
+    },
+    {
+      "epoch": 0.01691542288557214,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019986650254464437,
+      "loss": 0.5635,
+      "step": 357
+    },
+    {
+      "epoch": 0.016962805022506515,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019986573217064022,
+      "loss": 1.0004,
+      "step": 358
+    },
+    {
+      "epoch": 0.01701018715944089,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019986495958172054,
+      "loss": 1.3553,
+      "step": 359
+    },
+    {
+      "epoch": 0.017057569296375266,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019986418477790235,
+      "loss": 0.6413,
+      "step": 360
+    },
+    {
+      "epoch": 0.017104951433309642,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.00019986340775920297,
+      "loss": 0.0047,
+      "step": 361
+    },
+    {
+      "epoch": 0.017152333570244017,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019986262852563958,
+      "loss": 0.7254,
+      "step": 362
+    },
+    {
+      "epoch": 0.017199715707178393,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019986184707722945,
+      "loss": 0.4382,
+      "step": 363
+    },
+    {
+      "epoch": 0.01724709784411277,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019986106341398992,
+      "loss": 0.7554,
+      "step": 364
+    },
+    {
+      "epoch": 0.017294479981047144,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00019986027753593835,
+      "loss": 0.866,
+      "step": 365
+    },
+    {
+      "epoch": 0.01734186211798152,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019985948944309221,
+      "loss": 1.6714,
+      "step": 366
+    },
+    {
+      "epoch": 0.017389244254915896,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019985869913546894,
+      "loss": 0.4531,
+      "step": 367
+    },
+    {
+      "epoch": 0.01743662639185027,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019985790661308613,
+      "loss": 0.7205,
+      "step": 368
+    },
+    {
+      "epoch": 0.017484008528784647,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019985711187596125,
+      "loss": 0.3726,
+      "step": 369
+    },
+    {
+      "epoch": 0.017531390665719022,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019985631492411206,
+      "loss": 1.2968,
+      "step": 370
+    },
+    {
+      "epoch": 0.017578772802653398,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019985551575755613,
+      "loss": 0.7039,
+      "step": 371
+    },
+    {
+      "epoch": 0.017626154939587774,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001998547143763112,
+      "loss": 1.1429,
+      "step": 372
+    },
+    {
+      "epoch": 0.017673537076522153,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019985391078039514,
+      "loss": 0.865,
+      "step": 373
+    },
+    {
+      "epoch": 0.01772091921345653,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019985310496982564,
+      "loss": 1.5814,
+      "step": 374
+    },
+    {
+      "epoch": 0.017768301350390904,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019985229694462065,
+      "loss": 1.4083,
+      "step": 375
+    },
+    {
+      "epoch": 0.01781568348732528,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019985148670479804,
+      "loss": 0.3757,
+      "step": 376
+    },
+    {
+      "epoch": 0.017863065624259655,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019985067425037583,
+      "loss": 1.2033,
+      "step": 377
+    },
+    {
+      "epoch": 0.01791044776119403,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019984985958137203,
+      "loss": 1.2772,
+      "step": 378
+    },
+    {
+      "epoch": 0.017957829898128407,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001998490426978047,
+      "loss": 1.0075,
+      "step": 379
+    },
+    {
+      "epoch": 0.018005212035062782,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019984822359969196,
+      "loss": 0.7024,
+      "step": 380
+    },
+    {
+      "epoch": 0.018052594171997158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019984740228705196,
+      "loss": 1.2398,
+      "step": 381
+    },
+    {
+      "epoch": 0.018099976308931533,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019984657875990296,
+      "loss": 0.8402,
+      "step": 382
+    },
+    {
+      "epoch": 0.01814735844586591,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019984575301826315,
+      "loss": 1.4791,
+      "step": 383
+    },
+    {
+      "epoch": 0.018194740582800285,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019984492506215092,
+      "loss": 1.2764,
+      "step": 384
+    },
+    {
+      "epoch": 0.01824212271973466,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001998440948915846,
+      "loss": 0.9856,
+      "step": 385
+    },
+    {
+      "epoch": 0.018289504856669036,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001998432625065826,
+      "loss": 0.8699,
+      "step": 386
+    },
+    {
+      "epoch": 0.01833688699360341,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019984242790716339,
+      "loss": 0.9452,
+      "step": 387
+    },
+    {
+      "epoch": 0.018384269130537787,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00019984159109334547,
+      "loss": 0.0059,
+      "step": 388
+    },
+    {
+      "epoch": 0.018431651267472163,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019984075206514742,
+      "loss": 1.2753,
+      "step": 389
+    },
+    {
+      "epoch": 0.01847903340440654,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001998399108225878,
+      "loss": 1.1387,
+      "step": 390
+    },
+    {
+      "epoch": 0.018526415541340914,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019983906736568532,
+      "loss": 1.1879,
+      "step": 391
+    },
+    {
+      "epoch": 0.01857379767827529,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019983822169445867,
+      "loss": 0.5234,
+      "step": 392
+    },
+    {
+      "epoch": 0.018621179815209665,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019983737380892662,
+      "loss": 0.5158,
+      "step": 393
+    },
+    {
+      "epoch": 0.01866856195214404,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019983652370910796,
+      "loss": 0.9222,
+      "step": 394
+    },
+    {
+      "epoch": 0.018715944089078417,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019983567139502152,
+      "loss": 0.7296,
+      "step": 395
+    },
+    {
+      "epoch": 0.018763326226012792,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019983481686668627,
+      "loss": 0.8614,
+      "step": 396
+    },
+    {
+      "epoch": 0.018810708362947168,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019983396012412109,
+      "loss": 0.6279,
+      "step": 397
+    },
+    {
+      "epoch": 0.018858090499881543,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019983310116734502,
+      "loss": 1.1019,
+      "step": 398
+    },
+    {
+      "epoch": 0.01890547263681592,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001998322399963771,
+      "loss": 1.1183,
+      "step": 399
+    },
+    {
+      "epoch": 0.018952854773750295,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019983137661123642,
+      "loss": 0.9234,
+      "step": 400
+    },
+    {
+      "epoch": 0.01900023691068467,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019983051101194217,
+      "loss": 1.2332,
+      "step": 401
+    },
+    {
+      "epoch": 0.01904761904761905,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019982964319851352,
+      "loss": 0.9963,
+      "step": 402
+    },
+    {
+      "epoch": 0.019095001184553425,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001998287731709697,
+      "loss": 0.875,
+      "step": 403
+    },
+    {
+      "epoch": 0.0191423833214878,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019982790092933002,
+      "loss": 1.4477,
+      "step": 404
+    },
+    {
+      "epoch": 0.019189765458422176,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019982702647361385,
+      "loss": 1.0502,
+      "step": 405
+    },
+    {
+      "epoch": 0.019237147595356552,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019982614980384056,
+      "loss": 1.3743,
+      "step": 406
+    },
+    {
+      "epoch": 0.019284529732290927,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001998252709200296,
+      "loss": 0.9722,
+      "step": 407
+    },
+    {
+      "epoch": 0.019331911869225303,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019982438982220043,
+      "loss": 0.9775,
+      "step": 408
+    },
+    {
+      "epoch": 0.01937929400615968,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019982350651037264,
+      "loss": 0.4177,
+      "step": 409
+    },
+    {
+      "epoch": 0.019426676143094054,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019982262098456582,
+      "loss": 1.3982,
+      "step": 410
+    },
+    {
+      "epoch": 0.01947405828002843,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019982173324479955,
+      "loss": 0.9668,
+      "step": 411
+    },
+    {
+      "epoch": 0.019521440416962806,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001998208432910936,
+      "loss": 0.5722,
+      "step": 412
+    },
+    {
+      "epoch": 0.01956882255389718,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019981995112346764,
+      "loss": 0.0664,
+      "step": 413
+    },
+    {
+      "epoch": 0.019616204690831557,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019981905674194153,
+      "loss": 1.4617,
+      "step": 414
+    },
+    {
+      "epoch": 0.019663586827765932,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019981816014653502,
+      "loss": 1.0365,
+      "step": 415
+    },
+    {
+      "epoch": 0.019710968964700308,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019981726133726807,
+      "loss": 0.9083,
+      "step": 416
+    },
+    {
+      "epoch": 0.019758351101634684,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001998163603141606,
+      "loss": 0.5751,
+      "step": 417
+    },
+    {
+      "epoch": 0.01980573323856906,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00019981545707723256,
+      "loss": 0.3827,
+      "step": 418
+    },
+    {
+      "epoch": 0.019853115375503435,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019981455162650398,
+      "loss": 0.7891,
+      "step": 419
+    },
+    {
+      "epoch": 0.01990049751243781,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019981364396199497,
+      "loss": 1.0511,
+      "step": 420
+    },
+    {
+      "epoch": 0.019947879649372186,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019981273408372564,
+      "loss": 1.1038,
+      "step": 421
+    },
+    {
+      "epoch": 0.019995261786306562,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019981182199171622,
+      "loss": 1.2424,
+      "step": 422
+    },
+    {
+      "epoch": 0.020042643923240937,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001998109076859869,
+      "loss": 0.5426,
+      "step": 423
+    },
+    {
+      "epoch": 0.020090026060175313,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019980999116655794,
+      "loss": 1.1897,
+      "step": 424
+    },
+    {
+      "epoch": 0.02013740819710969,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019980907243344968,
+      "loss": 1.1586,
+      "step": 425
+    },
+    {
+      "epoch": 0.020184790334044064,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019980815148668251,
+      "loss": 0.6897,
+      "step": 426
+    },
+    {
+      "epoch": 0.02023217247097844,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001998072283262769,
+      "loss": 1.2089,
+      "step": 427
+    },
+    {
+      "epoch": 0.020279554607912816,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019980630295225323,
+      "loss": 0.7024,
+      "step": 428
+    },
+    {
+      "epoch": 0.02032693674484719,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019980537536463207,
+      "loss": 0.3366,
+      "step": 429
+    },
+    {
+      "epoch": 0.020374318881781567,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.000199804445563434,
+      "loss": 1.0042,
+      "step": 430
+    },
+    {
+      "epoch": 0.020421701018715942,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019980351354867963,
+      "loss": 0.5644,
+      "step": 431
+    },
+    {
+      "epoch": 0.02046908315565032,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019980257932038966,
+      "loss": 1.0856,
+      "step": 432
+    },
+    {
+      "epoch": 0.020516465292584697,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019980164287858475,
+      "loss": 0.4346,
+      "step": 433
+    },
+    {
+      "epoch": 0.020563847429519073,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019980070422328573,
+      "loss": 0.3099,
+      "step": 434
+    },
+    {
+      "epoch": 0.02061122956645345,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019979976335451338,
+      "loss": 0.991,
+      "step": 435
+    },
+    {
+      "epoch": 0.020658611703387824,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001997988202722886,
+      "loss": 1.0886,
+      "step": 436
+    },
+    {
+      "epoch": 0.0207059938403222,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019979787497663228,
+      "loss": 1.0824,
+      "step": 437
+    },
+    {
+      "epoch": 0.020753375977256575,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019979692746756536,
+      "loss": 1.2677,
+      "step": 438
+    },
+    {
+      "epoch": 0.02080075811419095,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019979597774510892,
+      "loss": 0.7553,
+      "step": 439
+    },
+    {
+      "epoch": 0.020848140251125327,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000199795025809284,
+      "loss": 1.0373,
+      "step": 440
+    },
+    {
+      "epoch": 0.020895522388059702,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019979407166011165,
+      "loss": 1.2302,
+      "step": 441
+    },
+    {
+      "epoch": 0.020942904524994078,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019979311529761312,
+      "loss": 0.3575,
+      "step": 442
+    },
+    {
+      "epoch": 0.020990286661928453,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019979215672180962,
+      "loss": 0.9975,
+      "step": 443
+    },
+    {
+      "epoch": 0.02103766879886283,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019979119593272236,
+      "loss": 0.4121,
+      "step": 444
+    },
+    {
+      "epoch": 0.021085050935797205,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019979023293037265,
+      "loss": 1.1058,
+      "step": 445
+    },
+    {
+      "epoch": 0.02113243307273158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019978926771478187,
+      "loss": 1.045,
+      "step": 446
+    },
+    {
+      "epoch": 0.021179815209665956,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019978830028597141,
+      "loss": 1.5996,
+      "step": 447
+    },
+    {
+      "epoch": 0.02122719734660033,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019978733064396277,
+      "loss": 0.6733,
+      "step": 448
+    },
+    {
+      "epoch": 0.021274579483534707,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019978635878877742,
+      "loss": 1.1244,
+      "step": 449
+    },
+    {
+      "epoch": 0.021321961620469083,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019978538472043692,
+      "loss": 1.5741,
+      "step": 450
+    },
+    {
+      "epoch": 0.02136934375740346,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019978440843896285,
+      "loss": 0.7887,
+      "step": 451
+    },
+    {
+      "epoch": 0.021416725894337834,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019978342994437688,
+      "loss": 0.4479,
+      "step": 452
+    },
+    {
+      "epoch": 0.02146410803127221,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019978244923670076,
+      "loss": 1.525,
+      "step": 453
+    },
+    {
+      "epoch": 0.021511490168206585,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019978146631595615,
+      "loss": 1.161,
+      "step": 454
+    },
+    {
+      "epoch": 0.02155887230514096,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019978048118216496,
+      "loss": 1.2118,
+      "step": 455
+    },
+    {
+      "epoch": 0.021606254442075336,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019977949383534894,
+      "loss": 0.6029,
+      "step": 456
+    },
+    {
+      "epoch": 0.021653636579009712,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019977850427553,
+      "loss": 0.9705,
+      "step": 457
+    },
+    {
+      "epoch": 0.021701018715944088,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001997775125027302,
+      "loss": 1.2315,
+      "step": 458
+    },
+    {
+      "epoch": 0.021748400852878463,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019977651851697136,
+      "loss": 1.1845,
+      "step": 459
+    },
+    {
+      "epoch": 0.02179578298981284,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00019977552231827564,
+      "loss": 0.4129,
+      "step": 460
+    },
+    {
+      "epoch": 0.021843165126747215,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019977452390666515,
+      "loss": 0.3194,
+      "step": 461
+    },
+    {
+      "epoch": 0.021890547263681594,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019977352328216197,
+      "loss": 0.784,
+      "step": 462
+    },
+    {
+      "epoch": 0.02193792940061597,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001997725204447883,
+      "loss": 1.1318,
+      "step": 463
+    },
+    {
+      "epoch": 0.021985311537550345,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019977151539456642,
+      "loss": 1.6012,
+      "step": 464
+    },
+    {
+      "epoch": 0.02203269367448472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001997705081315186,
+      "loss": 0.3294,
+      "step": 465
+    },
+    {
+      "epoch": 0.022080075811419096,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001997694986556672,
+      "loss": 0.5496,
+      "step": 466
+    },
+    {
+      "epoch": 0.022127457948353472,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019976848696703456,
+      "loss": 1.0092,
+      "step": 467
+    },
+    {
+      "epoch": 0.022174840085287847,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019976747306564314,
+      "loss": 0.9297,
+      "step": 468
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019976645695151546,
+      "loss": 1.193,
+      "step": 469
+    },
+    {
+      "epoch": 0.0222696043591566,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019976543862467404,
+      "loss": 1.4561,
+      "step": 470
+    },
+    {
+      "epoch": 0.022316986496090974,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001997644180851414,
+      "loss": 0.9823,
+      "step": 471
+    },
+    {
+      "epoch": 0.02236436863302535,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019976339533294028,
+      "loss": 0.5589,
+      "step": 472
+    },
+    {
+      "epoch": 0.022411750769959726,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019976237036809332,
+      "loss": 0.8081,
+      "step": 473
+    },
+    {
+      "epoch": 0.0224591329068941,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019976134319062323,
+      "loss": 0.6358,
+      "step": 474
+    },
+    {
+      "epoch": 0.022506515043828477,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001997603138005528,
+      "loss": 0.8958,
+      "step": 475
+    },
+    {
+      "epoch": 0.022553897180762852,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001997592821979049,
+      "loss": 1.0532,
+      "step": 476
+    },
+    {
+      "epoch": 0.022601279317697228,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019975824838270234,
+      "loss": 1.1853,
+      "step": 477
+    },
+    {
+      "epoch": 0.022648661454631604,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019975721235496811,
+      "loss": 0.4938,
+      "step": 478
+    },
+    {
+      "epoch": 0.02269604359156598,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019975617411472518,
+      "loss": 1.1591,
+      "step": 479
+    },
+    {
+      "epoch": 0.022743425728500355,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019975513366199654,
+      "loss": 1.3403,
+      "step": 480
+    },
+    {
+      "epoch": 0.02279080786543473,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001997540909968053,
+      "loss": 1.0315,
+      "step": 481
+    },
+    {
+      "epoch": 0.022838190002369106,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019975304611917456,
+      "loss": 0.3196,
+      "step": 482
+    },
+    {
+      "epoch": 0.022885572139303482,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019975199902912754,
+      "loss": 0.6876,
+      "step": 483
+    },
+    {
+      "epoch": 0.022932954276237857,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001997509497266874,
+      "loss": 1.6713,
+      "step": 484
+    },
+    {
+      "epoch": 0.022980336413172233,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019974989821187745,
+      "loss": 0.1393,
+      "step": 485
+    },
+    {
+      "epoch": 0.02302771855010661,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019974884448472103,
+      "loss": 0.2337,
+      "step": 486
+    },
+    {
+      "epoch": 0.023075100687040984,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019974778854524148,
+      "loss": 0.1739,
+      "step": 487
+    },
+    {
+      "epoch": 0.02312248282397536,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019974673039346223,
+      "loss": 1.2593,
+      "step": 488
+    },
+    {
+      "epoch": 0.023169864960909736,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019974567002940675,
+      "loss": 1.5246,
+      "step": 489
+    },
+    {
+      "epoch": 0.02321724709784411,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001997446074530985,
+      "loss": 0.9928,
+      "step": 490
+    },
+    {
+      "epoch": 0.023264629234778487,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019974354266456116,
+      "loss": 0.9088,
+      "step": 491
+    },
+    {
+      "epoch": 0.023312011371712866,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019974247566381824,
+      "loss": 1.3176,
+      "step": 492
+    },
+    {
+      "epoch": 0.02335939350864724,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019974140645089347,
+      "loss": 0.9397,
+      "step": 493
+    },
+    {
+      "epoch": 0.023406775645581617,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019974033502581054,
+      "loss": 0.8935,
+      "step": 494
+    },
+    {
+      "epoch": 0.023454157782515993,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019973926138859324,
+      "loss": 1.5524,
+      "step": 495
+    },
+    {
+      "epoch": 0.02350153991945037,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019973818553926535,
+      "loss": 1.1607,
+      "step": 496
+    },
+    {
+      "epoch": 0.023548922056384744,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019973710747785075,
+      "loss": 1.2721,
+      "step": 497
+    },
+    {
+      "epoch": 0.02359630419331912,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019973602720437336,
+      "loss": 0.7482,
+      "step": 498
+    },
+    {
+      "epoch": 0.023643686330253495,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019973494471885707,
+      "loss": 1.1923,
+      "step": 499
+    },
+    {
+      "epoch": 0.02369106846718787,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019973386002132597,
+      "loss": 1.0473,
+      "step": 500
+    },
+    {
+      "epoch": 0.023738450604122247,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019973277311180409,
+      "loss": 1.5173,
+      "step": 501
+    },
+    {
+      "epoch": 0.023785832741056622,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019973168399031548,
+      "loss": 0.5056,
+      "step": 502
+    },
+    {
+      "epoch": 0.023833214877990998,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019973059265688438,
+      "loss": 0.9238,
+      "step": 503
+    },
+    {
+      "epoch": 0.023880597014925373,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.000199729499111535,
+      "loss": 0.3908,
+      "step": 504
+    },
+    {
+      "epoch": 0.02392797915185975,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019972840335429152,
+      "loss": 1.0612,
+      "step": 505
+    },
+    {
+      "epoch": 0.023975361288794125,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019972730538517827,
+      "loss": 1.0104,
+      "step": 506
+    },
+    {
+      "epoch": 0.0240227434257285,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019972620520421964,
+      "loss": 1.2123,
+      "step": 507
+    },
+    {
+      "epoch": 0.024070125562662876,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019972510281144,
+      "loss": 1.2562,
+      "step": 508
+    },
+    {
+      "epoch": 0.02411750769959725,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019972399820686378,
+      "loss": 0.3947,
+      "step": 509
+    },
+    {
+      "epoch": 0.024164889836531627,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019972289139051551,
+      "loss": 0.5638,
+      "step": 510
+    },
+    {
+      "epoch": 0.024212271973466003,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019972178236241973,
+      "loss": 0.053,
+      "step": 511
+    },
+    {
+      "epoch": 0.02425965411040038,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019972067112260103,
+      "loss": 0.997,
+      "step": 512
+    },
+    {
+      "epoch": 0.024307036247334754,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001997195576710841,
+      "loss": 1.133,
+      "step": 513
+    },
+    {
+      "epoch": 0.02435441838426913,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019971844200789357,
+      "loss": 0.9191,
+      "step": 514
+    },
+    {
+      "epoch": 0.024401800521203505,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001997173241330542,
+      "loss": 0.9256,
+      "step": 515
+    },
+    {
+      "epoch": 0.02444918265813788,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001997162040465908,
+      "loss": 1.015,
+      "step": 516
+    },
+    {
+      "epoch": 0.024496564795072256,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019971508174852822,
+      "loss": 0.9148,
+      "step": 517
+    },
+    {
+      "epoch": 0.024543946932006632,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019971395723889133,
+      "loss": 1.2158,
+      "step": 518
+    },
+    {
+      "epoch": 0.024591329068941008,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001997128305177051,
+      "loss": 0.9604,
+      "step": 519
+    },
+    {
+      "epoch": 0.024638711205875383,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019971170158499443,
+      "loss": 1.173,
+      "step": 520
+    },
+    {
+      "epoch": 0.024686093342809762,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001997105704407845,
+      "loss": 1.0204,
+      "step": 521
+    },
+    {
+      "epoch": 0.024733475479744138,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001997094370851003,
+      "loss": 1.2085,
+      "step": 522
+    },
+    {
+      "epoch": 0.024780857616678514,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019970830151796697,
+      "loss": 0.9763,
+      "step": 523
+    },
+    {
+      "epoch": 0.02482823975361289,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019970716373940976,
+      "loss": 1.2341,
+      "step": 524
+    },
+    {
+      "epoch": 0.024875621890547265,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001997060237494538,
+      "loss": 0.5408,
+      "step": 525
+    },
+    {
+      "epoch": 0.02492300402748164,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0001997048815481245,
+      "loss": 0.0094,
+      "step": 526
+    },
+    {
+      "epoch": 0.024970386164416016,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001997037371354471,
+      "loss": 1.1377,
+      "step": 527
+    },
+    {
+      "epoch": 0.025017768301350392,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000199702590511447,
+      "loss": 0.3099,
+      "step": 528
+    },
+    {
+      "epoch": 0.025065150438284767,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019970144167614967,
+      "loss": 1.0221,
+      "step": 529
+    },
+    {
+      "epoch": 0.025112532575219143,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019970029062958054,
+      "loss": 0.4058,
+      "step": 530
+    },
+    {
+      "epoch": 0.02515991471215352,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019969913737176515,
+      "loss": 1.2623,
+      "step": 531
+    },
+    {
+      "epoch": 0.025207296849087894,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001996979819027291,
+      "loss": 1.618,
+      "step": 532
+    },
+    {
+      "epoch": 0.02525467898602227,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019969682422249803,
+      "loss": 0.4917,
+      "step": 533
+    },
+    {
+      "epoch": 0.025302061122956646,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019969566433109757,
+      "loss": 1.1174,
+      "step": 534
+    },
+    {
+      "epoch": 0.02534944325989102,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019969450222855347,
+      "loss": 1.2709,
+      "step": 535
+    },
+    {
+      "epoch": 0.025396825396825397,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001996933379148915,
+      "loss": 1.5649,
+      "step": 536
+    },
+    {
+      "epoch": 0.025444207533759772,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019969217139013745,
+      "loss": 0.5722,
+      "step": 537
+    },
+    {
+      "epoch": 0.025491589670694148,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019969100265431727,
+      "loss": 1.0741,
+      "step": 538
+    },
+    {
+      "epoch": 0.025538971807628524,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001996898317074568,
+      "loss": 0.2779,
+      "step": 539
+    },
+    {
+      "epoch": 0.0255863539445629,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019968865854958208,
+      "loss": 0.9093,
+      "step": 540
+    },
+    {
+      "epoch": 0.025633736081497275,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019968748318071908,
+      "loss": 1.1398,
+      "step": 541
+    },
+    {
+      "epoch": 0.02568111821843165,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001996863056008939,
+      "loss": 1.2883,
+      "step": 542
+    },
+    {
+      "epoch": 0.025728500355366026,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001996851258101326,
+      "loss": 1.5229,
+      "step": 543
+    },
+    {
+      "epoch": 0.025775882492300402,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019968394380846146,
+      "loss": 0.4906,
+      "step": 544
+    },
+    {
+      "epoch": 0.025823264629234777,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001996827595959066,
+      "loss": 0.8691,
+      "step": 545
+    },
+    {
+      "epoch": 0.025870646766169153,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019968157317249428,
+      "loss": 0.9585,
+      "step": 546
+    },
+    {
+      "epoch": 0.02591802890310353,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019968038453825084,
+      "loss": 0.9995,
+      "step": 547
+    },
+    {
+      "epoch": 0.025965411040037904,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019967919369320267,
+      "loss": 0.9304,
+      "step": 548
+    },
+    {
+      "epoch": 0.02601279317697228,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019967800063737617,
+      "loss": 0.2167,
+      "step": 549
+    },
+    {
+      "epoch": 0.026060175313906656,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019967680537079775,
+      "loss": 0.8259,
+      "step": 550
+    },
+    {
+      "epoch": 0.026107557450841035,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000199675607893494,
+      "loss": 0.0206,
+      "step": 551
+    },
+    {
+      "epoch": 0.02615493958777541,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001996744082054914,
+      "loss": 1.3086,
+      "step": 552
+    },
+    {
+      "epoch": 0.026202321724709786,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001996732063068166,
+      "loss": 1.187,
+      "step": 553
+    },
+    {
+      "epoch": 0.02624970386164416,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00019967200219749628,
+      "loss": 0.6433,
+      "step": 554
+    },
+    {
+      "epoch": 0.026297085998578537,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001996707958775571,
+      "loss": 0.8348,
+      "step": 555
+    },
+    {
+      "epoch": 0.026344468135512913,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019966958734702584,
+      "loss": 1.079,
+      "step": 556
+    },
+    {
+      "epoch": 0.02639185027244729,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019966837660592926,
+      "loss": 1.209,
+      "step": 557
+    },
+    {
+      "epoch": 0.026439232409381664,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001996671636542943,
+      "loss": 0.5099,
+      "step": 558
+    },
+    {
+      "epoch": 0.02648661454631604,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001996659484921478,
+      "loss": 1.0275,
+      "step": 559
+    },
+    {
+      "epoch": 0.026533996683250415,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019966473111951669,
+      "loss": 0.6466,
+      "step": 560
+    },
+    {
+      "epoch": 0.02658137882018479,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000199663511536428,
+      "loss": 1.2751,
+      "step": 561
+    },
+    {
+      "epoch": 0.026628760957119166,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001996622897429088,
+      "loss": 0.6194,
+      "step": 562
+    },
+    {
+      "epoch": 0.026676143094053542,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019966106573898618,
+      "loss": 0.6814,
+      "step": 563
+    },
+    {
+      "epoch": 0.026723525230987918,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019965983952468727,
+      "loss": 0.4884,
+      "step": 564
+    },
+    {
+      "epoch": 0.026770907367922293,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019965861110003927,
+      "loss": 0.7638,
+      "step": 565
+    },
+    {
+      "epoch": 0.02681828950485667,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019965738046506945,
+      "loss": 1.5112,
+      "step": 566
+    },
+    {
+      "epoch": 0.026865671641791045,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000199656147619805,
+      "loss": 0.6674,
+      "step": 567
+    },
+    {
+      "epoch": 0.02691305377872542,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019965491256427345,
+      "loss": 1.3322,
+      "step": 568
+    },
+    {
+      "epoch": 0.026960435915659796,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000199653675298502,
+      "loss": 1.0886,
+      "step": 569
+    },
+    {
+      "epoch": 0.02700781805259417,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019965243582251824,
+      "loss": 0.4547,
+      "step": 570
+    },
+    {
+      "epoch": 0.027055200189528547,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019965119413634956,
+      "loss": 0.7786,
+      "step": 571
+    },
+    {
+      "epoch": 0.027102582326462923,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001996499502400236,
+      "loss": 1.0901,
+      "step": 572
+    },
+    {
+      "epoch": 0.0271499644633973,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019964870413356783,
+      "loss": 0.6203,
+      "step": 573
+    },
+    {
+      "epoch": 0.027197346600331674,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019964745581700993,
+      "loss": 0.309,
+      "step": 574
+    },
+    {
+      "epoch": 0.02724472873726605,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019964620529037763,
+      "loss": 0.8369,
+      "step": 575
+    },
+    {
+      "epoch": 0.027292110874200425,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001996449525536986,
+      "loss": 1.0702,
+      "step": 576
+    },
+    {
+      "epoch": 0.0273394930111348,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019964369760700073,
+      "loss": 0.7614,
+      "step": 577
+    },
+    {
+      "epoch": 0.027386875148069176,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001996424404503117,
+      "loss": 0.8197,
+      "step": 578
+    },
+    {
+      "epoch": 0.027434257285003552,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019964118108365954,
+      "loss": 0.9399,
+      "step": 579
+    },
+    {
+      "epoch": 0.027481639421937928,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001996399195070721,
+      "loss": 1.1644,
+      "step": 580
+    },
+    {
+      "epoch": 0.027529021558872307,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019963865572057734,
+      "loss": 1.0886,
+      "step": 581
+    },
+    {
+      "epoch": 0.027576403695806682,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019963738972420336,
+      "loss": 1.1885,
+      "step": 582
+    },
+    {
+      "epoch": 0.027623785832741058,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019963612151797819,
+      "loss": 0.2672,
+      "step": 583
+    },
+    {
+      "epoch": 0.027671167969675434,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019963485110193,
+      "loss": 1.0301,
+      "step": 584
+    },
+    {
+      "epoch": 0.02771855010660981,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019963357847608692,
+      "loss": 0.1862,
+      "step": 585
+    },
+    {
+      "epoch": 0.027765932243544185,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001996323036404772,
+      "loss": 0.9836,
+      "step": 586
+    },
+    {
+      "epoch": 0.02781331438047856,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019963102659512912,
+      "loss": 1.0559,
+      "step": 587
+    },
+    {
+      "epoch": 0.027860696517412936,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019962974734007095,
+      "loss": 0.7318,
+      "step": 588
+    },
+    {
+      "epoch": 0.027908078654347312,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019962846587533113,
+      "loss": 0.8617,
+      "step": 589
+    },
+    {
+      "epoch": 0.027955460791281687,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019962718220093806,
+      "loss": 1.2558,
+      "step": 590
+    },
+    {
+      "epoch": 0.028002842928216063,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001996258963169202,
+      "loss": 1.2238,
+      "step": 591
+    },
+    {
+      "epoch": 0.02805022506515044,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019962460822330608,
+      "loss": 0.9095,
+      "step": 592
+    },
+    {
+      "epoch": 0.028097607202084814,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019962331792012426,
+      "loss": 1.1602,
+      "step": 593
+    },
+    {
+      "epoch": 0.02814498933901919,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001996220254074034,
+      "loss": 0.6363,
+      "step": 594
+    },
+    {
+      "epoch": 0.028192371475953566,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019962073068517205,
+      "loss": 0.1952,
+      "step": 595
+    },
+    {
+      "epoch": 0.02823975361288794,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001996194337534591,
+      "loss": 0.2863,
+      "step": 596
+    },
+    {
+      "epoch": 0.028287135749822317,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019961813461229314,
+      "loss": 1.1668,
+      "step": 597
+    },
+    {
+      "epoch": 0.028334517886756692,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001996168332617031,
+      "loss": 1.333,
+      "step": 598
+    },
+    {
+      "epoch": 0.028381900023691068,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019961552970171778,
+      "loss": 1.0673,
+      "step": 599
+    },
+    {
+      "epoch": 0.028429282160625444,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019961422393236617,
+      "loss": 1.4091,
+      "step": 600
+    },
+    {
+      "epoch": 0.02847666429755982,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019961291595367714,
+      "loss": 1.3616,
+      "step": 601
+    },
+    {
+      "epoch": 0.028524046434494195,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019961160576567973,
+      "loss": 1.1052,
+      "step": 602
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019961029336840302,
+      "loss": 0.095,
+      "step": 603
+    },
+    {
+      "epoch": 0.028618810708362946,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001996089787618761,
+      "loss": 1.1025,
+      "step": 604
+    },
+    {
+      "epoch": 0.028666192845297322,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019960766194612815,
+      "loss": 0.6186,
+      "step": 605
+    },
+    {
+      "epoch": 0.028713574982231697,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001996063429211883,
+      "loss": 0.8258,
+      "step": 606
+    },
+    {
+      "epoch": 0.028760957119166073,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001996050216870859,
+      "loss": 1.029,
+      "step": 607
+    },
+    {
+      "epoch": 0.02880833925610045,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001996036982438502,
+      "loss": 0.6272,
+      "step": 608
+    },
+    {
+      "epoch": 0.028855721393034824,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001996023725915106,
+      "loss": 0.7262,
+      "step": 609
+    },
+    {
+      "epoch": 0.0289031035299692,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019960104473009643,
+      "loss": 1.2872,
+      "step": 610
+    },
+    {
+      "epoch": 0.02895048566690358,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001995997146596372,
+      "loss": 0.0761,
+      "step": 611
+    },
+    {
+      "epoch": 0.028997867803837955,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001995983823801624,
+      "loss": 1.1936,
+      "step": 612
+    },
+    {
+      "epoch": 0.02904524994077233,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019959704789170152,
+      "loss": 1.2569,
+      "step": 613
+    },
+    {
+      "epoch": 0.029092632077706706,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001995957111942842,
+      "loss": 0.5025,
+      "step": 614
+    },
+    {
+      "epoch": 0.02914001421464108,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019959437228794013,
+      "loss": 0.9133,
+      "step": 615
+    },
+    {
+      "epoch": 0.029187396351575457,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019959303117269897,
+      "loss": 0.635,
+      "step": 616
+    },
+    {
+      "epoch": 0.029234778488509833,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019959168784859044,
+      "loss": 0.3316,
+      "step": 617
+    },
+    {
+      "epoch": 0.02928216062544421,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019959034231564434,
+      "loss": 0.7864,
+      "step": 618
+    },
+    {
+      "epoch": 0.029329542762378584,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019958899457389056,
+      "loss": 0.9831,
+      "step": 619
+    },
+    {
+      "epoch": 0.02937692489931296,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019958764462335894,
+      "loss": 1.1185,
+      "step": 620
+    },
+    {
+      "epoch": 0.029424307036247335,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019958629246407945,
+      "loss": 0.511,
+      "step": 621
+    },
+    {
+      "epoch": 0.02947168917318171,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019958493809608206,
+      "loss": 0.1045,
+      "step": 622
+    },
+    {
+      "epoch": 0.029519071310116086,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001995835815193968,
+      "loss": 0.8508,
+      "step": 623
+    },
+    {
+      "epoch": 0.029566453447050462,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001995822227340538,
+      "loss": 1.005,
+      "step": 624
+    },
+    {
+      "epoch": 0.029613835583984838,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019958086174008314,
+      "loss": 0.9452,
+      "step": 625
+    },
+    {
+      "epoch": 0.029661217720919213,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 0.00019957949853751506,
+      "loss": 0.0025,
+      "step": 626
+    },
+    {
+      "epoch": 0.02970859985785359,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019957813312637977,
+      "loss": 0.6284,
+      "step": 627
+    },
+    {
+      "epoch": 0.029755981994787965,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019957676550670753,
+      "loss": 0.4623,
+      "step": 628
+    },
+    {
+      "epoch": 0.02980336413172234,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019957539567852872,
+      "loss": 1.3487,
+      "step": 629
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019957402364187367,
+      "loss": 0.3309,
+      "step": 630
+    },
+    {
+      "epoch": 0.02989812840559109,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019957264939677287,
+      "loss": 0.4242,
+      "step": 631
+    },
+    {
+      "epoch": 0.029945510542525467,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019957127294325676,
+      "loss": 0.2819,
+      "step": 632
+    },
+    {
+      "epoch": 0.029992892679459843,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019956989428135584,
+      "loss": 0.5646,
+      "step": 633
+    },
+    {
+      "epoch": 0.03004027481639422,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019956851341110075,
+      "loss": 1.522,
+      "step": 634
+    },
+    {
+      "epoch": 0.030087656953328594,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019956713033252211,
+      "loss": 1.0439,
+      "step": 635
+    },
+    {
+      "epoch": 0.03013503909026297,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019956574504565054,
+      "loss": 0.4926,
+      "step": 636
+    },
+    {
+      "epoch": 0.030182421227197345,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001995643575505168,
+      "loss": 0.0841,
+      "step": 637
+    },
+    {
+      "epoch": 0.03022980336413172,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00019956296784715168,
+      "loss": 0.0148,
+      "step": 638
+    },
+    {
+      "epoch": 0.030277185501066096,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019956157593558596,
+      "loss": 0.3761,
+      "step": 639
+    },
+    {
+      "epoch": 0.030324567638000476,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019956018181585054,
+      "loss": 1.1064,
+      "step": 640
+    },
+    {
+      "epoch": 0.03037194977493485,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019955878548797636,
+      "loss": 0.987,
+      "step": 641
+    },
+    {
+      "epoch": 0.030419331911869227,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019955738695199432,
+      "loss": 0.3002,
+      "step": 642
+    },
+    {
+      "epoch": 0.030466714048803602,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019955598620793552,
+      "loss": 0.7511,
+      "step": 643
+    },
+    {
+      "epoch": 0.030514096185737978,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019955458325583096,
+      "loss": 0.936,
+      "step": 644
+    },
+    {
+      "epoch": 0.030561478322672354,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019955317809571178,
+      "loss": 1.0025,
+      "step": 645
+    },
+    {
+      "epoch": 0.03060886045960673,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001995517707276092,
+      "loss": 1.0406,
+      "step": 646
+    },
+    {
+      "epoch": 0.030656242596541105,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019955036115155435,
+      "loss": 1.1344,
+      "step": 647
+    },
+    {
+      "epoch": 0.03070362473347548,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001995489493675785,
+      "loss": 0.1149,
+      "step": 648
+    },
+    {
+      "epoch": 0.030751006870409856,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019954753537571303,
+      "loss": 1.2429,
+      "step": 649
+    },
+    {
+      "epoch": 0.030798389007344232,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019954611917598922,
+      "loss": 0.0303,
+      "step": 650
+    },
+    {
+      "epoch": 0.030845771144278607,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019954470076843854,
+      "loss": 0.8687,
+      "step": 651
+    },
+    {
+      "epoch": 0.030893153281212983,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019954328015309243,
+      "loss": 0.2572,
+      "step": 652
+    },
+    {
+      "epoch": 0.03094053541814736,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001995418573299824,
+      "loss": 0.5588,
+      "step": 653
+    },
+    {
+      "epoch": 0.030987917555081734,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019954043229913996,
+      "loss": 0.1976,
+      "step": 654
+    },
+    {
+      "epoch": 0.03103529969201611,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001995390050605968,
+      "loss": 0.5403,
+      "step": 655
+    },
+    {
+      "epoch": 0.031082681828950486,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019953757561438454,
+      "loss": 1.1379,
+      "step": 656
+    },
+    {
+      "epoch": 0.03113006396588486,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019953614396053487,
+      "loss": 0.4124,
+      "step": 657
+    },
+    {
+      "epoch": 0.031177446102819237,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019953471009907952,
+      "loss": 1.3503,
+      "step": 658
+    },
+    {
+      "epoch": 0.031224828239753612,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019953327403005034,
+      "loss": 0.8431,
+      "step": 659
+    },
+    {
+      "epoch": 0.03127221037668799,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019953183575347914,
+      "loss": 0.8172,
+      "step": 660
+    },
+    {
+      "epoch": 0.03131959251362237,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019953039526939784,
+      "loss": 0.2071,
+      "step": 661
+    },
+    {
+      "epoch": 0.03136697465055674,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019952895257783842,
+      "loss": 0.2853,
+      "step": 662
+    },
+    {
+      "epoch": 0.03141435678749112,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019952750767883283,
+      "loss": 0.8138,
+      "step": 663
+    },
+    {
+      "epoch": 0.03146173892442549,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019952606057241313,
+      "loss": 1.1287,
+      "step": 664
+    },
+    {
+      "epoch": 0.03150912106135987,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001995246112586114,
+      "loss": 0.8817,
+      "step": 665
+    },
+    {
+      "epoch": 0.03155650319829424,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019952315973745984,
+      "loss": 0.6365,
+      "step": 666
+    },
+    {
+      "epoch": 0.03160388533522862,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019952170600899058,
+      "loss": 1.1044,
+      "step": 667
+    },
+    {
+      "epoch": 0.03165126747216299,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019952025007323587,
+      "loss": 0.631,
+      "step": 668
+    },
+    {
+      "epoch": 0.03169864960909737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019951879193022806,
+      "loss": 0.6363,
+      "step": 669
+    },
+    {
+      "epoch": 0.031746031746031744,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001995173315799994,
+      "loss": 0.662,
+      "step": 670
+    },
+    {
+      "epoch": 0.03179341388296612,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019951586902258237,
+      "loss": 0.2984,
+      "step": 671
+    },
+    {
+      "epoch": 0.031840796019900496,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019951440425800934,
+      "loss": 0.2706,
+      "step": 672
+    },
+    {
+      "epoch": 0.031888178156834875,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019951293728631282,
+      "loss": 0.9611,
+      "step": 673
+    },
+    {
+      "epoch": 0.03193556029376925,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019951146810752535,
+      "loss": 0.2135,
+      "step": 674
+    },
+    {
+      "epoch": 0.031982942430703626,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001995099967216795,
+      "loss": 0.224,
+      "step": 675
+    },
+    {
+      "epoch": 0.032030324567638,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019950852312880795,
+      "loss": 0.7542,
+      "step": 676
+    },
+    {
+      "epoch": 0.03207770670457238,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001995070473289433,
+      "loss": 0.7266,
+      "step": 677
+    },
+    {
+      "epoch": 0.03212508884150675,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001995055693221184,
+      "loss": 0.2165,
+      "step": 678
+    },
+    {
+      "epoch": 0.03217247097844113,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001995040891083659,
+      "loss": 1.2225,
+      "step": 679
+    },
+    {
+      "epoch": 0.0322198531153755,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019950260668771873,
+      "loss": 1.2573,
+      "step": 680
+    },
+    {
+      "epoch": 0.03226723525230988,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001995011220602097,
+      "loss": 0.1674,
+      "step": 681
+    },
+    {
+      "epoch": 0.03231461738924425,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019949963522587178,
+      "loss": 0.4874,
+      "step": 682
+    },
+    {
+      "epoch": 0.03236199952617863,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019949814618473792,
+      "loss": 1.2075,
+      "step": 683
+    },
+    {
+      "epoch": 0.032409381663113,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019949665493684119,
+      "loss": 1.2133,
+      "step": 684
+    },
+    {
+      "epoch": 0.03245676380004738,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019949516148221462,
+      "loss": 1.1786,
+      "step": 685
+    },
+    {
+      "epoch": 0.03250414593698176,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019949366582089134,
+      "loss": 1.2238,
+      "step": 686
+    },
+    {
+      "epoch": 0.03255152807391613,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019949216795290452,
+      "loss": 0.5681,
+      "step": 687
+    },
+    {
+      "epoch": 0.03259891021085051,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019949066787828737,
+      "loss": 1.0104,
+      "step": 688
+    },
+    {
+      "epoch": 0.032646292347784885,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001994891655970732,
+      "loss": 1.0594,
+      "step": 689
+    },
+    {
+      "epoch": 0.032693674484719264,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019948766110929533,
+      "loss": 1.0748,
+      "step": 690
+    },
+    {
+      "epoch": 0.032741056621653636,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019948615441498708,
+      "loss": 0.7086,
+      "step": 691
+    },
+    {
+      "epoch": 0.032788438758588015,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001994846455141819,
+      "loss": 0.9541,
+      "step": 692
+    },
+    {
+      "epoch": 0.03283582089552239,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019948313440691325,
+      "loss": 1.1669,
+      "step": 693
+    },
+    {
+      "epoch": 0.032883203032456766,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019948162109321464,
+      "loss": 1.1247,
+      "step": 694
+    },
+    {
+      "epoch": 0.03293058516939114,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019948010557311964,
+      "loss": 1.3189,
+      "step": 695
+    },
+    {
+      "epoch": 0.03297796730632552,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019947858784666187,
+      "loss": 0.8425,
+      "step": 696
+    },
+    {
+      "epoch": 0.03302534944325989,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019947706791387498,
+      "loss": 1.0511,
+      "step": 697
+    },
+    {
+      "epoch": 0.03307273158019427,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019947554577479267,
+      "loss": 0.6043,
+      "step": 698
+    },
+    {
+      "epoch": 0.03312011371712864,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019947402142944868,
+      "loss": 0.9978,
+      "step": 699
+    },
+    {
+      "epoch": 0.03316749585406302,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019947249487787692,
+      "loss": 1.0231,
+      "step": 700
+    },
+    {
+      "epoch": 0.03321487799099739,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019947096612011112,
+      "loss": 0.7844,
+      "step": 701
+    },
+    {
+      "epoch": 0.03326226012793177,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019946943515618525,
+      "loss": 0.6324,
+      "step": 702
+    },
+    {
+      "epoch": 0.03330964226486614,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001994679019861333,
+      "loss": 0.88,
+      "step": 703
+    },
+    {
+      "epoch": 0.03335702440180052,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001994663666099892,
+      "loss": 0.0794,
+      "step": 704
+    },
+    {
+      "epoch": 0.033404406538734895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019946482902778704,
+      "loss": 1.3594,
+      "step": 705
+    },
+    {
+      "epoch": 0.033451788675669274,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019946328923956092,
+      "loss": 1.3723,
+      "step": 706
+    },
+    {
+      "epoch": 0.033499170812603646,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019946174724534498,
+      "loss": 0.7756,
+      "step": 707
+    },
+    {
+      "epoch": 0.033546552949538025,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019946020304517347,
+      "loss": 0.5788,
+      "step": 708
+    },
+    {
+      "epoch": 0.0335939350864724,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019945865663908055,
+      "loss": 0.1196,
+      "step": 709
+    },
+    {
+      "epoch": 0.033641317223406776,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00019945710802710056,
+      "loss": 0.7798,
+      "step": 710
+    },
+    {
+      "epoch": 0.03368869936034115,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019945555720926787,
+      "loss": 1.0431,
+      "step": 711
+    },
+    {
+      "epoch": 0.03373608149727553,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019945400418561686,
+      "loss": 0.2399,
+      "step": 712
+    },
+    {
+      "epoch": 0.0337834636342099,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019945244895618194,
+      "loss": 0.7883,
+      "step": 713
+    },
+    {
+      "epoch": 0.03383084577114428,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019945089152099765,
+      "loss": 0.5802,
+      "step": 714
+    },
+    {
+      "epoch": 0.03387822790807866,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019944933188009855,
+      "loss": 1.1913,
+      "step": 715
+    },
+    {
+      "epoch": 0.03392561004501303,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019944777003351916,
+      "loss": 0.844,
+      "step": 716
+    },
+    {
+      "epoch": 0.03397299218194741,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019944620598129418,
+      "loss": 1.5475,
+      "step": 717
+    },
+    {
+      "epoch": 0.03402037431888178,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019944463972345827,
+      "loss": 1.2742,
+      "step": 718
+    },
+    {
+      "epoch": 0.03406775645581616,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019944307126004614,
+      "loss": 0.3556,
+      "step": 719
+    },
+    {
+      "epoch": 0.03411513859275053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019944150059109266,
+      "loss": 0.7456,
+      "step": 720
+    },
+    {
+      "epoch": 0.03416252072968491,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001994399277166326,
+      "loss": 1.1194,
+      "step": 721
+    },
+    {
+      "epoch": 0.034209902866619284,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019943835263670084,
+      "loss": 1.0139,
+      "step": 722
+    },
+    {
+      "epoch": 0.03425728500355366,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001994367753513324,
+      "loss": 1.2507,
+      "step": 723
+    },
+    {
+      "epoch": 0.034304667140488035,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019943519586056212,
+      "loss": 0.8208,
+      "step": 724
+    },
+    {
+      "epoch": 0.034352049277422414,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019943361416442515,
+      "loss": 1.0066,
+      "step": 725
+    },
+    {
+      "epoch": 0.034399431414356786,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001994320302629565,
+      "loss": 1.3527,
+      "step": 726
+    },
+    {
+      "epoch": 0.034446813551291165,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019943044415619138,
+      "loss": 0.8505,
+      "step": 727
+    },
+    {
+      "epoch": 0.03449419568822554,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019942885584416488,
+      "loss": 0.097,
+      "step": 728
+    },
+    {
+      "epoch": 0.034541577825159916,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019942726532691228,
+      "loss": 0.2142,
+      "step": 729
+    },
+    {
+      "epoch": 0.03458895996209429,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019942567260446885,
+      "loss": 0.8172,
+      "step": 730
+    },
+    {
+      "epoch": 0.03463634209902867,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001994240776768699,
+      "loss": 0.2967,
+      "step": 731
+    },
+    {
+      "epoch": 0.03468372423596304,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001994224805441508,
+      "loss": 0.0132,
+      "step": 732
+    },
+    {
+      "epoch": 0.03473110637289742,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019942088120634694,
+      "loss": 1.1554,
+      "step": 733
+    },
+    {
+      "epoch": 0.03477848850983179,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019941927966349388,
+      "loss": 0.7384,
+      "step": 734
+    },
+    {
+      "epoch": 0.03482587064676617,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001994176759156271,
+      "loss": 1.2028,
+      "step": 735
+    },
+    {
+      "epoch": 0.03487325278370054,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019941606996278215,
+      "loss": 0.3103,
+      "step": 736
+    },
+    {
+      "epoch": 0.03492063492063492,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019941446180499466,
+      "loss": 0.557,
+      "step": 737
+    },
+    {
+      "epoch": 0.034968017057569294,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019941285144230029,
+      "loss": 0.9375,
+      "step": 738
+    },
+    {
+      "epoch": 0.03501539919450367,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019941123887473475,
+      "loss": 1.1975,
+      "step": 739
+    },
+    {
+      "epoch": 0.035062781331438045,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019940962410233386,
+      "loss": 0.589,
+      "step": 740
+    },
+    {
+      "epoch": 0.035110163468372424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001994080071251334,
+      "loss": 1.0395,
+      "step": 741
+    },
+    {
+      "epoch": 0.035157545605306796,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001994063879431692,
+      "loss": 1.1947,
+      "step": 742
+    },
+    {
+      "epoch": 0.035204927742241175,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001994047665564772,
+      "loss": 1.1165,
+      "step": 743
+    },
+    {
+      "epoch": 0.03525230987917555,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019940314296509337,
+      "loss": 0.31,
+      "step": 744
+    },
+    {
+      "epoch": 0.035299692016109926,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019940151716905371,
+      "loss": 0.68,
+      "step": 745
+    },
+    {
+      "epoch": 0.035347074153044306,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019939988916839425,
+      "loss": 0.7773,
+      "step": 746
+    },
+    {
+      "epoch": 0.03539445628997868,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019939825896315115,
+      "loss": 1.0553,
+      "step": 747
+    },
+    {
+      "epoch": 0.03544183842691306,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019939662655336053,
+      "loss": 0.7158,
+      "step": 748
+    },
+    {
+      "epoch": 0.03548922056384743,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019939499193905862,
+      "loss": 0.2265,
+      "step": 749
+    },
+    {
+      "epoch": 0.03553660270078181,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019939335512028164,
+      "loss": 0.96,
+      "step": 750
+    },
+    {
+      "epoch": 0.03558398483771618,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001993917160970659,
+      "loss": 0.7222,
+      "step": 751
+    },
+    {
+      "epoch": 0.03563136697465056,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001993900748694478,
+      "loss": 1.0403,
+      "step": 752
+    },
+    {
+      "epoch": 0.03567874911158493,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019938843143746369,
+      "loss": 0.9728,
+      "step": 753
+    },
+    {
+      "epoch": 0.03572613124851931,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019938678580115005,
+      "loss": 1.0014,
+      "step": 754
+    },
+    {
+      "epoch": 0.03577351338545368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019938513796054333,
+      "loss": 1.2322,
+      "step": 755
+    },
+    {
+      "epoch": 0.03582089552238806,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019938348791568013,
+      "loss": 1.2852,
+      "step": 756
+    },
+    {
+      "epoch": 0.035868277659322434,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019938183566659703,
+      "loss": 1.0591,
+      "step": 757
+    },
+    {
+      "epoch": 0.03591565979625681,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019938018121333064,
+      "loss": 1.7937,
+      "step": 758
+    },
+    {
+      "epoch": 0.035963041933191185,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019937852455591772,
+      "loss": 1.0624,
+      "step": 759
+    },
+    {
+      "epoch": 0.036010424070125564,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019937686569439496,
+      "loss": 0.564,
+      "step": 760
+    },
+    {
+      "epoch": 0.036057806207059936,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019937520462879918,
+      "loss": 0.9257,
+      "step": 761
+    },
+    {
+      "epoch": 0.036105188343994316,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019937354135916721,
+      "loss": 1.023,
+      "step": 762
+    },
+    {
+      "epoch": 0.03615257048092869,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00019937187588553595,
+      "loss": 0.7796,
+      "step": 763
+    },
+    {
+      "epoch": 0.03619995261786307,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019937020820794233,
+      "loss": 0.0903,
+      "step": 764
+    },
+    {
+      "epoch": 0.03624733475479744,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019936853832642332,
+      "loss": 0.5821,
+      "step": 765
+    },
+    {
+      "epoch": 0.03629471689173182,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019936686624101596,
+      "loss": 0.3511,
+      "step": 766
+    },
+    {
+      "epoch": 0.03634209902866619,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001993651919517574,
+      "loss": 0.314,
+      "step": 767
+    },
+    {
+      "epoch": 0.03638948116560057,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019936351545868467,
+      "loss": 0.8487,
+      "step": 768
+    },
+    {
+      "epoch": 0.03643686330253494,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000199361836761835,
+      "loss": 0.6685,
+      "step": 769
+    },
+    {
+      "epoch": 0.03648424543946932,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001993601558612457,
+      "loss": 0.7398,
+      "step": 770
+    },
+    {
+      "epoch": 0.03653162757640369,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019935847275695393,
+      "loss": 1.2331,
+      "step": 771
+    },
+    {
+      "epoch": 0.03657900971333807,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019935678744899705,
+      "loss": 1.2426,
+      "step": 772
+    },
+    {
+      "epoch": 0.036626391850272444,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019935509993741245,
+      "loss": 0.0053,
+      "step": 773
+    },
+    {
+      "epoch": 0.03667377398720682,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001993534102222376,
+      "loss": 0.9312,
+      "step": 774
+    },
+    {
+      "epoch": 0.0367211561241412,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001993517183035099,
+      "loss": 1.0097,
+      "step": 775
+    },
+    {
+      "epoch": 0.036768538261075574,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00019935002418126693,
+      "loss": 0.004,
+      "step": 776
+    },
+    {
+      "epoch": 0.03681592039800995,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019934832785554625,
+      "loss": 1.0161,
+      "step": 777
+    },
+    {
+      "epoch": 0.036863302534944326,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019934662932638548,
+      "loss": 0.9238,
+      "step": 778
+    },
+    {
+      "epoch": 0.036910684671878705,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019934492859382226,
+      "loss": 0.1305,
+      "step": 779
+    },
+    {
+      "epoch": 0.03695806680881308,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019934322565789438,
+      "loss": 1.173,
+      "step": 780
+    },
+    {
+      "epoch": 0.037005448945747456,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019934152051863957,
+      "loss": 0.2643,
+      "step": 781
+    },
+    {
+      "epoch": 0.03705283108268183,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019933981317609562,
+      "loss": 1.4096,
+      "step": 782
+    },
+    {
+      "epoch": 0.03710021321961621,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019933810363030046,
+      "loss": 0.2747,
+      "step": 783
+    },
+    {
+      "epoch": 0.03714759535655058,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019933639188129195,
+      "loss": 0.9685,
+      "step": 784
+    },
+    {
+      "epoch": 0.03719497749348496,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019933467792910805,
+      "loss": 1.578,
+      "step": 785
+    },
+    {
+      "epoch": 0.03724235963041933,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019933296177378684,
+      "loss": 1.4234,
+      "step": 786
+    },
+    {
+      "epoch": 0.03728974176735371,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019933124341536633,
+      "loss": 1.0826,
+      "step": 787
+    },
+    {
+      "epoch": 0.03733712390428808,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019932952285388463,
+      "loss": 0.0578,
+      "step": 788
+    },
+    {
+      "epoch": 0.03738450604122246,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019932780008937993,
+      "loss": 1.0469,
+      "step": 789
+    },
+    {
+      "epoch": 0.03743188817815683,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019932607512189042,
+      "loss": 0.0993,
+      "step": 790
+    },
+    {
+      "epoch": 0.03747927031509121,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019932434795145437,
+      "loss": 0.4258,
+      "step": 791
+    },
+    {
+      "epoch": 0.037526652452025584,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001993226185781101,
+      "loss": 0.0317,
+      "step": 792
+    },
+    {
+      "epoch": 0.03757403458895996,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001993208870018959,
+      "loss": 0.2297,
+      "step": 793
+    },
+    {
+      "epoch": 0.037621416725894335,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019931915322285025,
+      "loss": 0.0231,
+      "step": 794
+    },
+    {
+      "epoch": 0.037668798862828715,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019931741724101153,
+      "loss": 0.7809,
+      "step": 795
+    },
+    {
+      "epoch": 0.03771618099976309,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019931567905641834,
+      "loss": 0.8968,
+      "step": 796
+    },
+    {
+      "epoch": 0.037763563136697466,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019931393866910914,
+      "loss": 0.3473,
+      "step": 797
+    },
+    {
+      "epoch": 0.03781094527363184,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019931219607912258,
+      "loss": 0.0396,
+      "step": 798
+    },
+    {
+      "epoch": 0.03785832741056622,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019931045128649725,
+      "loss": 0.4401,
+      "step": 799
+    },
+    {
+      "epoch": 0.03790570954750059,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019930870429127193,
+      "loss": 0.345,
+      "step": 800
+    },
+    {
+      "epoch": 0.03795309168443497,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019930695509348534,
+      "loss": 1.2678,
+      "step": 801
+    },
+    {
+      "epoch": 0.03800047382136934,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019930520369317622,
+      "loss": 0.1882,
+      "step": 802
+    },
+    {
+      "epoch": 0.03804785595830372,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019930345009038351,
+      "loss": 0.4099,
+      "step": 803
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000199301694285146,
+      "loss": 0.2211,
+      "step": 804
+    },
+    {
+      "epoch": 0.03814262023217247,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019929993627750272,
+      "loss": 0.8162,
+      "step": 805
+    },
+    {
+      "epoch": 0.03819000236910685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001992981760674926,
+      "loss": 1.9835,
+      "step": 806
+    },
+    {
+      "epoch": 0.03823738450604122,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019929641365515474,
+      "loss": 0.9136,
+      "step": 807
+    },
+    {
+      "epoch": 0.0382847666429756,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019929464904052812,
+      "loss": 0.6974,
+      "step": 808
+    },
+    {
+      "epoch": 0.03833214877990997,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019929288222365202,
+      "loss": 0.7456,
+      "step": 809
+    },
+    {
+      "epoch": 0.03837953091684435,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001992911132045655,
+      "loss": 1.1999,
+      "step": 810
+    },
+    {
+      "epoch": 0.038426913053778725,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001992893419833079,
+      "loss": 0.7235,
+      "step": 811
+    },
+    {
+      "epoch": 0.038474295190713104,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001992875685599184,
+      "loss": 1.0206,
+      "step": 812
+    },
+    {
+      "epoch": 0.038521677327647476,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001992857929344364,
+      "loss": 0.3944,
+      "step": 813
+    },
+    {
+      "epoch": 0.038569059464581855,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001992840151069013,
+      "loss": 1.2379,
+      "step": 814
+    },
+    {
+      "epoch": 0.03861644160151623,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019928223507735248,
+      "loss": 1.1114,
+      "step": 815
+    },
+    {
+      "epoch": 0.038663823738450606,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019928045284582941,
+      "loss": 0.2301,
+      "step": 816
+    },
+    {
+      "epoch": 0.03871120587538498,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019927866841237167,
+      "loss": 1.174,
+      "step": 817
+    },
+    {
+      "epoch": 0.03875858801231936,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019927688177701883,
+      "loss": 0.6894,
+      "step": 818
+    },
+    {
+      "epoch": 0.03880597014925373,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019927509293981048,
+      "loss": 1.3543,
+      "step": 819
+    },
+    {
+      "epoch": 0.03885335228618811,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001992733019007863,
+      "loss": 0.2246,
+      "step": 820
+    },
+    {
+      "epoch": 0.03890073442312248,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019927150865998604,
+      "loss": 0.6539,
+      "step": 821
+    },
+    {
+      "epoch": 0.03894811656005686,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019926971321744942,
+      "loss": 1.4651,
+      "step": 822
+    },
+    {
+      "epoch": 0.03899549869699123,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019926791557321635,
+      "loss": 0.8091,
+      "step": 823
+    },
+    {
+      "epoch": 0.03904288083392561,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019926611572732662,
+      "loss": 1.3761,
+      "step": 824
+    },
+    {
+      "epoch": 0.03909026297085998,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001992643136798202,
+      "loss": 0.3683,
+      "step": 825
+    },
+    {
+      "epoch": 0.03913764510779436,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019926250943073698,
+      "loss": 0.5105,
+      "step": 826
+    },
+    {
+      "epoch": 0.039185027244728735,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001992607029801171,
+      "loss": 0.542,
+      "step": 827
+    },
+    {
+      "epoch": 0.039232409381663114,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001992588943280005,
+      "loss": 0.2755,
+      "step": 828
+    },
+    {
+      "epoch": 0.039279791518597486,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001992570834744274,
+      "loss": 0.2822,
+      "step": 829
+    },
+    {
+      "epoch": 0.039327173655531865,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001992552704194379,
+      "loss": 0.9901,
+      "step": 830
+    },
+    {
+      "epoch": 0.03937455579246624,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.00019925345516307217,
+      "loss": 0.0039,
+      "step": 831
+    },
+    {
+      "epoch": 0.039421937929400616,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019925163770537059,
+      "loss": 1.0106,
+      "step": 832
+    },
+    {
+      "epoch": 0.03946932006633499,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 0.00019924981804637337,
+      "loss": 0.0026,
+      "step": 833
+    },
+    {
+      "epoch": 0.03951670220326937,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001992479961861209,
+      "loss": 0.9969,
+      "step": 834
+    },
+    {
+      "epoch": 0.039564084340203746,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001992461721246536,
+      "loss": 0.2964,
+      "step": 835
+    },
+    {
+      "epoch": 0.03961146647713812,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019924434586201191,
+      "loss": 1.2602,
+      "step": 836
+    },
+    {
+      "epoch": 0.0396588486140725,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019924251739823637,
+      "loss": 0.7571,
+      "step": 837
+    },
+    {
+      "epoch": 0.03970623075100687,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00019924068673336746,
+      "loss": 0.0111,
+      "step": 838
+    },
+    {
+      "epoch": 0.03975361288794125,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019923885386744582,
+      "loss": 0.02,
+      "step": 839
+    },
+    {
+      "epoch": 0.03980099502487562,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019923701880051212,
+      "loss": 1.8461,
+      "step": 840
+    },
+    {
+      "epoch": 0.03984837716181,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019923518153260706,
+      "loss": 0.3594,
+      "step": 841
+    },
+    {
+      "epoch": 0.03989575929874437,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019923334206377135,
+      "loss": 0.2136,
+      "step": 842
+    },
+    {
+      "epoch": 0.03994314143567875,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019923150039404582,
+      "loss": 0.7874,
+      "step": 843
+    },
+    {
+      "epoch": 0.039990523572613124,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019922965652347134,
+      "loss": 1.266,
+      "step": 844
+    },
+    {
+      "epoch": 0.0400379057095475,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019922781045208875,
+      "loss": 1.0366,
+      "step": 845
+    },
+    {
+      "epoch": 0.040085287846481875,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000199225962179939,
+      "loss": 1.495,
+      "step": 846
+    },
+    {
+      "epoch": 0.040132669983416254,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019922411170706313,
+      "loss": 0.9435,
+      "step": 847
+    },
+    {
+      "epoch": 0.040180052120350626,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019922225903350212,
+      "loss": 1.0605,
+      "step": 848
+    },
+    {
+      "epoch": 0.040227434257285005,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001992204041592971,
+      "loss": 1.4564,
+      "step": 849
+    },
+    {
+      "epoch": 0.04027481639421938,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001992185470844892,
+      "loss": 0.6306,
+      "step": 850
+    },
+    {
+      "epoch": 0.040322198531153756,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019921668780911963,
+      "loss": 0.2552,
+      "step": 851
+    },
+    {
+      "epoch": 0.04036958066808813,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001992148263332296,
+      "loss": 0.1815,
+      "step": 852
+    },
+    {
+      "epoch": 0.04041696280502251,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001992129626568604,
+      "loss": 1.2073,
+      "step": 853
+    },
+    {
+      "epoch": 0.04046434494195688,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00019921109678005335,
+      "loss": 0.0031,
+      "step": 854
+    },
+    {
+      "epoch": 0.04051172707889126,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019920922870284984,
+      "loss": 1.0907,
+      "step": 855
+    },
+    {
+      "epoch": 0.04055910921582563,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001992073584252913,
+      "loss": 0.8843,
+      "step": 856
+    },
+    {
+      "epoch": 0.04060649135276001,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019920548594741928,
+      "loss": 1.1248,
+      "step": 857
+    },
+    {
+      "epoch": 0.04065387348969438,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019920361126927522,
+      "loss": 0.2766,
+      "step": 858
+    },
+    {
+      "epoch": 0.04070125562662876,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019920173439090072,
+      "loss": 0.9496,
+      "step": 859
+    },
+    {
+      "epoch": 0.040748637763563134,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019919985531233742,
+      "loss": 0.2032,
+      "step": 860
+    },
+    {
+      "epoch": 0.04079601990049751,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000199197974033627,
+      "loss": 0.7034,
+      "step": 861
+    },
+    {
+      "epoch": 0.040843402037431885,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019919609055481116,
+      "loss": 1.1872,
+      "step": 862
+    },
+    {
+      "epoch": 0.040890784174366264,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001991942048759317,
+      "loss": 1.3044,
+      "step": 863
+    },
+    {
+      "epoch": 0.04093816631130064,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019919231699703046,
+      "loss": 1.1731,
+      "step": 864
+    },
+    {
+      "epoch": 0.040985548448235015,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019919042691814924,
+      "loss": 0.2639,
+      "step": 865
+    },
+    {
+      "epoch": 0.041032930585169394,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019918853463933003,
+      "loss": 0.83,
+      "step": 866
+    },
+    {
+      "epoch": 0.041080312722103766,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019918664016061474,
+      "loss": 1.0699,
+      "step": 867
+    },
+    {
+      "epoch": 0.041127694859038146,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019918474348204544,
+      "loss": 1.2956,
+      "step": 868
+    },
+    {
+      "epoch": 0.04117507699597252,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001991828446036642,
+      "loss": 0.8589,
+      "step": 869
+    },
+    {
+      "epoch": 0.0412224591329069,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019918094352551312,
+      "loss": 0.6539,
+      "step": 870
+    },
+    {
+      "epoch": 0.04126984126984127,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019917904024763428,
+      "loss": 0.4389,
+      "step": 871
+    },
+    {
+      "epoch": 0.04131722340677565,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019917713477007003,
+      "loss": 1.028,
+      "step": 872
+    },
+    {
+      "epoch": 0.04136460554371002,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019917522709286256,
+      "loss": 0.2418,
+      "step": 873
+    },
+    {
+      "epoch": 0.0414119876806444,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019917331721605418,
+      "loss": 0.2146,
+      "step": 874
+    },
+    {
+      "epoch": 0.04145936981757877,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019917140513968725,
+      "loss": 1.2755,
+      "step": 875
+    },
+    {
+      "epoch": 0.04150675195451315,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001991694908638042,
+      "loss": 0.8287,
+      "step": 876
+    },
+    {
+      "epoch": 0.04155413409144752,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001991675743884475,
+      "loss": 0.2201,
+      "step": 877
+    },
+    {
+      "epoch": 0.0416015162283819,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001991656557136596,
+      "loss": 0.0061,
+      "step": 878
+    },
+    {
+      "epoch": 0.041648898365316274,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019916373483948308,
+      "loss": 1.242,
+      "step": 879
+    },
+    {
+      "epoch": 0.04169628050225065,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019916181176596055,
+      "loss": 1.2381,
+      "step": 880
+    },
+    {
+      "epoch": 0.041743662639185025,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019915988649313467,
+      "loss": 1.2693,
+      "step": 881
+    },
+    {
+      "epoch": 0.041791044776119404,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001991579590210481,
+      "loss": 1.3847,
+      "step": 882
+    },
+    {
+      "epoch": 0.041838426913053776,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019915602934974364,
+      "loss": 1.032,
+      "step": 883
+    },
+    {
+      "epoch": 0.041885809049988156,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019915409747926405,
+      "loss": 1.0213,
+      "step": 884
+    },
+    {
+      "epoch": 0.04193319118692253,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0001991521634096522,
+      "loss": 0.7964,
+      "step": 885
+    },
+    {
+      "epoch": 0.04198057332385691,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019915022714095098,
+      "loss": 0.8511,
+      "step": 886
+    },
+    {
+      "epoch": 0.04202795546079128,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019914828867320335,
+      "loss": 0.734,
+      "step": 887
+    },
+    {
+      "epoch": 0.04207533759772566,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019914634800645225,
+      "loss": 0.9853,
+      "step": 888
+    },
+    {
+      "epoch": 0.04212271973466003,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019914440514074078,
+      "loss": 0.9674,
+      "step": 889
+    },
+    {
+      "epoch": 0.04217010187159441,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000199142460076112,
+      "loss": 1.0232,
+      "step": 890
+    },
+    {
+      "epoch": 0.04221748400852878,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019914051281260905,
+      "loss": 0.8152,
+      "step": 891
+    },
+    {
+      "epoch": 0.04226486614546316,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019913856335027514,
+      "loss": 0.0387,
+      "step": 892
+    },
+    {
+      "epoch": 0.04231224828239753,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001991366116891535,
+      "loss": 0.8438,
+      "step": 893
+    },
+    {
+      "epoch": 0.04235963041933191,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019913465782928736,
+      "loss": 0.1963,
+      "step": 894
+    },
+    {
+      "epoch": 0.04240701255626629,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019913270177072015,
+      "loss": 0.8991,
+      "step": 895
+    },
+    {
+      "epoch": 0.04245439469320066,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001991307435134952,
+      "loss": 0.6034,
+      "step": 896
+    },
+    {
+      "epoch": 0.04250177683013504,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019912878305765593,
+      "loss": 1.2035,
+      "step": 897
+    },
+    {
+      "epoch": 0.042549158967069414,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019912682040324587,
+      "loss": 1.2343,
+      "step": 898
+    },
+    {
+      "epoch": 0.04259654110400379,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001991248555503085,
+      "loss": 0.1443,
+      "step": 899
+    },
+    {
+      "epoch": 0.042643923240938165,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019912288849888743,
+      "loss": 1.281,
+      "step": 900
+    },
+    {
+      "epoch": 0.042691305377872545,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019912091924902624,
+      "loss": 0.0633,
+      "step": 901
+    },
+    {
+      "epoch": 0.04273868751480692,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019911894780076867,
+      "loss": 1.2478,
+      "step": 902
+    },
+    {
+      "epoch": 0.042786069651741296,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001991169741541584,
+      "loss": 0.7873,
+      "step": 903
+    },
+    {
+      "epoch": 0.04283345178867567,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019911499830923922,
+      "loss": 0.9015,
+      "step": 904
+    },
+    {
+      "epoch": 0.04288083392561005,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019911302026605495,
+      "loss": 1.2971,
+      "step": 905
+    },
+    {
+      "epoch": 0.04292821606254442,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019911104002464947,
+      "loss": 0.965,
+      "step": 906
+    },
+    {
+      "epoch": 0.0429755981994788,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019910905758506667,
+      "loss": 1.2898,
+      "step": 907
+    },
+    {
+      "epoch": 0.04302298033641317,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019910707294735057,
+      "loss": 0.9892,
+      "step": 908
+    },
+    {
+      "epoch": 0.04307036247334755,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019910508611154515,
+      "loss": 0.2499,
+      "step": 909
+    },
+    {
+      "epoch": 0.04311774461028192,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019910309707769447,
+      "loss": 0.1086,
+      "step": 910
+    },
+    {
+      "epoch": 0.0431651267472163,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019910110584584265,
+      "loss": 0.824,
+      "step": 911
+    },
+    {
+      "epoch": 0.04321250888415067,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019909911241603386,
+      "loss": 1.3195,
+      "step": 912
+    },
+    {
+      "epoch": 0.04325989102108505,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001990971167883123,
+      "loss": 0.1024,
+      "step": 913
+    },
+    {
+      "epoch": 0.043307273158019424,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019909511896272229,
+      "loss": 0.6601,
+      "step": 914
+    },
+    {
+      "epoch": 0.0433546552949538,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019909311893930807,
+      "loss": 0.3921,
+      "step": 915
+    },
+    {
+      "epoch": 0.043402037431888175,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019909111671811402,
+      "loss": 0.7832,
+      "step": 916
+    },
+    {
+      "epoch": 0.043449419568822555,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019908911229918452,
+      "loss": 1.384,
+      "step": 917
+    },
+    {
+      "epoch": 0.04349680170575693,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001990871056825641,
+      "loss": 1.0337,
+      "step": 918
+    },
+    {
+      "epoch": 0.043544183842691306,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001990850968682972,
+      "loss": 0.2049,
+      "step": 919
+    },
+    {
+      "epoch": 0.04359156597962568,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019908308585642838,
+      "loss": 0.6315,
+      "step": 920
+    },
+    {
+      "epoch": 0.04363894811656006,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019908107264700225,
+      "loss": 1.143,
+      "step": 921
+    },
+    {
+      "epoch": 0.04368633025349443,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001990790572400635,
+      "loss": 0.3487,
+      "step": 922
+    },
+    {
+      "epoch": 0.04373371239042881,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019907703963565677,
+      "loss": 1.0009,
+      "step": 923
+    },
+    {
+      "epoch": 0.04378109452736319,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019907501983382683,
+      "loss": 0.0066,
+      "step": 924
+    },
+    {
+      "epoch": 0.04382847666429756,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019907299783461852,
+      "loss": 0.513,
+      "step": 925
+    },
+    {
+      "epoch": 0.04387585880123194,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001990709736380766,
+      "loss": 0.1052,
+      "step": 926
+    },
+    {
+      "epoch": 0.04392324093816631,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000199068947244246,
+      "loss": 1.0993,
+      "step": 927
+    },
+    {
+      "epoch": 0.04397062307510069,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019906691865317173,
+      "loss": 0.859,
+      "step": 928
+    },
+    {
+      "epoch": 0.04401800521203506,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019906488786489867,
+      "loss": 1.018,
+      "step": 929
+    },
+    {
+      "epoch": 0.04406538734896944,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019906285487947197,
+      "loss": 1.2107,
+      "step": 930
+    },
+    {
+      "epoch": 0.04411276948590381,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001990608196969366,
+      "loss": 1.1488,
+      "step": 931
+    },
+    {
+      "epoch": 0.04416015162283819,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019905878231733781,
+      "loss": 1.1261,
+      "step": 932
+    },
+    {
+      "epoch": 0.044207533759772565,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019905674274072076,
+      "loss": 0.1464,
+      "step": 933
+    },
+    {
+      "epoch": 0.044254915896706944,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001990547009671306,
+      "loss": 0.7288,
+      "step": 934
+    },
+    {
+      "epoch": 0.044302298033641316,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019905265699661273,
+      "loss": 0.3238,
+      "step": 935
+    },
+    {
+      "epoch": 0.044349680170575695,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019905061082921242,
+      "loss": 0.8499,
+      "step": 936
+    },
+    {
+      "epoch": 0.04439706230751007,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019904856246497508,
+      "loss": 1.4673,
+      "step": 937
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001990465119039461,
+      "loss": 1.1882,
+      "step": 938
+    },
+    {
+      "epoch": 0.04449182658137882,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000199044459146171,
+      "loss": 1.2626,
+      "step": 939
+    },
+    {
+      "epoch": 0.0445392087183132,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001990424041916953,
+      "loss": 0.0356,
+      "step": 940
+    },
+    {
+      "epoch": 0.04458659085524757,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019904034704056454,
+      "loss": 0.2159,
+      "step": 941
+    },
+    {
+      "epoch": 0.04463397299218195,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001990382876928244,
+      "loss": 0.8416,
+      "step": 942
+    },
+    {
+      "epoch": 0.04468135512911632,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019903622614852055,
+      "loss": 0.7343,
+      "step": 943
+    },
+    {
+      "epoch": 0.0447287372660507,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019903416240769865,
+      "loss": 0.2001,
+      "step": 944
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019903209647040458,
+      "loss": 0.928,
+      "step": 945
+    },
+    {
+      "epoch": 0.04482350153991945,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019903002833668402,
+      "loss": 0.2404,
+      "step": 946
+    },
+    {
+      "epoch": 0.04487088367685382,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00019902795800658295,
+      "loss": 0.0152,
+      "step": 947
+    },
+    {
+      "epoch": 0.0449182658137882,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019902588548014724,
+      "loss": 0.8822,
+      "step": 948
+    },
+    {
+      "epoch": 0.044965647950722575,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001990238107574229,
+      "loss": 0.8807,
+      "step": 949
+    },
+    {
+      "epoch": 0.045013030087656954,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019902173383845587,
+      "loss": 0.8176,
+      "step": 950
+    },
+    {
+      "epoch": 0.045060412224591326,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019901965472329228,
+      "loss": 0.0183,
+      "step": 951
+    },
+    {
+      "epoch": 0.045107794361525705,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001990175734119782,
+      "loss": 0.7378,
+      "step": 952
+    },
+    {
+      "epoch": 0.045155176498460084,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001990154899045598,
+      "loss": 1.3953,
+      "step": 953
+    },
+    {
+      "epoch": 0.045202558635394456,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019901340420108333,
+      "loss": 1.292,
+      "step": 954
+    },
+    {
+      "epoch": 0.045249940772328835,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019901131630159502,
+      "loss": 1.3624,
+      "step": 955
+    },
+    {
+      "epoch": 0.04529732290926321,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019900922620614119,
+      "loss": 1.2252,
+      "step": 956
+    },
+    {
+      "epoch": 0.045344705046197586,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019900713391476815,
+      "loss": 0.4908,
+      "step": 957
+    },
+    {
+      "epoch": 0.04539208718313196,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019900503942752235,
+      "loss": 1.1443,
+      "step": 958
+    },
+    {
+      "epoch": 0.04543946932006634,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001990029427444502,
+      "loss": 0.9862,
+      "step": 959
+    },
+    {
+      "epoch": 0.04548685145700071,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00019900084386559826,
+      "loss": 0.0205,
+      "step": 960
+    },
+    {
+      "epoch": 0.04553423359393509,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019899874279101306,
+      "loss": 0.1988,
+      "step": 961
+    },
+    {
+      "epoch": 0.04558161573086946,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019899663952074122,
+      "loss": 1.3923,
+      "step": 962
+    },
+    {
+      "epoch": 0.04562899786780384,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019899453405482933,
+      "loss": 1.2035,
+      "step": 963
+    },
+    {
+      "epoch": 0.04567638000473821,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019899242639332413,
+      "loss": 0.1353,
+      "step": 964
+    },
+    {
+      "epoch": 0.04572376214167259,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001989903165362723,
+      "loss": 0.1484,
+      "step": 965
+    },
+    {
+      "epoch": 0.045771144278606964,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019898820448372074,
+      "loss": 1.3354,
+      "step": 966
+    },
+    {
+      "epoch": 0.04581852641554134,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019898609023571626,
+      "loss": 0.4631,
+      "step": 967
+    },
+    {
+      "epoch": 0.045865908552475715,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001989839737923057,
+      "loss": 0.4975,
+      "step": 968
+    },
+    {
+      "epoch": 0.045913290689410094,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019898185515353608,
+      "loss": 1.0471,
+      "step": 969
+    },
+    {
+      "epoch": 0.045960672826344466,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001989797343194543,
+      "loss": 0.8438,
+      "step": 970
+    },
+    {
+      "epoch": 0.046008054963278845,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019897761129010743,
+      "loss": 1.1028,
+      "step": 971
+    },
+    {
+      "epoch": 0.04605543710021322,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019897548606554258,
+      "loss": 0.7336,
+      "step": 972
+    },
+    {
+      "epoch": 0.046102819237147596,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001989733586458069,
+      "loss": 0.954,
+      "step": 973
+    },
+    {
+      "epoch": 0.04615020137408197,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019897122903094752,
+      "loss": 1.1067,
+      "step": 974
+    },
+    {
+      "epoch": 0.04619758351101635,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001989690972210117,
+      "loss": 1.6079,
+      "step": 975
+    },
+    {
+      "epoch": 0.04624496564795072,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019896696321604674,
+      "loss": 0.3096,
+      "step": 976
+    },
+    {
+      "epoch": 0.0462923477848851,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019896482701609993,
+      "loss": 0.9771,
+      "step": 977
+    },
+    {
+      "epoch": 0.04633972992181947,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019896268862121868,
+      "loss": 0.9484,
+      "step": 978
+    },
+    {
+      "epoch": 0.04638711205875385,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019896054803145039,
+      "loss": 0.2152,
+      "step": 979
+    },
+    {
+      "epoch": 0.04643449419568822,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019895840524684257,
+      "loss": 0.9165,
+      "step": 980
+    },
+    {
+      "epoch": 0.0464818763326226,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001989562602674427,
+      "loss": 1.0838,
+      "step": 981
+    },
+    {
+      "epoch": 0.046529258469556974,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019895411309329845,
+      "loss": 1.2539,
+      "step": 982
+    },
+    {
+      "epoch": 0.04657664060649135,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001989519637244573,
+      "loss": 1.1238,
+      "step": 983
+    },
+    {
+      "epoch": 0.04662402274342573,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019894981216096703,
+      "loss": 1.024,
+      "step": 984
+    },
+    {
+      "epoch": 0.046671404880360104,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019894765840287532,
+      "loss": 1.2483,
+      "step": 985
+    },
+    {
+      "epoch": 0.04671878701729448,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019894550245022993,
+      "loss": 1.0601,
+      "step": 986
+    },
+    {
+      "epoch": 0.046766169154228855,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019894334430307868,
+      "loss": 0.9888,
+      "step": 987
+    },
+    {
+      "epoch": 0.046813551291163234,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001989411839614695,
+      "loss": 0.8472,
+      "step": 988
+    },
+    {
+      "epoch": 0.046860933428097606,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019893902142545015,
+      "loss": 0.0727,
+      "step": 989
+    },
+    {
+      "epoch": 0.046908315565031986,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019893685669506876,
+      "loss": 0.3552,
+      "step": 990
+    },
+    {
+      "epoch": 0.04695569770196636,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019893468977037325,
+      "loss": 0.9933,
+      "step": 991
+    },
+    {
+      "epoch": 0.04700307983890074,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001989325206514117,
+      "loss": 0.0244,
+      "step": 992
+    },
+    {
+      "epoch": 0.04705046197583511,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019893034933823222,
+      "loss": 0.9056,
+      "step": 993
+    },
+    {
+      "epoch": 0.04709784411276949,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.000198928175830883,
+      "loss": 0.3743,
+      "step": 994
+    },
+    {
+      "epoch": 0.04714522624970386,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019892600012941217,
+      "loss": 0.8073,
+      "step": 995
+    },
+    {
+      "epoch": 0.04719260838663824,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019892382223386806,
+      "loss": 0.947,
+      "step": 996
+    },
+    {
+      "epoch": 0.04723999052357261,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001989216421442989,
+      "loss": 1.0194,
+      "step": 997
+    },
+    {
+      "epoch": 0.04728737266050699,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001989194598607531,
+      "loss": 1.2811,
+      "step": 998
+    },
+    {
+      "epoch": 0.04733475479744136,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001989172753832791,
+      "loss": 0.7862,
+      "step": 999
+    },
+    {
+      "epoch": 0.04738213693437574,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019891508871192523,
+      "loss": 1.246,
+      "step": 1000
+    },
+    {
+      "epoch": 0.047429519071310114,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001989128998467401,
+      "loss": 0.2165,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04747690120824449,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019891070878777213,
+      "loss": 1.1898,
+      "step": 1002
+    },
+    {
+      "epoch": 0.047524283345178865,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019890851553507006,
+      "loss": 0.3253,
+      "step": 1003
+    },
+    {
+      "epoch": 0.047571665482113244,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019890632008868244,
+      "loss": 0.8785,
+      "step": 1004
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.000198904122448658,
+      "loss": 0.6022,
+      "step": 1005
+    },
+    {
+      "epoch": 0.047666429755981995,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019890192261504548,
+      "loss": 1.3854,
+      "step": 1006
+    },
+    {
+      "epoch": 0.04771381189291637,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019889972058789366,
+      "loss": 0.0278,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04776119402985075,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019889751636725138,
+      "loss": 1.2147,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04780857616678512,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019889530995316753,
+      "loss": 1.0106,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0478559583037195,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019889310134569104,
+      "loss": 0.8297,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04790334044065387,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019889089054487088,
+      "loss": 0.8559,
+      "step": 1011
+    },
+    {
+      "epoch": 0.04795072257758825,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019888867755075613,
+      "loss": 0.322,
+      "step": 1012
+    },
+    {
+      "epoch": 0.04799810471452263,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.00019888646236339584,
+      "loss": 0.0064,
+      "step": 1013
+    },
+    {
+      "epoch": 0.048045486851457,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019888424498283914,
+      "loss": 0.1653,
+      "step": 1014
+    },
+    {
+      "epoch": 0.04809286898839138,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019888202540913522,
+      "loss": 0.8939,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04814025112532575,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019887980364233327,
+      "loss": 0.6065,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04818763326226013,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00019887757968248263,
+      "loss": 0.007,
+      "step": 1017
+    },
+    {
+      "epoch": 0.0482350153991945,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019887535352963257,
+      "loss": 1.1738,
+      "step": 1018
+    },
+    {
+      "epoch": 0.04828239753612888,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001988731251838325,
+      "loss": 0.4478,
+      "step": 1019
+    },
+    {
+      "epoch": 0.048329779673063254,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019887089464513182,
+      "loss": 1.1147,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04837716180999763,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019886866191358,
+      "loss": 0.625,
+      "step": 1021
+    },
+    {
+      "epoch": 0.048424543946932005,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001988664269892266,
+      "loss": 0.8374,
+      "step": 1022
+    },
+    {
+      "epoch": 0.048471926083866385,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019886418987212113,
+      "loss": 0.7784,
+      "step": 1023
+    },
+    {
+      "epoch": 0.04851930822080076,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019886195056231326,
+      "loss": 0.1788,
+      "step": 1024
+    },
+    {
+      "epoch": 0.048566690357735136,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001988597090598526,
+      "loss": 1.3,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04861407249466951,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001988574653647889,
+      "loss": 0.0471,
+      "step": 1026
+    },
+    {
+      "epoch": 0.04866145463160389,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019885521947717193,
+      "loss": 0.835,
+      "step": 1027
+    },
+    {
+      "epoch": 0.04870883676853826,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001988529713970515,
+      "loss": 1.1535,
+      "step": 1028
+    },
+    {
+      "epoch": 0.04875621890547264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019885072112447741,
+      "loss": 0.7651,
+      "step": 1029
+    },
+    {
+      "epoch": 0.04880360104240701,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019884846865949967,
+      "loss": 0.3877,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04885098317934139,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019884621400216815,
+      "loss": 0.2153,
+      "step": 1031
+    },
+    {
+      "epoch": 0.04889836531627576,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001988439571525329,
+      "loss": 1.0758,
+      "step": 1032
+    },
+    {
+      "epoch": 0.04894574745321014,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019884169811064395,
+      "loss": 0.015,
+      "step": 1033
+    },
+    {
+      "epoch": 0.04899312959014451,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019883943687655143,
+      "loss": 0.0581,
+      "step": 1034
+    },
+    {
+      "epoch": 0.04904051172707889,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001988371734503055,
+      "loss": 1.1398,
+      "step": 1035
+    },
+    {
+      "epoch": 0.049087893864013264,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019883490783195628,
+      "loss": 1.2845,
+      "step": 1036
+    },
+    {
+      "epoch": 0.04913527600094764,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019883264002155414,
+      "loss": 1.0436,
+      "step": 1037
+    },
+    {
+      "epoch": 0.049182658137882015,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001988303700191493,
+      "loss": 0.2564,
+      "step": 1038
+    },
+    {
+      "epoch": 0.049230040274816395,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001988280978247921,
+      "loss": 1.0179,
+      "step": 1039
+    },
+    {
+      "epoch": 0.04927742241175077,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019882582343853298,
+      "loss": 0.7744,
+      "step": 1040
+    },
+    {
+      "epoch": 0.049324804548685146,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019882354686042236,
+      "loss": 1.5745,
+      "step": 1041
+    },
+    {
+      "epoch": 0.049372186685619525,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019882126809051071,
+      "loss": 0.6615,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0494195688225539,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001988189871288486,
+      "loss": 0.9245,
+      "step": 1043
+    },
+    {
+      "epoch": 0.049466950959488276,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019881670397548664,
+      "loss": 0.8864,
+      "step": 1044
+    },
+    {
+      "epoch": 0.04951433309642265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019881441863047543,
+      "loss": 0.6943,
+      "step": 1045
+    },
+    {
+      "epoch": 0.04956171523335703,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019881213109386567,
+      "loss": 1.1239,
+      "step": 1046
+    },
+    {
+      "epoch": 0.0496090973702914,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019880984136570805,
+      "loss": 1.1835,
+      "step": 1047
+    },
+    {
+      "epoch": 0.04965647950722578,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019880754944605344,
+      "loss": 0.058,
+      "step": 1048
+    },
+    {
+      "epoch": 0.04970386164416015,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019880525533495262,
+      "loss": 0.9848,
+      "step": 1049
+    },
+    {
+      "epoch": 0.04975124378109453,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001988029590324565,
+      "loss": 0.0888,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0497986259180289,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019880066053861594,
+      "loss": 1.3071,
+      "step": 1051
+    },
+    {
+      "epoch": 0.04984600805496328,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019879835985348198,
+      "loss": 0.1956,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04989339019189765,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019879605697710565,
+      "loss": 0.8406,
+      "step": 1053
+    },
+    {
+      "epoch": 0.04994077232883203,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019879375190953803,
+      "loss": 0.2062,
+      "step": 1054
+    },
+    {
+      "epoch": 0.049988154465766405,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019879144465083018,
+      "loss": 1.3425,
+      "step": 1055
+    },
+    {
+      "epoch": 0.050035536602700784,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019878913520103334,
+      "loss": 0.5386,
+      "step": 1056
+    },
+    {
+      "epoch": 0.050082918739635156,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019878682356019872,
+      "loss": 0.7189,
+      "step": 1057
+    },
+    {
+      "epoch": 0.050130300876569535,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019878450972837753,
+      "loss": 1.426,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05017768301350391,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019878219370562117,
+      "loss": 0.7201,
+      "step": 1059
+    },
+    {
+      "epoch": 0.050225065150438286,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019877987549198097,
+      "loss": 0.0785,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05027244728737266,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019877755508750832,
+      "loss": 0.1805,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05031982942430704,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019877523249225477,
+      "loss": 1.2644,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05036721156124141,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001987729077062717,
+      "loss": 0.9881,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05041459369817579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001987705807296108,
+      "loss": 0.7509,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05046197583511016,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0001987682515623236,
+      "loss": 0.0115,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05050935797204454,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019876592020446178,
+      "loss": 0.2251,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05055674010897891,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019876358665607706,
+      "loss": 1.3754,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05060412224591329,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001987612509172212,
+      "loss": 0.1373,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05065150438284766,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019875891298794596,
+      "loss": 0.0087,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05069888651978204,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019875657286830324,
+      "loss": 1.1929,
+      "step": 1070
+    },
+    {
+      "epoch": 0.050746268656716415,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019875423055834492,
+      "loss": 0.0568,
+      "step": 1071
+    },
+    {
+      "epoch": 0.050793650793650794,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019875188605812297,
+      "loss": 0.7779,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05084103293058517,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019874953936768936,
+      "loss": 1.5075,
+      "step": 1073
+    },
+    {
+      "epoch": 0.050888415067519545,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019874719048709616,
+      "loss": 1.1444,
+      "step": 1074
+    },
+    {
+      "epoch": 0.050935797204453924,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019874483941639546,
+      "loss": 1.0448,
+      "step": 1075
+    },
+    {
+      "epoch": 0.050983179341388296,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019874248615563936,
+      "loss": 1.245,
+      "step": 1076
+    },
+    {
+      "epoch": 0.051030561478322675,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019874013070488014,
+      "loss": 0.8945,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05107794361525705,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019873777306416996,
+      "loss": 0.7757,
+      "step": 1078
+    },
+    {
+      "epoch": 0.051125325752191426,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019873541323356118,
+      "loss": 1.0556,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0511727078891258,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019873305121310609,
+      "loss": 0.1753,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05122009002606018,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019873068700285704,
+      "loss": 1.4021,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05126747216299455,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019872832060286656,
+      "loss": 0.8074,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05131485429992893,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019872595201318708,
+      "loss": 0.9757,
+      "step": 1083
+    },
+    {
+      "epoch": 0.0513622364368633,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019872358123387116,
+      "loss": 0.1936,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05140961857379768,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019872120826497136,
+      "loss": 1.2161,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05145700071073205,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019871883310654031,
+      "loss": 1.0108,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05150438284766643,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001987164557586307,
+      "loss": 0.9926,
+      "step": 1087
+    },
+    {
+      "epoch": 0.051551764984600804,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019871407622129523,
+      "loss": 1.2252,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05159914712153518,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001987116944945867,
+      "loss": 0.2188,
+      "step": 1089
+    },
+    {
+      "epoch": 0.051646529258469555,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 0.00019870931057855792,
+      "loss": 0.0031,
+      "step": 1090
+    },
+    {
+      "epoch": 0.051693911395403934,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001987069244732618,
+      "loss": 1.4619,
+      "step": 1091
+    },
+    {
+      "epoch": 0.051741293532338306,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019870453617875123,
+      "loss": 0.9762,
+      "step": 1092
+    },
+    {
+      "epoch": 0.051788675669272685,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019870214569507914,
+      "loss": 0.3705,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05183605780620706,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019869975302229864,
+      "loss": 1.2158,
+      "step": 1094
+    },
+    {
+      "epoch": 0.051883439943141436,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001986973581604627,
+      "loss": 0.9881,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05193082208007581,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019869496110962452,
+      "loss": 1.101,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05197820421701019,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019869256186983724,
+      "loss": 0.5085,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05202558635394456,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019869016044115405,
+      "loss": 0.767,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05207296849087894,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001986877568236282,
+      "loss": 0.0881,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05212035062781331,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019868535101731305,
+      "loss": 0.2183,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05216773276474769,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019868294302226192,
+      "loss": 1.084,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05221511490168207,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001986805328385282,
+      "loss": 0.9669,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05226249703861644,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0001986781204661654,
+      "loss": 0.0065,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05230987917555082,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019867570590522698,
+      "loss": 1.0672,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05235726131248519,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001986732891557665,
+      "loss": 0.205,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05240464344941957,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001986708702178376,
+      "loss": 0.5019,
+      "step": 1106
+    },
+    {
+      "epoch": 0.052452025586353944,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019866844909149388,
+      "loss": 0.9741,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05249940772328832,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019866602577678905,
+      "loss": 1.3385,
+      "step": 1108
+    },
+    {
+      "epoch": 0.052546789860222695,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019866360027377686,
+      "loss": 0.6885,
+      "step": 1109
+    },
+    {
+      "epoch": 0.052594171997157074,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019866117258251112,
+      "loss": 0.1899,
+      "step": 1110
+    },
+    {
+      "epoch": 0.052641554134091446,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019865874270304565,
+      "loss": 0.7239,
+      "step": 1111
+    },
+    {
+      "epoch": 0.052688936271025825,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.00019865631063543438,
+      "loss": 0.0036,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0527363184079602,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001986538763797312,
+      "loss": 0.9317,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05278370054489458,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019865143993599014,
+      "loss": 0.0368,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05283108268182895,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001986490013042652,
+      "loss": 1.0516,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05287846481876333,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001986465604846105,
+      "loss": 1.0051,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0529258469556977,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019864411747708016,
+      "loss": 0.5711,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05297322909263208,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019864167228172836,
+      "loss": 1.2666,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05302061122956645,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019863922489860936,
+      "loss": 1.4304,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05306799336650083,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019863677532777738,
+      "loss": 0.9424,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0531153755034352,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001986343235692868,
+      "loss": 1.1107,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05316275764036958,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000198631869623192,
+      "loss": 1.2194,
+      "step": 1122
+    },
+    {
+      "epoch": 0.053210139777303954,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019862941348954737,
+      "loss": 0.9687,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05325752191423833,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019862695516840743,
+      "loss": 0.7604,
+      "step": 1124
+    },
+    {
+      "epoch": 0.053304904051172705,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019862449465982666,
+      "loss": 1.1728,
+      "step": 1125
+    },
+    {
+      "epoch": 0.053352286188107084,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019862203196385964,
+      "loss": 1.0629,
+      "step": 1126
+    },
+    {
+      "epoch": 0.053399668325041456,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000198619567080561,
+      "loss": 0.3663,
+      "step": 1127
+    },
+    {
+      "epoch": 0.053447050461975835,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001986171000099854,
+      "loss": 1.1313,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05349443259891021,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001986146307521876,
+      "loss": 0.9301,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05354181473584459,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019861215930722234,
+      "loss": 0.9257,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05358919687277896,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019860968567514436,
+      "loss": 1.0074,
+      "step": 1131
+    },
+    {
+      "epoch": 0.05363657900971334,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001986072098560086,
+      "loss": 0.9213,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05368396114664772,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019860473184987,
+      "loss": 0.7373,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05373134328358209,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019860225165678345,
+      "loss": 1.1619,
+      "step": 1134
+    },
+    {
+      "epoch": 0.05377872542051647,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00019859976927680397,
+      "loss": 1.2205,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05382610755745084,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019859728470998666,
+      "loss": 1.1632,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05387348969438522,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00019859479795638658,
+      "loss": 0.0056,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05392087183131959,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001985923090160589,
+      "loss": 0.0159,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05396825396825397,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019858981788905883,
+      "loss": 1.1557,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05401563610518834,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001985873245754416,
+      "loss": 0.7595,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05406301824212272,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019858482907526254,
+      "loss": 1.2538,
+      "step": 1141
+    },
+    {
+      "epoch": 0.054110400379057094,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019858233138857697,
+      "loss": 0.5855,
+      "step": 1142
+    },
+    {
+      "epoch": 0.05415778251599147,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001985798315154403,
+      "loss": 0.5724,
+      "step": 1143
+    },
+    {
+      "epoch": 0.054205164652925845,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00019857732945590794,
+      "loss": 0.2013,
+      "step": 1144
+    },
+    {
+      "epoch": 0.054252546789860225,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019857482521003545,
+      "loss": 1.0684,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0542999289267946,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001985723187778783,
+      "loss": 0.821,
+      "step": 1146
+    },
+    {
+      "epoch": 0.054347311063728976,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019856981015949215,
+      "loss": 0.8093,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05439469320066335,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019856729935493258,
+      "loss": 0.831,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05444207533759773,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001985647863642553,
+      "loss": 0.9024,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0544894574745321,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019856227118751605,
+      "loss": 0.8989,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05453683961146648,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001985597538247706,
+      "loss": 0.4668,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05458422174840085,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019855723427607483,
+      "loss": 0.0173,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05463160388533523,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001985547125414845,
+      "loss": 1.0429,
+      "step": 1153
+    },
+    {
+      "epoch": 0.0546789860222696,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019855218862105568,
+      "loss": 1.1953,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05472636815920398,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001985496625148443,
+      "loss": 0.9374,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05477375029613835,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019854713422290637,
+      "loss": 1.2047,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05482113243307273,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019854460374529794,
+      "loss": 1.514,
+      "step": 1157
+    },
+    {
+      "epoch": 0.054868514570007104,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019854207108207521,
+      "loss": 0.2,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05491589670694148,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019853953623329428,
+      "loss": 0.8601,
+      "step": 1159
+    },
+    {
+      "epoch": 0.054963278843875855,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019853699919901137,
+      "loss": 0.6545,
+      "step": 1160
+    },
+    {
+      "epoch": 0.055010660980810235,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001985344599792828,
+      "loss": 1.0417,
+      "step": 1161
+    },
+    {
+      "epoch": 0.055058043117744614,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019853191857416487,
+      "loss": 1.3017,
+      "step": 1162
+    },
+    {
+      "epoch": 0.055105425254678986,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019852937498371394,
+      "loss": 0.6266,
+      "step": 1163
+    },
+    {
+      "epoch": 0.055152807391613365,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001985268292079864,
+      "loss": 0.9562,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05520018952854774,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019852428124703876,
+      "loss": 0.0217,
+      "step": 1165
+    },
+    {
+      "epoch": 0.055247571665482116,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001985217311009275,
+      "loss": 0.8682,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05529495380241649,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019851917876970916,
+      "loss": 1.2461,
+      "step": 1167
+    },
+    {
+      "epoch": 0.05534233593935087,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019851662425344037,
+      "loss": 0.0585,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05538971807628524,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001985140675521778,
+      "loss": 0.1793,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05543710021321962,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019851150866597816,
+      "loss": 0.1714,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05548448235015399,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001985089475948982,
+      "loss": 0.1487,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05553186448708837,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019850638433899467,
+      "loss": 0.7608,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05557924662402274,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019850381889832447,
+      "loss": 1.2617,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05562662876095712,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001985012512729445,
+      "loss": 1.1914,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05567401089789149,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019849868146291168,
+      "loss": 1.0033,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05572139303482587,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019849610946828306,
+      "loss": 0.8603,
+      "step": 1176
+    },
+    {
+      "epoch": 0.055768775171760245,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001984935352891156,
+      "loss": 0.8712,
+      "step": 1177
+    },
+    {
+      "epoch": 0.055816157308694624,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019849095892546646,
+      "loss": 0.648,
+      "step": 1178
+    },
+    {
+      "epoch": 0.055863539445628996,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019848838037739275,
+      "loss": 1.5737,
+      "step": 1179
+    },
+    {
+      "epoch": 0.055910921582563375,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001984857996449517,
+      "loss": 0.2043,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05595830371949775,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019848321672820047,
+      "loss": 0.8374,
+      "step": 1181
+    },
+    {
+      "epoch": 0.056005685856432126,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001984806316271964,
+      "loss": 0.6035,
+      "step": 1182
+    },
+    {
+      "epoch": 0.0560530679933665,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019847804434199685,
+      "loss": 0.3609,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05610045013030088,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019847545487265914,
+      "loss": 0.8299,
+      "step": 1184
+    },
+    {
+      "epoch": 0.05614783226723525,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019847286321924074,
+      "loss": 0.2107,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05619521440416963,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019847026938179914,
+      "loss": 1.2353,
+      "step": 1186
+    },
+    {
+      "epoch": 0.056242596541104,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00019846767336039183,
+      "loss": 0.0066,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05628997867803838,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001984650751550764,
+      "loss": 0.1846,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05633736081497275,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001984624747659105,
+      "loss": 0.9159,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05638474295190713,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001984598721929518,
+      "loss": 0.0875,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05643212508884151,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019845726743625798,
+      "loss": 1.0923,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05647950722577588,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019845466049588686,
+      "loss": 1.2815,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05652688936271026,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019845205137189624,
+      "loss": 1.1373,
+      "step": 1193
+    },
+    {
+      "epoch": 0.056574271499644634,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019844944006434397,
+      "loss": 1.4879,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05662165363657901,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000198446826573288,
+      "loss": 0.7608,
+      "step": 1195
+    },
+    {
+      "epoch": 0.056669035773513385,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019844421089878627,
+      "loss": 0.2131,
+      "step": 1196
+    },
+    {
+      "epoch": 0.056716417910447764,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001984415930408968,
+      "loss": 1.0517,
+      "step": 1197
+    },
+    {
+      "epoch": 0.056763800047382136,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019843897299967765,
+      "loss": 0.9982,
+      "step": 1198
+    },
+    {
+      "epoch": 0.056811182184316515,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019843635077518691,
+      "loss": 1.1964,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05685856432125089,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019843372636748282,
+      "loss": 1.0349,
+      "step": 1200
+    },
+    {
+      "epoch": 0.056905946458185266,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019843109977662348,
+      "loss": 0.8252,
+      "step": 1201
+    },
+    {
+      "epoch": 0.05695332859511964,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019842847100266718,
+      "loss": 0.7719,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05700071073205402,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019842584004567225,
+      "loss": 1.2267,
+      "step": 1203
+    },
+    {
+      "epoch": 0.05704809286898839,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019842320690569704,
+      "loss": 1.3288,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05709547500592277,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019842057158279993,
+      "loss": 1.3921,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001984179340770394,
+      "loss": 0.2586,
+      "step": 1206
+    },
+    {
+      "epoch": 0.05719023927979152,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019841529438847387,
+      "loss": 0.8311,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05723762141672589,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019841265251716196,
+      "loss": 0.0966,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05728500355366027,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019841000846316224,
+      "loss": 1.0222,
+      "step": 1209
+    },
+    {
+      "epoch": 0.057332385690594644,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019840736222653335,
+      "loss": 0.2326,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05737976782752902,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019840471380733402,
+      "loss": 1.1139,
+      "step": 1211
+    },
+    {
+      "epoch": 0.057427149964463395,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019840206320562295,
+      "loss": 0.9745,
+      "step": 1212
+    },
+    {
+      "epoch": 0.057474532101397774,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019839941042145887,
+      "loss": 0.6654,
+      "step": 1213
+    },
+    {
+      "epoch": 0.057521914238332146,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019839675545490072,
+      "loss": 0.2254,
+      "step": 1214
+    },
+    {
+      "epoch": 0.057569296375266525,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019839409830600733,
+      "loss": 0.9748,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0576166785122009,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019839143897483767,
+      "loss": 0.2524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.057664060649135276,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019838877746145066,
+      "loss": 0.1985,
+      "step": 1217
+    },
+    {
+      "epoch": 0.05771144278606965,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019838611376590538,
+      "loss": 0.1742,
+      "step": 1218
+    },
+    {
+      "epoch": 0.05775882492300403,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001983834478882609,
+      "loss": 1.2711,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0578062070599384,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019838077982857634,
+      "loss": 0.7279,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05785358919687278,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019837810958691087,
+      "loss": 0.6179,
+      "step": 1221
+    },
+    {
+      "epoch": 0.05790097133380716,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001983754371633237,
+      "loss": 0.9773,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05794835347074153,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019837276255787415,
+      "loss": 0.3931,
+      "step": 1223
+    },
+    {
+      "epoch": 0.05799573560767591,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001983700857706215,
+      "loss": 1.3234,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05804311774461028,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019836740680162512,
+      "loss": 0.8854,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05809049988154466,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019836472565094442,
+      "loss": 0.9522,
+      "step": 1226
+    },
+    {
+      "epoch": 0.05813788201847903,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019836204231863888,
+      "loss": 1.0914,
+      "step": 1227
+    },
+    {
+      "epoch": 0.05818526415541341,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019835935680476803,
+      "loss": 1.223,
+      "step": 1228
+    },
+    {
+      "epoch": 0.058232646292347784,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019835666910939138,
+      "loss": 1.3825,
+      "step": 1229
+    },
+    {
+      "epoch": 0.05828002842928216,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001983539792325686,
+      "loss": 1.5757,
+      "step": 1230
+    },
+    {
+      "epoch": 0.058327410566216535,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019835128717435934,
+      "loss": 0.8738,
+      "step": 1231
+    },
+    {
+      "epoch": 0.058374792703150914,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019834859293482328,
+      "loss": 0.8583,
+      "step": 1232
+    },
+    {
+      "epoch": 0.058422174840085286,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001983458965140202,
+      "loss": 0.2097,
+      "step": 1233
+    },
+    {
+      "epoch": 0.058469556977019665,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019834319791200983,
+      "loss": 0.9326,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05851693911395404,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001983404971288521,
+      "loss": 1.0178,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05856432125088842,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019833779416460692,
+      "loss": 1.1593,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05861170338782279,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001983350890193342,
+      "loss": 1.1695,
+      "step": 1237
+    },
+    {
+      "epoch": 0.05865908552475717,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019833238169309395,
+      "loss": 0.4912,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05870646766169154,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001983296721859462,
+      "loss": 0.3485,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05875384979862592,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019832696049795108,
+      "loss": 0.9958,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05880123193556029,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001983242466291687,
+      "loss": 1.262,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05884861407249467,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019832153057965926,
+      "loss": 0.1754,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05889599620942904,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019831881234948296,
+      "loss": 0.0748,
+      "step": 1243
+    },
+    {
+      "epoch": 0.05894337834636342,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019831609193870015,
+      "loss": 0.2001,
+      "step": 1244
+    },
+    {
+      "epoch": 0.058990760483297794,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019831336934737117,
+      "loss": 1.4368,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05903814262023217,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019831064457555636,
+      "loss": 0.8099,
+      "step": 1246
+    },
+    {
+      "epoch": 0.059085524757166545,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019830791762331617,
+      "loss": 1.5033,
+      "step": 1247
+    },
+    {
+      "epoch": 0.059132906894100924,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001983051884907111,
+      "loss": 1.082,
+      "step": 1248
+    },
+    {
+      "epoch": 0.059180289031035296,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001983024571778016,
+      "loss": 0.442,
+      "step": 1249
+    },
+    {
+      "epoch": 0.059227671167969675,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019829972368464835,
+      "loss": 0.855,
+      "step": 1250
+    },
+    {
+      "epoch": 0.059275053304904055,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 0.00019829698801131194,
+      "loss": 0.0025,
+      "step": 1251
+    },
+    {
+      "epoch": 0.05932243544183843,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019829425015785305,
+      "loss": 1.1192,
+      "step": 1252
+    },
+    {
+      "epoch": 0.059369817578772806,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019829151012433233,
+      "loss": 0.1857,
+      "step": 1253
+    },
+    {
+      "epoch": 0.05941719971570718,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019828876791081065,
+      "loss": 0.0894,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05946458185264156,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019828602351734879,
+      "loss": 1.186,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05951196398957593,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001982832769440076,
+      "loss": 0.7192,
+      "step": 1256
+    },
+    {
+      "epoch": 0.05955934612651031,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019828052819084803,
+      "loss": 1.0787,
+      "step": 1257
+    },
+    {
+      "epoch": 0.05960672826344468,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019827777725793104,
+      "loss": 1.3895,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05965411040037906,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001982750241453176,
+      "loss": 0.7737,
+      "step": 1259
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001982722688530688,
+      "loss": 0.8477,
+      "step": 1260
+    },
+    {
+      "epoch": 0.05974887467424781,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019826951138124578,
+      "loss": 0.0817,
+      "step": 1261
+    },
+    {
+      "epoch": 0.05979625681118218,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019826675172990966,
+      "loss": 1.0333,
+      "step": 1262
+    },
+    {
+      "epoch": 0.05984363894811656,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019826398989912167,
+      "loss": 0.9203,
+      "step": 1263
+    },
+    {
+      "epoch": 0.059891021085050934,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019826122588894305,
+      "loss": 0.1183,
+      "step": 1264
+    },
+    {
+      "epoch": 0.05993840322198531,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019825845969943505,
+      "loss": 1.0712,
+      "step": 1265
+    },
+    {
+      "epoch": 0.059985785358919685,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019825569133065913,
+      "loss": 0.9647,
+      "step": 1266
+    },
+    {
+      "epoch": 0.060033167495854065,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001982529207826766,
+      "loss": 0.3805,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06008054963278844,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000198250148055549,
+      "loss": 0.5593,
+      "step": 1268
+    },
+    {
+      "epoch": 0.060127931769722816,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001982473731493377,
+      "loss": 0.9391,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06017531390665719,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019824459606410436,
+      "loss": 0.6489,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06022269604359157,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001982418167999105,
+      "loss": 0.7921,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06027007818052594,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019823903535681777,
+      "loss": 0.7933,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06031746031746032,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001982362517348879,
+      "loss": 0.8601,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06036484245439469,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001982334659341826,
+      "loss": 0.7573,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06041222459132907,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019823067795476365,
+      "loss": 0.2582,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06045960672826344,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001982278877966929,
+      "loss": 0.8929,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06050698886519782,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001982250954600322,
+      "loss": 0.9258,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06055437100213219,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019822230094484355,
+      "loss": 1.0504,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06060175313906657,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019821950425118887,
+      "loss": 0.8885,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06064913527600095,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019821670537913022,
+      "loss": 0.1725,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06069651741293532,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001982139043287296,
+      "loss": 0.65,
+      "step": 1281
+    },
+    {
+      "epoch": 0.0607438995498697,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001982111011000493,
+      "loss": 0.0851,
+      "step": 1282
+    },
+    {
+      "epoch": 0.060791281686804075,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001982082956931513,
+      "loss": 1.096,
+      "step": 1283
+    },
+    {
+      "epoch": 0.060838663823738454,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001982054881080979,
+      "loss": 0.0401,
+      "step": 1284
+    },
+    {
+      "epoch": 0.060886045960672826,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019820267834495144,
+      "loss": 1.1779,
+      "step": 1285
+    },
+    {
+      "epoch": 0.060933428097607205,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019819986640377414,
+      "loss": 1.2213,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06098081023454158,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019819705228462842,
+      "loss": 1.2332,
+      "step": 1287
+    },
+    {
+      "epoch": 0.061028192371475956,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019819423598757668,
+      "loss": 1.1858,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06107557450841033,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019819141751268134,
+      "loss": 0.0969,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06112295664534471,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019818859686000496,
+      "loss": 1.1394,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06117033878227908,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019818577402961013,
+      "loss": 1.091,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06121772091921346,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019818294902155937,
+      "loss": 0.5775,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06126510305614783,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019818012183591538,
+      "loss": 0.8449,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06131248519308221,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001981772924727409,
+      "loss": 1.2039,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06135986733001658,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001981744609320986,
+      "loss": 0.0453,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06140724946695096,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019817162721405134,
+      "loss": 0.814,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06145463160388533,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019816879131866197,
+      "loss": 0.6862,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06150201374081971,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019816595324599335,
+      "loss": 0.1128,
+      "step": 1298
+    },
+    {
+      "epoch": 0.061549395877754084,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001981631129961085,
+      "loss": 0.0284,
+      "step": 1299
+    },
+    {
+      "epoch": 0.061596778014688464,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019816027056907034,
+      "loss": 1.3758,
+      "step": 1300
+    },
+    {
+      "epoch": 0.061644160151622836,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019815742596494192,
+      "loss": 0.558,
+      "step": 1301
+    },
+    {
+      "epoch": 0.061691542288557215,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019815457918378635,
+      "loss": 0.3337,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06173892442549159,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019815173022566675,
+      "loss": 1.1537,
+      "step": 1303
+    },
+    {
+      "epoch": 0.061786306562425966,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019814887909064632,
+      "loss": 1.2278,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06183368869936034,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019814602577878835,
+      "loss": 1.1332,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06188107083629472,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000198143170290156,
+      "loss": 1.1716,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06192845297322909,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001981403126248127,
+      "loss": 0.3662,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06197583511016347,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001981374527828218,
+      "loss": 1.4196,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06202321724709784,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019813459076424672,
+      "loss": 0.5595,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06207059938403222,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019813172656915094,
+      "loss": 0.2389,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0621179815209666,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000198128860197598,
+      "loss": 0.6669,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06216536365790097,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019812599164965148,
+      "loss": 0.5671,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06221274579483535,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019812312092537497,
+      "loss": 1.0529,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06226012793176972,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019812024802483212,
+      "loss": 0.9078,
+      "step": 1314
+    },
+    {
+      "epoch": 0.0623075100687041,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019811737294808673,
+      "loss": 1.1995,
+      "step": 1315
+    },
+    {
+      "epoch": 0.062354892205638474,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019811449569520252,
+      "loss": 1.0019,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06240227434257285,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019811161626624328,
+      "loss": 0.8394,
+      "step": 1317
+    },
+    {
+      "epoch": 0.062449656479507225,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019810873466127288,
+      "loss": 1.4806,
+      "step": 1318
+    },
+    {
+      "epoch": 0.062497038616441604,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019810585088035526,
+      "loss": 0.7276,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06254442075337598,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001981029649235544,
+      "loss": 0.8709,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06259180289031036,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019810007679093422,
+      "loss": 0.0994,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06263918502724473,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019809718648255888,
+      "loss": 1.3156,
+      "step": 1322
+    },
+    {
+      "epoch": 0.0626865671641791,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019809429399849238,
+      "loss": 0.8856,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06273394930111348,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019809139933879897,
+      "loss": 1.0598,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06278133143804786,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019808850250354278,
+      "loss": 1.2765,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06282871357498224,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019808560349278808,
+      "loss": 1.4646,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0628760957119166,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001980827023065992,
+      "loss": 1.0001,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06292347784885098,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019807979894504043,
+      "loss": 0.1755,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06297085998578536,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019807689340817618,
+      "loss": 1.2925,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06301824212271974,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001980739856960709,
+      "loss": 0.8089,
+      "step": 1330
+    },
+    {
+      "epoch": 0.0630656242596541,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001980710758087891,
+      "loss": 1.3729,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06311300639658848,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001980681637463953,
+      "loss": 0.2018,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06316038853352286,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00019806524950895406,
+      "loss": 0.626,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06320777067045724,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019806233309653008,
+      "loss": 1.0182,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06325515280739161,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019805941450918798,
+      "loss": 1.1142,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06330253494432599,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001980564937469925,
+      "loss": 0.4311,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06334991708126037,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019805357081000845,
+      "loss": 0.3452,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06339729921819474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019805064569830067,
+      "loss": 0.0286,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06344468135512911,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000198047718411934,
+      "loss": 1.5651,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06349206349206349,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019804478895097335,
+      "loss": 0.9273,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06353944562899787,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019804185731548367,
+      "loss": 0.4678,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06358682776593225,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001980389235055301,
+      "loss": 0.9688,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06363420990286663,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001980359875211776,
+      "loss": 0.4255,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06368159203980099,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019803304936249133,
+      "loss": 1.298,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06372897417673537,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019803010902953643,
+      "loss": 0.4256,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06377635631366975,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019802716652237811,
+      "loss": 1.1817,
+      "step": 1346
+    },
+    {
+      "epoch": 0.06382373845060413,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001980242218410817,
+      "loss": 0.2303,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0638711205875385,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019802127498571244,
+      "loss": 0.8684,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06391850272447287,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019801832595633568,
+      "loss": 1.2578,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06396588486140725,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001980153747530169,
+      "loss": 0.9623,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06401326699834163,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019801242137582148,
+      "loss": 0.8354,
+      "step": 1351
+    },
+    {
+      "epoch": 0.064060649135276,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019800946582481497,
+      "loss": 0.9811,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06410803127221038,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001980065081000629,
+      "loss": 0.7545,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06415541340914475,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019800354820163088,
+      "loss": 0.7627,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06420279554607913,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019800058612958453,
+      "loss": 0.6435,
+      "step": 1355
+    },
+    {
+      "epoch": 0.0642501776830135,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019799762188398953,
+      "loss": 0.9989,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06429755981994788,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019799465546491173,
+      "loss": 0.8533,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06434494195688226,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001979916868724168,
+      "loss": 0.2298,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06439232409381664,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019798871610657068,
+      "loss": 1.1798,
+      "step": 1359
+    },
+    {
+      "epoch": 0.064439706230751,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019798574316743915,
+      "loss": 0.8694,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06448708836768538,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019798276805508826,
+      "loss": 1.2998,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06453447050461976,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001979797907695839,
+      "loss": 0.8954,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06458185264155414,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001979768113109922,
+      "loss": 1.4044,
+      "step": 1363
+    },
+    {
+      "epoch": 0.0646292347784885,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019797382967937912,
+      "loss": 0.919,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06467661691542288,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001979708458748109,
+      "loss": 0.0239,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06472399905235726,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019796785989735364,
+      "loss": 1.1159,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06477138118929164,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019796487174707363,
+      "loss": 0.0594,
+      "step": 1367
+    },
+    {
+      "epoch": 0.064818763326226,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019796188142403708,
+      "loss": 1.0021,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06486614546316039,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001979588889283104,
+      "loss": 0.4291,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06491352760009476,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019795589425995985,
+      "loss": 1.3191,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06496090973702914,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001979528974190519,
+      "loss": 1.1967,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06500829187396352,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019794989840565307,
+      "loss": 1.1394,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06505567401089789,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019794689721982977,
+      "loss": 0.9012,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06510305614783227,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019794389386164864,
+      "loss": 0.1004,
+      "step": 1374
+    },
+    {
+      "epoch": 0.06515043828476665,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019794088833117627,
+      "loss": 0.4262,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06519782042170102,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019793788062847932,
+      "loss": 1.051,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06524520255863539,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019793487075362448,
+      "loss": 0.4535,
+      "step": 1377
+    },
+    {
+      "epoch": 0.06529258469556977,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019793185870667856,
+      "loss": 0.2997,
+      "step": 1378
+    },
+    {
+      "epoch": 0.06533996683250415,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019792884448770827,
+      "loss": 0.7945,
+      "step": 1379
+    },
+    {
+      "epoch": 0.06538734896943853,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019792582809678057,
+      "loss": 0.6099,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06543473110637289,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001979228095339623,
+      "loss": 0.5786,
+      "step": 1381
+    },
+    {
+      "epoch": 0.06548211324330727,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019791978879932042,
+      "loss": 0.0349,
+      "step": 1382
+    },
+    {
+      "epoch": 0.06552949538024165,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019791676589292189,
+      "loss": 0.3752,
+      "step": 1383
+    },
+    {
+      "epoch": 0.06557687751717603,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001979137408148338,
+      "loss": 0.8369,
+      "step": 1384
+    },
+    {
+      "epoch": 0.0656242596541104,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.00019791071356512326,
+      "loss": 0.0036,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06567164179104477,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019790768414385736,
+      "loss": 0.0213,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06571902392797915,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019790465255110334,
+      "loss": 0.1986,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06576640606491353,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019790161878692836,
+      "loss": 1.0587,
+      "step": 1388
+    },
+    {
+      "epoch": 0.0658137882018479,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001978985828513998,
+      "loss": 1.2339,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06586117033878228,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019789554474458493,
+      "loss": 1.2089,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06590855247571666,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019789250446655116,
+      "loss": 1.0333,
+      "step": 1391
+    },
+    {
+      "epoch": 0.06595593461265103,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001978894620173659,
+      "loss": 0.8205,
+      "step": 1392
+    },
+    {
+      "epoch": 0.0660033167495854,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019788641739709663,
+      "loss": 0.6403,
+      "step": 1393
+    },
+    {
+      "epoch": 0.06605069888651978,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019788337060581092,
+      "loss": 0.7673,
+      "step": 1394
+    },
+    {
+      "epoch": 0.06609808102345416,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019788032164357627,
+      "loss": 0.7747,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06614546316038854,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001978772705104604,
+      "loss": 0.916,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0661928452973229,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00019787421720653087,
+      "loss": 0.1706,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06624022743425728,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019787116173185546,
+      "loss": 1.6059,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06628760957119166,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019786810408650195,
+      "loss": 0.4818,
+      "step": 1399
+    },
+    {
+      "epoch": 0.06633499170812604,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019786504427053814,
+      "loss": 0.3894,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06638237384506042,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001978619822840319,
+      "loss": 0.8414,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06642975598199478,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019785891812705106,
+      "loss": 0.6433,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06647713811892916,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019785585179966372,
+      "loss": 1.1861,
+      "step": 1403
+    },
+    {
+      "epoch": 0.06652452025586354,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001978527833019378,
+      "loss": 0.5291,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06657190239279792,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019784971263394136,
+      "loss": 1.133,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06661928452973229,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019784663979574254,
+      "loss": 0.9909,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019784356478740945,
+      "loss": 0.7536,
+      "step": 1407
+    },
+    {
+      "epoch": 0.06671404880360104,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019784048760901031,
+      "loss": 0.0098,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06676143094053542,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001978374082606134,
+      "loss": 1.0359,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06680881307746979,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019783432674228696,
+      "loss": 0.9342,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06685619521440417,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001978312430540994,
+      "loss": 1.0898,
+      "step": 1411
+    },
+    {
+      "epoch": 0.06690357735133855,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000197828157196119,
+      "loss": 1.1065,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06695095948827293,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019782506916841435,
+      "loss": 0.5028,
+      "step": 1413
+    },
+    {
+      "epoch": 0.06699834162520729,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019782197897105384,
+      "loss": 1.3315,
+      "step": 1414
+    },
+    {
+      "epoch": 0.06704572376214167,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019781888660410602,
+      "loss": 0.0758,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06709310589907605,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001978157920676395,
+      "loss": 1.1478,
+      "step": 1416
+    },
+    {
+      "epoch": 0.06714048803601043,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019781269536172288,
+      "loss": 0.1899,
+      "step": 1417
+    },
+    {
+      "epoch": 0.0671878701729448,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019780959648642489,
+      "loss": 1.3329,
+      "step": 1418
+    },
+    {
+      "epoch": 0.06723525230987917,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019780649544181423,
+      "loss": 1.0074,
+      "step": 1419
+    },
+    {
+      "epoch": 0.06728263444681355,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019780339222795964,
+      "loss": 1.6396,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06733001658374793,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019780028684493,
+      "loss": 0.5259,
+      "step": 1421
+    },
+    {
+      "epoch": 0.0673773987206823,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019779717929279422,
+      "loss": 1.2355,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06742478085761668,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001977940695716211,
+      "loss": 0.1711,
+      "step": 1423
+    },
+    {
+      "epoch": 0.06747216299455105,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019779095768147973,
+      "loss": 0.0939,
+      "step": 1424
+    },
+    {
+      "epoch": 0.06751954513148543,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001977878436224391,
+      "loss": 0.0777,
+      "step": 1425
+    },
+    {
+      "epoch": 0.0675669272684198,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001977847273945682,
+      "loss": 0.2649,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06761430940535418,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019778160899793624,
+      "loss": 1.2112,
+      "step": 1427
+    },
+    {
+      "epoch": 0.06766169154228856,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019777848843261232,
+      "loss": 1.1193,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06770907367922294,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001977753656986657,
+      "loss": 0.0224,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06775645581615732,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019777224079616562,
+      "loss": 1.253,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06780383795309168,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019776911372518135,
+      "loss": 0.5958,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06785122009002606,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019776598448578229,
+      "loss": 1.0039,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06789860222696044,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019776285307803782,
+      "loss": 1.2828,
+      "step": 1433
+    },
+    {
+      "epoch": 0.06794598436389482,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019775971950201742,
+      "loss": 1.4507,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06799336650082918,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019775658375779056,
+      "loss": 0.9144,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06804074863776356,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001977534458454268,
+      "loss": 1.1339,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06808813077469794,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019775030576499574,
+      "loss": 0.8736,
+      "step": 1437
+    },
+    {
+      "epoch": 0.06813551291163232,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.000197747163516567,
+      "loss": 0.0034,
+      "step": 1438
+    },
+    {
+      "epoch": 0.06818289504856669,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019774401910021031,
+      "loss": 0.9527,
+      "step": 1439
+    },
+    {
+      "epoch": 0.06823027718550106,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019774087251599537,
+      "loss": 0.6436,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06827765932243544,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.000197737723763992,
+      "loss": 0.2099,
+      "step": 1441
+    },
+    {
+      "epoch": 0.06832504145936982,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019773457284427,
+      "loss": 1.2293,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06837242359630419,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001977314197568993,
+      "loss": 0.7126,
+      "step": 1443
+    },
+    {
+      "epoch": 0.06841980573323857,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019772826450194982,
+      "loss": 0.0341,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06846718787017295,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019772510707949154,
+      "loss": 1.6576,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06851457000710733,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019772194748959442,
+      "loss": 1.482,
+      "step": 1446
+    },
+    {
+      "epoch": 0.06856195214404169,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001977187857323286,
+      "loss": 0.8749,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06860933428097607,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019771562180776424,
+      "loss": 0.2753,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06865671641791045,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019771245571597142,
+      "loss": 0.1535,
+      "step": 1449
+    },
+    {
+      "epoch": 0.06870409855484483,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019770928745702047,
+      "loss": 0.5296,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0687514806917792,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019770611703098155,
+      "loss": 0.1664,
+      "step": 1451
+    },
+    {
+      "epoch": 0.06879886282871357,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019770294443792507,
+      "loss": 1.132,
+      "step": 1452
+    },
+    {
+      "epoch": 0.06884624496564795,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00019769976967792132,
+      "loss": 0.0082,
+      "step": 1453
+    },
+    {
+      "epoch": 0.06889362710258233,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019769659275104074,
+      "loss": 1.4106,
+      "step": 1454
+    },
+    {
+      "epoch": 0.0689410092395167,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019769341365735382,
+      "loss": 0.3698,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06898839137645107,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019769023239693103,
+      "loss": 0.9214,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06903577351338545,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019768704896984293,
+      "loss": 0.7105,
+      "step": 1457
+    },
+    {
+      "epoch": 0.06908315565031983,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019768386337616013,
+      "loss": 1.3286,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06913053778725421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001976806756159533,
+      "loss": 1.117,
+      "step": 1459
+    },
+    {
+      "epoch": 0.06917791992418858,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019767748568929317,
+      "loss": 0.8185,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06922530206112296,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019767429359625037,
+      "loss": 1.0807,
+      "step": 1461
+    },
+    {
+      "epoch": 0.06927268419805734,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019767109933689584,
+      "loss": 0.8929,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06932006633499171,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019766790291130036,
+      "loss": 0.708,
+      "step": 1463
+    },
+    {
+      "epoch": 0.06936744847192608,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 0.00019766470431953478,
+      "loss": 0.0032,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06941483060886046,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001976615035616701,
+      "loss": 0.58,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06946221274579484,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019765830063777734,
+      "loss": 0.0398,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06950959488272922,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019765509554792746,
+      "loss": 1.1605,
+      "step": 1467
+    },
+    {
+      "epoch": 0.06955697701966358,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019765188829219156,
+      "loss": 1.0195,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06960435915659796,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001976486788706408,
+      "loss": 1.1521,
+      "step": 1469
+    },
+    {
+      "epoch": 0.06965174129353234,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019764546728334636,
+      "loss": 0.8,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06969912343046672,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019764225353037946,
+      "loss": 1.2107,
+      "step": 1471
+    },
+    {
+      "epoch": 0.06974650556740108,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001976390376118114,
+      "loss": 1.4391,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06979388770433546,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019763581952771347,
+      "loss": 0.6741,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06984126984126984,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019763259927815704,
+      "loss": 0.7293,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06988865197820422,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001976293768632136,
+      "loss": 0.9597,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06993603411513859,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001976261522829545,
+      "loss": 0.9556,
+      "step": 1476
+    },
+    {
+      "epoch": 0.06998341625207297,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019762292553745142,
+      "loss": 1.0678,
+      "step": 1477
+    },
+    {
+      "epoch": 0.07003079838900735,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019761969662677578,
+      "loss": 1.1374,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07007818052594172,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019761646555099924,
+      "loss": 1.0522,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07012556266287609,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019761323231019348,
+      "loss": 1.129,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07017294479981047,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019760999690443022,
+      "loss": 0.1712,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07022032693674485,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001976067593337812,
+      "loss": 1.0533,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07026770907367923,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019760351959831817,
+      "loss": 1.1679,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07031509121061359,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001976002776981131,
+      "loss": 0.9336,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07036247334754797,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001975970336332378,
+      "loss": 0.0366,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07040985548448235,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019759378740376426,
+      "loss": 0.9989,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07045723762141673,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019759053900976446,
+      "loss": 0.8363,
+      "step": 1487
+    },
+    {
+      "epoch": 0.0705046197583511,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019758728845131044,
+      "loss": 0.4774,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07055200189528547,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019758403572847435,
+      "loss": 0.1674,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07059938403221985,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019758078084132827,
+      "loss": 1.1625,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07064676616915423,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019757752378994438,
+      "loss": 1.334,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07069414830608861,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019757426457439497,
+      "loss": 1.1302,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07074153044302298,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019757100319475233,
+      "loss": 0.3671,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07078891257995736,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019756773965108875,
+      "loss": 0.384,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07083629471689173,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001975644739434766,
+      "loss": 0.8549,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07088367685382611,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019756120607198835,
+      "loss": 0.8868,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07093105899076048,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001975579360366965,
+      "loss": 1.2682,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07097844112769486,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001975546638376735,
+      "loss": 0.968,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07102582326462924,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000197551389474992,
+      "loss": 1.1436,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07107320540156362,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019754811294872456,
+      "loss": 0.1859,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07112058753849798,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001975448342589439,
+      "loss": 0.9387,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07116796967543236,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019754155340572272,
+      "loss": 0.7409,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07121535181236674,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019753827038913375,
+      "loss": 1.5493,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07126273394930112,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019753498520924987,
+      "loss": 0.1718,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07131011608623548,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001975316978661439,
+      "loss": 1.1719,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07135749822316986,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019752840835988872,
+      "loss": 1.3937,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07140488036010424,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019752511669055738,
+      "loss": 1.2344,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07145226249703862,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001975218228582228,
+      "loss": 0.1861,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07149964463397299,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019751852686295806,
+      "loss": 1.4951,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07154702677090737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001975152287048363,
+      "loss": 0.6003,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07159440890784174,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019751192838393062,
+      "loss": 1.3992,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07164179104477612,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019750862590031424,
+      "loss": 0.9387,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07168917318171049,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019750532125406035,
+      "loss": 1.4259,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07173655531864487,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019750201444524235,
+      "loss": 1.1457,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07178393745557925,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0001974987054739335,
+      "loss": 0.0032,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07183131959251363,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001974953943402072,
+      "loss": 0.745,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07187870172944799,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019749208104413694,
+      "loss": 1.2319,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07192608386638237,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019748876558579612,
+      "loss": 1.3171,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07197346600331675,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019748544796525835,
+      "loss": 0.0755,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07202084814025113,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001974821281825972,
+      "loss": 1.3535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07206823027718551,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019747880623788625,
+      "loss": 0.876,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07211561241411987,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019747548213119918,
+      "loss": 0.7843,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07216299455105425,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019747215586260982,
+      "loss": 1.4776,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07221037668798863,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001974688274321918,
+      "loss": 0.6193,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07225775882492301,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019746549684001902,
+      "loss": 0.9984,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07230514096185738,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019746216408616536,
+      "loss": 1.0317,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07235252309879175,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001974588291707047,
+      "loss": 0.0613,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07239990523572613,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000197455492093711,
+      "loss": 0.7384,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07244728737266051,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001974521528552583,
+      "loss": 0.0511,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07249466950959488,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019744881145542068,
+      "loss": 0.969,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07254205164652926,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001974454678942722,
+      "loss": 0.6165,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07258943378346364,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000197442122171887,
+      "loss": 0.8376,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07263681592039802,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019743877428833934,
+      "loss": 0.1542,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07268419805733238,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019743542424370346,
+      "loss": 0.0372,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07273158019426676,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019743207203805368,
+      "loss": 1.7193,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07277896233120114,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019742871767146428,
+      "loss": 1.0667,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07282634446813552,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019742536114400973,
+      "loss": 0.1448,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07287372660506988,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019742200245576443,
+      "loss": 1.2105,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07292110874200426,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001974186416068029,
+      "loss": 1.3034,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07296849087893864,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019741527859719966,
+      "loss": 1.3104,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07301587301587302,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019741191342702928,
+      "loss": 0.1806,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07306325515280739,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019740854609636644,
+      "loss": 1.0215,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07311063728974176,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019740517660528579,
+      "loss": 1.149,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07315801942667614,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001974018049538621,
+      "loss": 0.8894,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07320540156361052,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019739843114217006,
+      "loss": 1.1542,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07325278370054489,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019739505517028463,
+      "loss": 1.0084,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07330016583747927,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019739167703828058,
+      "loss": 0.6863,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07334754797441365,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001973882967462329,
+      "loss": 0.8286,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07339493011134803,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001973849142942165,
+      "loss": 0.8003,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0734423122482824,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019738152968230645,
+      "loss": 0.8031,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07348969438521677,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001973781429105778,
+      "loss": 0.6852,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07353707652215115,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019737475397910563,
+      "loss": 0.027,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07358445865908553,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019737136288796515,
+      "loss": 0.1264,
+      "step": 1553
+    },
+    {
+      "epoch": 0.0736318407960199,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001973679696372316,
+      "loss": 0.6384,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07367922293295427,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019736457422698015,
+      "loss": 0.8887,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07372660506988865,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019736117665728617,
+      "loss": 0.9889,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07377398720682303,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000197357776928225,
+      "loss": 1.2968,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07382136934375741,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019735437503987202,
+      "loss": 0.8644,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07386875148069177,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001973509709923027,
+      "loss": 1.1251,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07391613361762615,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019734756478559254,
+      "loss": 0.2686,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07396351575456053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001973441564198171,
+      "loss": 0.7783,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07401089789149491,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019734074589505195,
+      "loss": 1.1619,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07405828002842928,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019733733321137273,
+      "loss": 0.8184,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07410566216536366,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019733391836885514,
+      "loss": 1.2703,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07415304430229804,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001973305013675749,
+      "loss": 0.1781,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07420042643923241,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019732708220760782,
+      "loss": 1.2109,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07424780857616678,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019732366088902976,
+      "loss": 0.7686,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07429519071310116,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019732023741191653,
+      "loss": 0.5418,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07434257285003554,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019731681177634412,
+      "loss": 0.6996,
+      "step": 1569
+    },
+    {
+      "epoch": 0.07438995498696992,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019731338398238846,
+      "loss": 1.1675,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07443733712390428,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001973099540301256,
+      "loss": 0.891,
+      "step": 1571
+    },
+    {
+      "epoch": 0.07448471926083866,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001973065219196316,
+      "loss": 1.2264,
+      "step": 1572
+    },
+    {
+      "epoch": 0.07453210139777304,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00019730308765098263,
+      "loss": 1.1571,
+      "step": 1573
+    },
+    {
+      "epoch": 0.07457948353470742,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001972996512242548,
+      "loss": 0.6126,
+      "step": 1574
+    },
+    {
+      "epoch": 0.07462686567164178,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001972962126395243,
+      "loss": 0.6741,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07467424780857616,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00019729277189686748,
+      "loss": 0.4545,
+      "step": 1576
+    },
+    {
+      "epoch": 0.07472162994551054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001972893289963606,
+      "loss": 0.7829,
+      "step": 1577
+    },
+    {
+      "epoch": 0.07476901208244492,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019728588393808005,
+      "loss": 1.5995,
+      "step": 1578
+    },
+    {
+      "epoch": 0.0748163942193793,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019728243672210223,
+      "loss": 0.4342,
+      "step": 1579
+    },
+    {
+      "epoch": 0.07486377635631367,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019727898734850355,
+      "loss": 0.7056,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07491115849324805,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019727553581736054,
+      "loss": 0.0762,
+      "step": 1581
+    },
+    {
+      "epoch": 0.07495854063018242,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019727208212874978,
+      "loss": 0.3199,
+      "step": 1582
+    },
+    {
+      "epoch": 0.0750059227671168,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019726862628274784,
+      "loss": 1.4885,
+      "step": 1583
+    },
+    {
+      "epoch": 0.07505330490405117,
+      "grad_norm": 0.022216796875,
+      "learning_rate": 0.0001972651682794314,
+      "loss": 0.0019,
+      "step": 1584
+    },
+    {
+      "epoch": 0.07510068704098555,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001972617081188771,
+      "loss": 0.8026,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07514806917791993,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019725824580116172,
+      "loss": 0.9925,
+      "step": 1586
+    },
+    {
+      "epoch": 0.0751954513148543,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019725478132636207,
+      "loss": 1.1016,
+      "step": 1587
+    },
+    {
+      "epoch": 0.07524283345178867,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001972513146945549,
+      "loss": 0.6538,
+      "step": 1588
+    },
+    {
+      "epoch": 0.07529021558872305,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001972478459058172,
+      "loss": 1.4446,
+      "step": 1589
+    },
+    {
+      "epoch": 0.07533759772565743,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019724437496022588,
+      "loss": 0.8096,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07538497986259181,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019724090185785787,
+      "loss": 0.0483,
+      "step": 1591
+    },
+    {
+      "epoch": 0.07543236199952617,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019723742659879024,
+      "loss": 1.3905,
+      "step": 1592
+    },
+    {
+      "epoch": 0.07547974413646055,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019723394918310007,
+      "loss": 0.8611,
+      "step": 1593
+    },
+    {
+      "epoch": 0.07552712627339493,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019723046961086444,
+      "loss": 1.3968,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07557450841032931,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001972269878821606,
+      "loss": 1.1161,
+      "step": 1595
+    },
+    {
+      "epoch": 0.07562189054726368,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001972235039970657,
+      "loss": 1.2614,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07566927268419806,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019722001795565705,
+      "loss": 1.1092,
+      "step": 1597
+    },
+    {
+      "epoch": 0.07571665482113243,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019721652975801196,
+      "loss": 1.2873,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07576403695806681,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001972130394042078,
+      "loss": 1.0667,
+      "step": 1599
+    },
+    {
+      "epoch": 0.07581141909500118,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019720954689432199,
+      "loss": 0.9517,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07585880123193556,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019720605222843193,
+      "loss": 1.1561,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07590618336886994,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001972025554066152,
+      "loss": 1.052,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07595356550580432,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001971990564289493,
+      "loss": 0.2228,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07600094764273868,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001971955552955119,
+      "loss": 1.0376,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07604832977967306,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019719205200638057,
+      "loss": 0.0396,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07609571191660744,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019718854656163308,
+      "loss": 0.8643,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07614309405354182,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019718503896134712,
+      "loss": 1.0733,
+      "step": 1607
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019718152920560056,
+      "loss": 1.1269,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07623785832741056,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00019717801729447114,
+      "loss": 0.4496,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07628524046434494,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019717450322803684,
+      "loss": 1.1536,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07633262260127932,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019717098700637554,
+      "loss": 0.6839,
+      "step": 1611
+    },
+    {
+      "epoch": 0.0763800047382137,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019716746862956527,
+      "loss": 1.3748,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07642738687514807,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019716394809768403,
+      "loss": 1.0119,
+      "step": 1613
+    },
+    {
+      "epoch": 0.07647476901208244,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019716042541080992,
+      "loss": 1.0805,
+      "step": 1614
+    },
+    {
+      "epoch": 0.07652215114901682,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019715690056902108,
+      "loss": 1.3178,
+      "step": 1615
+    },
+    {
+      "epoch": 0.0765695332859512,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019715337357239566,
+      "loss": 1.1677,
+      "step": 1616
+    },
+    {
+      "epoch": 0.07661691542288557,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001971498444210119,
+      "loss": 0.7856,
+      "step": 1617
+    },
+    {
+      "epoch": 0.07666429755981995,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019714631311494807,
+      "loss": 0.8945,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07671167969675433,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001971427796542825,
+      "loss": 1.3548,
+      "step": 1619
+    },
+    {
+      "epoch": 0.0767590618336887,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019713924403909352,
+      "loss": 1.1525,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07680644397062307,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001971357062694596,
+      "loss": 0.0897,
+      "step": 1621
+    },
+    {
+      "epoch": 0.07685382610755745,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001971321663454592,
+      "loss": 0.826,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07690120824449183,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019712862426717075,
+      "loss": 0.8226,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07694859038142621,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001971250800346729,
+      "loss": 0.8715,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07699597251836057,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019712153364804424,
+      "loss": 0.9103,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07704335465529495,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001971179851073634,
+      "loss": 1.124,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07709073679222933,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019711443441270912,
+      "loss": 0.8573,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07713811892916371,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019711088156416012,
+      "loss": 0.524,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07718550106609808,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019710732656179518,
+      "loss": 1.0545,
+      "step": 1629
+    },
+    {
+      "epoch": 0.07723288320303245,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019710376940569317,
+      "loss": 0.681,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07728026533996683,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000197100210095933,
+      "loss": 1.4263,
+      "step": 1631
+    },
+    {
+      "epoch": 0.07732764747690121,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019709664863259358,
+      "loss": 0.2156,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07737502961383558,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019709308501575398,
+      "loss": 1.1149,
+      "step": 1633
+    },
+    {
+      "epoch": 0.07742241175076996,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019708951924549307,
+      "loss": 1.2251,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07746979388770434,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001970859513218901,
+      "loss": 0.7541,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07751717602463871,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019708238124502417,
+      "loss": 1.5339,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07756455816157308,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019707880901497437,
+      "loss": 1.0511,
+      "step": 1637
+    },
+    {
+      "epoch": 0.07761194029850746,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019707523463182,
+      "loss": 0.0204,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07765932243544184,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019707165809564034,
+      "loss": 0.9405,
+      "step": 1639
+    },
+    {
+      "epoch": 0.07770670457237622,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019706807940651473,
+      "loss": 1.1919,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0777540867093106,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019706449856452248,
+      "loss": 0.7183,
+      "step": 1641
+    },
+    {
+      "epoch": 0.07780146884624496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019706091556974303,
+      "loss": 1.4933,
+      "step": 1642
+    },
+    {
+      "epoch": 0.07784885098317934,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019705733042225588,
+      "loss": 1.0473,
+      "step": 1643
+    },
+    {
+      "epoch": 0.07789623312011372,
+      "grad_norm": 0.01409912109375,
+      "learning_rate": 0.0001970537431221405,
+      "loss": 0.001,
+      "step": 1644
+    },
+    {
+      "epoch": 0.0779436152570481,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019705015366947647,
+      "loss": 0.9944,
+      "step": 1645
+    },
+    {
+      "epoch": 0.07799099739398246,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019704656206434343,
+      "loss": 0.9846,
+      "step": 1646
+    },
+    {
+      "epoch": 0.07803837953091684,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019704296830682098,
+      "loss": 0.9973,
+      "step": 1647
+    },
+    {
+      "epoch": 0.07808576166785122,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001970393723969889,
+      "loss": 0.0165,
+      "step": 1648
+    },
+    {
+      "epoch": 0.0781331438047856,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019703577433492688,
+      "loss": 1.2609,
+      "step": 1649
+    },
+    {
+      "epoch": 0.07818052594171997,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019703217412071476,
+      "loss": 0.2079,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07822790807865435,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00019702857175443234,
+      "loss": 0.2338,
+      "step": 1651
+    },
+    {
+      "epoch": 0.07827529021558872,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019702496723615956,
+      "loss": 0.1863,
+      "step": 1652
+    },
+    {
+      "epoch": 0.0783226723525231,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019702136056597635,
+      "loss": 0.7679,
+      "step": 1653
+    },
+    {
+      "epoch": 0.07837005448945747,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019701775174396272,
+      "loss": 1.2124,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07841743662639185,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019701414077019871,
+      "loss": 0.0344,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07846481876332623,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019701052764476437,
+      "loss": 1.0308,
+      "step": 1656
+    },
+    {
+      "epoch": 0.0785122009002606,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019700691236773987,
+      "loss": 0.0132,
+      "step": 1657
+    },
+    {
+      "epoch": 0.07855958303719497,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019700329493920534,
+      "loss": 1.3492,
+      "step": 1658
+    },
+    {
+      "epoch": 0.07860696517412935,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019699967535924108,
+      "loss": 0.1897,
+      "step": 1659
+    },
+    {
+      "epoch": 0.07865434731106373,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019699605362792736,
+      "loss": 1.41,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07870172944799811,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019699242974534445,
+      "loss": 1.0693,
+      "step": 1661
+    },
+    {
+      "epoch": 0.07874911158493247,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001969888037115728,
+      "loss": 1.2154,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07879649372186685,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019698517552669276,
+      "loss": 0.7537,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07884387585880123,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019698154519078484,
+      "loss": 1.3872,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07889125799573561,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019697791270392956,
+      "loss": 1.2254,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07893864013266998,
+      "grad_norm": 0.01055908203125,
+      "learning_rate": 0.00019697427806620744,
+      "loss": 0.0009,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07898602226960436,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019697064127769916,
+      "loss": 0.8843,
+      "step": 1667
+    },
+    {
+      "epoch": 0.07903340440653873,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019696700233848532,
+      "loss": 0.954,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07908078654347311,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019696336124864667,
+      "loss": 0.8493,
+      "step": 1669
+    },
+    {
+      "epoch": 0.07912816868040749,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019695971800826395,
+      "loss": 0.1835,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07917555081734186,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019695607261741794,
+      "loss": 0.2005,
+      "step": 1671
+    },
+    {
+      "epoch": 0.07922293295427624,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019695242507618952,
+      "loss": 0.7346,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07927031509121062,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001969487753846596,
+      "loss": 0.7522,
+      "step": 1673
+    },
+    {
+      "epoch": 0.079317697228145,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019694512354290908,
+      "loss": 0.6101,
+      "step": 1674
+    },
+    {
+      "epoch": 0.07936507936507936,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000196941469551019,
+      "loss": 0.7977,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07941246150201374,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019693781340907038,
+      "loss": 0.2747,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07945984363894812,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019693415511714432,
+      "loss": 0.9906,
+      "step": 1677
+    },
+    {
+      "epoch": 0.0795072257758825,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019693049467532192,
+      "loss": 1.2076,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07955460791281686,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019692683208368444,
+      "loss": 0.1237,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07960199004975124,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019692316734231302,
+      "loss": 0.645,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07964937218668562,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019691950045128901,
+      "loss": 1.1272,
+      "step": 1681
+    },
+    {
+      "epoch": 0.07969675432362,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019691583141069372,
+      "loss": 1.2546,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07974413646055437,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001969121602206085,
+      "loss": 0.1895,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07979151859748874,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001969084868811148,
+      "loss": 0.8877,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07983890073442312,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001969048113922941,
+      "loss": 0.7107,
+      "step": 1685
+    },
+    {
+      "epoch": 0.0798862828713575,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019690113375422789,
+      "loss": 0.0773,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07993366500829187,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001968974539669977,
+      "loss": 1.6357,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07998104714522625,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019689377203068524,
+      "loss": 1.1769,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08002842928216063,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019689008794537212,
+      "loss": 0.2874,
+      "step": 1689
+    },
+    {
+      "epoch": 0.080075811419095,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019688640171114006,
+      "loss": 1.636,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08012319355602937,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001968827133280708,
+      "loss": 0.9001,
+      "step": 1691
+    },
+    {
+      "epoch": 0.08017057569296375,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019687902279624616,
+      "loss": 0.9937,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08021795782989813,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019687533011574797,
+      "loss": 0.0951,
+      "step": 1693
+    },
+    {
+      "epoch": 0.08026533996683251,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019687163528665815,
+      "loss": 1.5956,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08031272210376687,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019686793830905864,
+      "loss": 0.0098,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08036010424070125,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019686423918303144,
+      "loss": 0.219,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08040748637763563,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001968605379086586,
+      "loss": 0.236,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08045486851457001,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001968568344860222,
+      "loss": 0.1505,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08050225065150439,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019685312891520437,
+      "loss": 0.9413,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08054963278843875,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001968494211962873,
+      "loss": 0.0553,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08059701492537313,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019684571132935324,
+      "loss": 1.17,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08064439706230751,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019684199931448444,
+      "loss": 1.0493,
+      "step": 1702
+    },
+    {
+      "epoch": 0.08069177919924189,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019683828515176325,
+      "loss": 0.108,
+      "step": 1703
+    },
+    {
+      "epoch": 0.08073916133617626,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019683456884127205,
+      "loss": 0.8528,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08078654347311064,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019683085038309326,
+      "loss": 0.2052,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08083392561004502,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019682712977730935,
+      "loss": 0.7148,
+      "step": 1706
+    },
+    {
+      "epoch": 0.0808813077469794,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00019682340702400285,
+      "loss": 0.2189,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08092868988391376,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019681968212325628,
+      "loss": 0.773,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08097607202084814,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001968159550751523,
+      "loss": 0.2618,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08102345415778252,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019681222587977356,
+      "loss": 0.6582,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0810708362947169,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019680849453720275,
+      "loss": 1.3603,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08111821843165126,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019680476104752269,
+      "loss": 0.078,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08116560056858564,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019680102541081614,
+      "loss": 1.651,
+      "step": 1713
+    },
+    {
+      "epoch": 0.08121298270552002,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019679728762716592,
+      "loss": 0.8839,
+      "step": 1714
+    },
+    {
+      "epoch": 0.0812603648424544,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019679354769665497,
+      "loss": 0.1678,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08130774697938876,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019678980561936625,
+      "loss": 0.1759,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08135512911632314,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019678606139538274,
+      "loss": 1.0799,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08140251125325752,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019678231502478745,
+      "loss": 0.816,
+      "step": 1718
+    },
+    {
+      "epoch": 0.0814498933901919,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019677856650766353,
+      "loss": 1.0411,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08149727552712627,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019677481584409406,
+      "loss": 0.0126,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08154465766406065,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019677106303416227,
+      "loss": 1.2617,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08159203980099503,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001967673080779514,
+      "loss": 0.6694,
+      "step": 1722
+    },
+    {
+      "epoch": 0.0816394219379294,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019676355097554468,
+      "loss": 1.0204,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08168680407486377,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001967597917270255,
+      "loss": 1.0117,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08173418621179815,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019675603033247717,
+      "loss": 0.4565,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08178156834873253,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001967522667919832,
+      "loss": 1.0873,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08182895048566691,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019674850110562692,
+      "loss": 0.0202,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08187633262260129,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000196744733273492,
+      "loss": 0.1749,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08192371475953565,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019674096329566192,
+      "loss": 0.2461,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08197109689647003,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001967371911722203,
+      "loss": 0.9353,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08201847903340441,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019673341690325087,
+      "loss": 1.253,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08206586117033879,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019672964048883727,
+      "loss": 0.072,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08211324330727315,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019672586192906325,
+      "loss": 0.829,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08216062544420753,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019672208122401268,
+      "loss": 0.2471,
+      "step": 1734
+    },
+    {
+      "epoch": 0.08220800758114191,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019671829837376935,
+      "loss": 0.9004,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08225538971807629,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019671451337841718,
+      "loss": 1.1285,
+      "step": 1736
+    },
+    {
+      "epoch": 0.08230277185501066,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019671072623804012,
+      "loss": 0.0457,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08235015399194504,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019670693695272216,
+      "loss": 1.2687,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08239753612887941,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019670314552254736,
+      "loss": 0.8366,
+      "step": 1739
+    },
+    {
+      "epoch": 0.0824449182658138,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019669935194759978,
+      "loss": 1.333,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08249230040274816,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019669555622796358,
+      "loss": 1.0499,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08253968253968254,
+      "grad_norm": 0.009033203125,
+      "learning_rate": 0.00019669175836372293,
+      "loss": 0.001,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08258706467661692,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001966879583549621,
+      "loss": 0.8029,
+      "step": 1743
+    },
+    {
+      "epoch": 0.0826344468135513,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001966841562017653,
+      "loss": 0.8387,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08268182895048566,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019668035190421694,
+      "loss": 0.2109,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08272921108742004,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019667654546240133,
+      "loss": 0.2607,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08277659322435442,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019667273687640292,
+      "loss": 1.5496,
+      "step": 1747
+    },
+    {
+      "epoch": 0.0828239753612888,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019666892614630618,
+      "loss": 0.4689,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08287135749822316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019666511327219563,
+      "loss": 0.9652,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08291873963515754,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019666129825415582,
+      "loss": 0.789,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08296612177209192,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001966574810922714,
+      "loss": 0.3402,
+      "step": 1751
+    },
+    {
+      "epoch": 0.0830135039090263,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019665366178662697,
+      "loss": 0.1491,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08306088604596067,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001966498403373073,
+      "loss": 0.0646,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08310826818289505,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019664601674439708,
+      "loss": 0.2229,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08315565031982942,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019664219100798118,
+      "loss": 0.0445,
+      "step": 1755
+    },
+    {
+      "epoch": 0.0832030324567638,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001966383631281444,
+      "loss": 0.9004,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08325041459369818,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001966345331049717,
+      "loss": 1.1838,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08329779673063255,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019663070093854797,
+      "loss": 0.8954,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08334517886756693,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019662686662895822,
+      "loss": 0.9397,
+      "step": 1759
+    },
+    {
+      "epoch": 0.0833925610045013,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019662303017628746,
+      "loss": 1.0675,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08343994314143569,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019661919158062084,
+      "loss": 0.0825,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08348732527837005,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019661535084204346,
+      "loss": 0.9899,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08353470741530443,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001966115079606405,
+      "loss": 0.5338,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08358208955223881,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00019660766293649718,
+      "loss": 0.8799,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08362947168917319,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019660381576969886,
+      "loss": 0.9777,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08367685382610755,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019659996646033076,
+      "loss": 0.7217,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08372423596304193,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001965961150084783,
+      "loss": 0.978,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08377161809997631,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001965922614142269,
+      "loss": 0.4204,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08381900023691069,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019658840567766205,
+      "loss": 0.9237,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08386638237384506,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001965845477988692,
+      "loss": 1.1636,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08391376451077943,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000196580687777934,
+      "loss": 0.9592,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08396114664771381,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019657682561494198,
+      "loss": 0.9794,
+      "step": 1772
+    },
+    {
+      "epoch": 0.08400852878464819,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019657296130997886,
+      "loss": 1.0554,
+      "step": 1773
+    },
+    {
+      "epoch": 0.08405591092158256,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019656909486313033,
+      "loss": 0.0215,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08410329305851694,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00019656522627448212,
+      "loss": 0.4928,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08415067519545132,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019656135554412004,
+      "loss": 1.4965,
+      "step": 1776
+    },
+    {
+      "epoch": 0.0841980573323857,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019655748267212998,
+      "loss": 0.1763,
+      "step": 1777
+    },
+    {
+      "epoch": 0.08424543946932006,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019655360765859778,
+      "loss": 1.0559,
+      "step": 1778
+    },
+    {
+      "epoch": 0.08429282160625444,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019654973050360942,
+      "loss": 0.0206,
+      "step": 1779
+    },
+    {
+      "epoch": 0.08434020374318882,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019654585120725085,
+      "loss": 1.4635,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0843875858801232,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019654196976960818,
+      "loss": 1.0031,
+      "step": 1781
+    },
+    {
+      "epoch": 0.08443496801705756,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001965380861907674,
+      "loss": 0.6398,
+      "step": 1782
+    },
+    {
+      "epoch": 0.08448235015399194,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019653420047081473,
+      "loss": 1.2147,
+      "step": 1783
+    },
+    {
+      "epoch": 0.08452973229092632,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001965303126098363,
+      "loss": 0.259,
+      "step": 1784
+    },
+    {
+      "epoch": 0.0845771144278607,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019652642260791837,
+      "loss": 0.988,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08462449656479507,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001965225304651472,
+      "loss": 0.7848,
+      "step": 1786
+    },
+    {
+      "epoch": 0.08467187870172944,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001965186361816091,
+      "loss": 0.022,
+      "step": 1787
+    },
+    {
+      "epoch": 0.08471926083866382,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019651473975739047,
+      "loss": 0.0087,
+      "step": 1788
+    },
+    {
+      "epoch": 0.0847666429755982,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001965108411925777,
+      "loss": 1.2164,
+      "step": 1789
+    },
+    {
+      "epoch": 0.08481402511253258,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019650694048725732,
+      "loss": 0.8642,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08486140724946695,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019650303764151574,
+      "loss": 1.1453,
+      "step": 1791
+    },
+    {
+      "epoch": 0.08490878938640133,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001964991326554396,
+      "loss": 0.0116,
+      "step": 1792
+    },
+    {
+      "epoch": 0.0849561715233357,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019649522552911547,
+      "loss": 0.1656,
+      "step": 1793
+    },
+    {
+      "epoch": 0.08500355366027008,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019649131626263002,
+      "loss": 0.791,
+      "step": 1794
+    },
+    {
+      "epoch": 0.08505093579720445,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019648740485606996,
+      "loss": 0.0354,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08509831793413883,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019648349130952207,
+      "loss": 0.9545,
+      "step": 1796
+    },
+    {
+      "epoch": 0.08514570007107321,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019647957562307305,
+      "loss": 0.036,
+      "step": 1797
+    },
+    {
+      "epoch": 0.08519308220800759,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019647565779680983,
+      "loss": 0.1932,
+      "step": 1798
+    },
+    {
+      "epoch": 0.08524046434494195,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019647173783081932,
+      "loss": 0.8699,
+      "step": 1799
+    },
+    {
+      "epoch": 0.08528784648187633,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.00019646781572518838,
+      "loss": 0.0034,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08533522861881071,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019646389148000404,
+      "loss": 0.0264,
+      "step": 1801
+    },
+    {
+      "epoch": 0.08538261075574509,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019645996509535334,
+      "loss": 0.0255,
+      "step": 1802
+    },
+    {
+      "epoch": 0.08542999289267945,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019645603657132335,
+      "loss": 1.1372,
+      "step": 1803
+    },
+    {
+      "epoch": 0.08547737502961383,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019645210590800124,
+      "loss": 1.076,
+      "step": 1804
+    },
+    {
+      "epoch": 0.08552475716654821,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019644817310547416,
+      "loss": 0.0127,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08557213930348259,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001964442381638293,
+      "loss": 0.7601,
+      "step": 1806
+    },
+    {
+      "epoch": 0.08561952144041696,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019644030108315397,
+      "loss": 1.2834,
+      "step": 1807
+    },
+    {
+      "epoch": 0.08566690357735134,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001964363618635355,
+      "loss": 0.9358,
+      "step": 1808
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019643242050506124,
+      "loss": 0.1022,
+      "step": 1809
+    },
+    {
+      "epoch": 0.0857616678512201,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001964284770078186,
+      "loss": 1.0071,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08580904998815446,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019642453137189505,
+      "loss": 0.0967,
+      "step": 1811
+    },
+    {
+      "epoch": 0.08585643212508884,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00019642058359737813,
+      "loss": 0.9234,
+      "step": 1812
+    },
+    {
+      "epoch": 0.08590381426202322,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019641663368435533,
+      "loss": 1.2388,
+      "step": 1813
+    },
+    {
+      "epoch": 0.0859511963989576,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001964126816329143,
+      "loss": 0.007,
+      "step": 1814
+    },
+    {
+      "epoch": 0.08599857853589196,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001964087274431427,
+      "loss": 0.0974,
+      "step": 1815
+    },
+    {
+      "epoch": 0.08604596067282634,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001964047711151282,
+      "loss": 0.4457,
+      "step": 1816
+    },
+    {
+      "epoch": 0.08609334280976072,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019640081264895857,
+      "loss": 1.1292,
+      "step": 1817
+    },
+    {
+      "epoch": 0.0861407249466951,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001963968520447216,
+      "loss": 0.7719,
+      "step": 1818
+    },
+    {
+      "epoch": 0.08618810708362948,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019639288930250516,
+      "loss": 0.7508,
+      "step": 1819
+    },
+    {
+      "epoch": 0.08623548922056384,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019638892442239706,
+      "loss": 1.1699,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08628287135749822,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019638495740448528,
+      "loss": 0.9526,
+      "step": 1821
+    },
+    {
+      "epoch": 0.0863302534944326,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019638098824885784,
+      "loss": 1.1326,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08637763563136698,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019637701695560274,
+      "loss": 0.1677,
+      "step": 1823
+    },
+    {
+      "epoch": 0.08642501776830135,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019637304352480806,
+      "loss": 1.0371,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08647239990523573,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001963690679565619,
+      "loss": 0.208,
+      "step": 1825
+    },
+    {
+      "epoch": 0.0865197820421701,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019636509025095248,
+      "loss": 0.0226,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08656716417910448,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019636111040806798,
+      "loss": 0.1305,
+      "step": 1827
+    },
+    {
+      "epoch": 0.08661454631603885,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019635712842799673,
+      "loss": 1.0012,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08666192845297323,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019635314431082698,
+      "loss": 0.9523,
+      "step": 1829
+    },
+    {
+      "epoch": 0.0867093105899076,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019634915805664713,
+      "loss": 0.7971,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08675669272684199,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001963451696655456,
+      "loss": 0.8661,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08680407486377635,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001963411791376108,
+      "loss": 1.1867,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08685145700071073,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019633718647293127,
+      "loss": 0.7809,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08689883913764511,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019633319167159557,
+      "loss": 0.9822,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08694622127457949,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019632919473369228,
+      "loss": 1.0346,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08699360341151385,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019632519565931006,
+      "loss": 0.1719,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08704098554844823,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019632119444853757,
+      "loss": 0.3176,
+      "step": 1837
+    },
+    {
+      "epoch": 0.08708836768538261,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019631719110146362,
+      "loss": 1.0714,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08713574982231699,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019631318561817696,
+      "loss": 0.2127,
+      "step": 1839
+    },
+    {
+      "epoch": 0.08718313195925136,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019630917799876642,
+      "loss": 1.0841,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08723051409618574,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001963051682433209,
+      "loss": 0.55,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08727789623312011,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019630115635192933,
+      "loss": 0.7196,
+      "step": 1842
+    },
+    {
+      "epoch": 0.0873252783700545,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001962971423246807,
+      "loss": 1.0728,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08737266050698886,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000196293126161664,
+      "loss": 0.0204,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08742004264392324,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019628910786296833,
+      "loss": 0.126,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08746742478085762,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019628508742868285,
+      "loss": 1.0929,
+      "step": 1846
+    },
+    {
+      "epoch": 0.087514806917792,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019628106485889663,
+      "loss": 1.3432,
+      "step": 1847
+    },
+    {
+      "epoch": 0.08756218905472637,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000196277040153699,
+      "loss": 1.217,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08760957119166074,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001962730133131791,
+      "loss": 1.0241,
+      "step": 1849
+    },
+    {
+      "epoch": 0.08765695332859512,
+      "grad_norm": 0.01507568359375,
+      "learning_rate": 0.00019626898433742636,
+      "loss": 0.0011,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0877043354655295,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001962649532265301,
+      "loss": 1.0595,
+      "step": 1851
+    },
+    {
+      "epoch": 0.08775171760246388,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001962609199805797,
+      "loss": 0.9597,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08779909973939824,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019625688459966464,
+      "loss": 0.3966,
+      "step": 1853
+    },
+    {
+      "epoch": 0.08784648187633262,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019625284708387435,
+      "loss": 0.17,
+      "step": 1854
+    },
+    {
+      "epoch": 0.087893864013267,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00019624880743329847,
+      "loss": 0.006,
+      "step": 1855
+    },
+    {
+      "epoch": 0.08794124615020138,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019624476564802657,
+      "loss": 0.2036,
+      "step": 1856
+    },
+    {
+      "epoch": 0.08798862828713575,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019624072172814828,
+      "loss": 1.4628,
+      "step": 1857
+    },
+    {
+      "epoch": 0.08803601042407012,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001962366756737533,
+      "loss": 0.377,
+      "step": 1858
+    },
+    {
+      "epoch": 0.0880833925610045,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019623262748493135,
+      "loss": 1.082,
+      "step": 1859
+    },
+    {
+      "epoch": 0.08813077469793888,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019622857716177224,
+      "loss": 0.1617,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08817815683487325,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019622452470436578,
+      "loss": 1.0647,
+      "step": 1861
+    },
+    {
+      "epoch": 0.08822553897180763,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019622047011280184,
+      "loss": 1.0775,
+      "step": 1862
+    },
+    {
+      "epoch": 0.088272921108742,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019621641338717037,
+      "loss": 1.3918,
+      "step": 1863
+    },
+    {
+      "epoch": 0.08832030324567638,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019621235452756136,
+      "loss": 0.1497,
+      "step": 1864
+    },
+    {
+      "epoch": 0.08836768538261075,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019620829353406477,
+      "loss": 1.1239,
+      "step": 1865
+    },
+    {
+      "epoch": 0.08841506751954513,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019620423040677074,
+      "loss": 0.9789,
+      "step": 1866
+    },
+    {
+      "epoch": 0.08846244965647951,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019620016514576932,
+      "loss": 1.0354,
+      "step": 1867
+    },
+    {
+      "epoch": 0.08850983179341389,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019619609775115072,
+      "loss": 1.2406,
+      "step": 1868
+    },
+    {
+      "epoch": 0.08855721393034825,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019619202822300513,
+      "loss": 1.5745,
+      "step": 1869
+    },
+    {
+      "epoch": 0.08860459606728263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001961879565614228,
+      "loss": 1.1745,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08865197820421701,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019618388276649407,
+      "loss": 0.2836,
+      "step": 1871
+    },
+    {
+      "epoch": 0.08869936034115139,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019617980683830926,
+      "loss": 0.379,
+      "step": 1872
+    },
+    {
+      "epoch": 0.08874674247808576,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00019617572877695882,
+      "loss": 0.1641,
+      "step": 1873
+    },
+    {
+      "epoch": 0.08879412461502013,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019617164858253311,
+      "loss": 0.0742,
+      "step": 1874
+    },
+    {
+      "epoch": 0.08884150675195451,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001961675662551227,
+      "loss": 1.1144,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019616348179481808,
+      "loss": 0.0538,
+      "step": 1876
+    },
+    {
+      "epoch": 0.08893627102582327,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019615939520170988,
+      "loss": 0.9442,
+      "step": 1877
+    },
+    {
+      "epoch": 0.08898365316275764,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019615530647588872,
+      "loss": 1.5793,
+      "step": 1878
+    },
+    {
+      "epoch": 0.08903103529969202,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019615121561744524,
+      "loss": 0.7052,
+      "step": 1879
+    },
+    {
+      "epoch": 0.0890784174366264,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019614712262647026,
+      "loss": 1.4601,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08912579957356077,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019614302750305452,
+      "loss": 1.1465,
+      "step": 1881
+    },
+    {
+      "epoch": 0.08917318171049514,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001961389302472888,
+      "loss": 1.4365,
+      "step": 1882
+    },
+    {
+      "epoch": 0.08922056384742952,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000196134830859264,
+      "loss": 1.0184,
+      "step": 1883
+    },
+    {
+      "epoch": 0.0892679459843639,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001961307293390711,
+      "loss": 1.4051,
+      "step": 1884
+    },
+    {
+      "epoch": 0.08931532812129828,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019612662568680098,
+      "loss": 0.7338,
+      "step": 1885
+    },
+    {
+      "epoch": 0.08936271025823264,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001961225199025447,
+      "loss": 1.3631,
+      "step": 1886
+    },
+    {
+      "epoch": 0.08941009239516702,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019611841198639333,
+      "loss": 0.8304,
+      "step": 1887
+    },
+    {
+      "epoch": 0.0894574745321014,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019611430193843793,
+      "loss": 0.8106,
+      "step": 1888
+    },
+    {
+      "epoch": 0.08950485666903578,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019611018975876972,
+      "loss": 0.0576,
+      "step": 1889
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019610607544747987,
+      "loss": 0.5524,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08959962094290452,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019610195900465963,
+      "loss": 1.0035,
+      "step": 1891
+    },
+    {
+      "epoch": 0.0896470030798389,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019609784043040032,
+      "loss": 1.0234,
+      "step": 1892
+    },
+    {
+      "epoch": 0.08969438521677328,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001960937197247932,
+      "loss": 1.0955,
+      "step": 1893
+    },
+    {
+      "epoch": 0.08974176735370765,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001960895968879298,
+      "loss": 0.0322,
+      "step": 1894
+    },
+    {
+      "epoch": 0.08978914949064203,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019608547191990146,
+      "loss": 0.7134,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0898365316275764,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019608134482079973,
+      "loss": 0.8305,
+      "step": 1896
+    },
+    {
+      "epoch": 0.08988391376451078,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019607721559071608,
+      "loss": 1.1003,
+      "step": 1897
+    },
+    {
+      "epoch": 0.08993129590144515,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019607308422974216,
+      "loss": 0.9224,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08997867803837953,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001960689507379695,
+      "loss": 0.2233,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09002606017531391,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001960648151154899,
+      "loss": 0.1031,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09007344231224829,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.000196060677362395,
+      "loss": 0.1767,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09012082444918265,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0001960565374787766,
+      "loss": 0.0024,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09016820658611703,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001960523954647265,
+      "loss": 1.0201,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09021558872305141,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019604825132033663,
+      "loss": 1.044,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09026297085998579,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001960441050456988,
+      "loss": 0.0299,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09031035299692017,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.000196039956640905,
+      "loss": 1.0385,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09035773513385453,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001960358061060473,
+      "loss": 1.0603,
+      "step": 1907
+    },
+    {
+      "epoch": 0.09040511727078891,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019603165344121774,
+      "loss": 0.8072,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09045249940772329,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019602749864650835,
+      "loss": 0.3386,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09049988154465767,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019602334172201137,
+      "loss": 1.4393,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09054726368159204,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001960191826678189,
+      "loss": 1.5312,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09059464581852641,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00019601502148402323,
+      "loss": 0.3824,
+      "step": 1912
+    },
+    {
+      "epoch": 0.0906420279554608,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001960108581707167,
+      "loss": 0.2593,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09068941009239517,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019600669272799157,
+      "loss": 0.102,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09073679222932954,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019600252515594027,
+      "loss": 0.9024,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09078417436626392,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001959983554546552,
+      "loss": 0.6981,
+      "step": 1916
+    },
+    {
+      "epoch": 0.0908315565031983,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001959941836242289,
+      "loss": 1.0576,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09087893864013268,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019599000966475383,
+      "loss": 0.6964,
+      "step": 1918
+    },
+    {
+      "epoch": 0.09092632077706704,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019598583357632258,
+      "loss": 0.0821,
+      "step": 1919
+    },
+    {
+      "epoch": 0.09097370291400142,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019598165535902778,
+      "loss": 0.8816,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0910210850509358,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001959774750129621,
+      "loss": 1.0731,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09106846718787018,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001959732925382183,
+      "loss": 0.7151,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09111584932480454,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019596910793488902,
+      "loss": 1.176,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09116323146173892,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001959649212030672,
+      "loss": 0.3397,
+      "step": 1924
+    },
+    {
+      "epoch": 0.0912106135986733,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019596073234284561,
+      "loss": 1.1506,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09125799573560768,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019595654135431723,
+      "loss": 0.7399,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09130537787254205,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019595234823757495,
+      "loss": 0.1751,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09135276000947642,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001959481529927118,
+      "loss": 1.2673,
+      "step": 1928
+    },
+    {
+      "epoch": 0.0914001421464108,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019594395561982081,
+      "loss": 0.9565,
+      "step": 1929
+    },
+    {
+      "epoch": 0.09144752428334518,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019593975611899506,
+      "loss": 0.4006,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09149490642027955,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019593555449032773,
+      "loss": 0.62,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09154228855721393,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000195931350733912,
+      "loss": 1.1285,
+      "step": 1932
+    },
+    {
+      "epoch": 0.0915896706941483,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019592714484984106,
+      "loss": 0.8639,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09163705283108269,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019592293683820826,
+      "loss": 1.1909,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09168443496801706,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001959187266991069,
+      "loss": 1.2091,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09173181710495143,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001959145144326303,
+      "loss": 1.3906,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09177919924188581,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019591030003887196,
+      "loss": 0.248,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09182658137882019,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019590608351792533,
+      "loss": 1.1041,
+      "step": 1938
+    },
+    {
+      "epoch": 0.09187396351575457,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019590186486988391,
+      "loss": 1.1581,
+      "step": 1939
+    },
+    {
+      "epoch": 0.09192134565268893,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001958976440948413,
+      "loss": 0.9681,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09196872778962331,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019589342119289105,
+      "loss": 0.8862,
+      "step": 1941
+    },
+    {
+      "epoch": 0.09201610992655769,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001958891961641269,
+      "loss": 1.1146,
+      "step": 1942
+    },
+    {
+      "epoch": 0.09206349206349207,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001958849690086425,
+      "loss": 0.5821,
+      "step": 1943
+    },
+    {
+      "epoch": 0.09211087420042643,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001958807397265316,
+      "loss": 0.6119,
+      "step": 1944
+    },
+    {
+      "epoch": 0.09215825633736081,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00019587650831788805,
+      "loss": 0.5591,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09220563847429519,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019587227478280562,
+      "loss": 0.3989,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09225302061122957,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001958680391213783,
+      "loss": 1.3922,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09230040274816394,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019586380133369998,
+      "loss": 0.1463,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09234778488509832,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019585956141986467,
+      "loss": 0.922,
+      "step": 1949
+    },
+    {
+      "epoch": 0.0923951670220327,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019585531937996636,
+      "loss": 0.2105,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09244254915896707,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019585107521409917,
+      "loss": 0.9343,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09248993129590144,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019584682892235725,
+      "loss": 1.093,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09253731343283582,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019584258050483474,
+      "loss": 1.3191,
+      "step": 1953
+    },
+    {
+      "epoch": 0.0925846955697702,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001958383299616259,
+      "loss": 0.0516,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09263207770670458,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019583407729282498,
+      "loss": 0.8796,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09267945984363894,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019582982249852625,
+      "loss": 0.5609,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09272684198057332,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001958255655788242,
+      "loss": 0.6923,
+      "step": 1957
+    },
+    {
+      "epoch": 0.0927742241175077,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019582130653381314,
+      "loss": 0.2134,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09282160625444208,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019581704536358755,
+      "loss": 1.4877,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09286898839137644,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000195812782068242,
+      "loss": 0.8046,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09291637052831082,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019580851664787098,
+      "loss": 1.401,
+      "step": 1961
+    },
+    {
+      "epoch": 0.0929637526652452,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001958042491025691,
+      "loss": 0.017,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09301113480217958,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019579997943243102,
+      "loss": 0.961,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09305851693911395,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019579570763755144,
+      "loss": 0.157,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09310589907604833,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001957914337180251,
+      "loss": 0.1928,
+      "step": 1965
+    },
+    {
+      "epoch": 0.0931532812129827,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019578715767394682,
+      "loss": 0.2222,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09320066334991708,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019578287950541137,
+      "loss": 0.2013,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09324804548685146,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019577859921251372,
+      "loss": 0.1711,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09329542762378583,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019577431679534876,
+      "loss": 1.3195,
+      "step": 1969
+    },
+    {
+      "epoch": 0.09334280976072021,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019577003225401142,
+      "loss": 0.9525,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09339019189765459,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019576574558859678,
+      "loss": 0.8443,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09343757403458897,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019576145679919992,
+      "loss": 0.8663,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09348495617152333,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019575716588591597,
+      "loss": 0.1549,
+      "step": 1973
+    },
+    {
+      "epoch": 0.09353233830845771,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019575287284884005,
+      "loss": 0.211,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09357972044539209,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019574857768806742,
+      "loss": 0.5792,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09362710258232647,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001957442804036933,
+      "loss": 1.4567,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09367448471926083,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00019573998099581304,
+      "loss": 0.012,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09372186685619521,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019573567946452197,
+      "loss": 0.8247,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09376924899312959,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001957313758099155,
+      "loss": 0.0189,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09381663113006397,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019572707003208913,
+      "loss": 1.1493,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09386401326699834,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019572276213113824,
+      "loss": 1.4851,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09391139540393272,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019571845210715849,
+      "loss": 1.2883,
+      "step": 1982
+    },
+    {
+      "epoch": 0.0939587775408671,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019571413996024542,
+      "loss": 1.0645,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09400615967780147,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019570982569049467,
+      "loss": 0.0704,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09405354181473584,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019570550929800192,
+      "loss": 0.9932,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09410092395167022,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019570119078286293,
+      "loss": 0.9952,
+      "step": 1986
+    },
+    {
+      "epoch": 0.0941483060886046,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019569687014517348,
+      "loss": 0.2108,
+      "step": 1987
+    },
+    {
+      "epoch": 0.09419568822553898,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001956925473850294,
+      "loss": 0.8497,
+      "step": 1988
+    },
+    {
+      "epoch": 0.09424307036247334,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019568822250252648,
+      "loss": 1.1613,
+      "step": 1989
+    },
+    {
+      "epoch": 0.09429045249940772,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001956838954977607,
+      "loss": 0.5845,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0943378346363421,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001956795663708281,
+      "loss": 1.0806,
+      "step": 1991
+    },
+    {
+      "epoch": 0.09438521677327648,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001956752351218246,
+      "loss": 0.8906,
+      "step": 1992
+    },
+    {
+      "epoch": 0.09443259891021084,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019567090175084633,
+      "loss": 1.265,
+      "step": 1993
+    },
+    {
+      "epoch": 0.09447998104714522,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001956665662579893,
+      "loss": 0.295,
+      "step": 1994
+    },
+    {
+      "epoch": 0.0945273631840796,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001956622286433498,
+      "loss": 0.0936,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09457474532101398,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001956578889070239,
+      "loss": 0.4778,
+      "step": 1996
+    },
+    {
+      "epoch": 0.09462212745794836,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019565354704910792,
+      "loss": 1.3791,
+      "step": 1997
+    },
+    {
+      "epoch": 0.09466950959488273,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019564920306969818,
+      "loss": 1.0183,
+      "step": 1998
+    },
+    {
+      "epoch": 0.0947168917318171,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019564485696889098,
+      "loss": 1.125,
+      "step": 1999
+    },
+    {
+      "epoch": 0.09476427386875148,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019564050874678276,
+      "loss": 0.881,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09481165600568586,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0001956361584034699,
+      "loss": 0.0042,
+      "step": 2001
+    },
+    {
+      "epoch": 0.09485903814262023,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001956318059390489,
+      "loss": 0.9589,
+      "step": 2002
+    },
+    {
+      "epoch": 0.0949064202795546,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019562745135361635,
+      "loss": 0.8589,
+      "step": 2003
+    },
+    {
+      "epoch": 0.09495380241648899,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019562309464726875,
+      "loss": 0.8546,
+      "step": 2004
+    },
+    {
+      "epoch": 0.09500118455342337,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001956187358201028,
+      "loss": 1.1615,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09504856669035773,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001956143748722151,
+      "loss": 1.1912,
+      "step": 2006
+    },
+    {
+      "epoch": 0.09509594882729211,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001956100118037024,
+      "loss": 1.4969,
+      "step": 2007
+    },
+    {
+      "epoch": 0.09514333096422649,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001956056466146615,
+      "loss": 1.0647,
+      "step": 2008
+    },
+    {
+      "epoch": 0.09519071310116087,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019560127930518922,
+      "loss": 0.7003,
+      "step": 2009
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019559690987538234,
+      "loss": 0.4448,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09528547737502961,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019559253832533786,
+      "loss": 1.2646,
+      "step": 2011
+    },
+    {
+      "epoch": 0.09533285951196399,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001955881646551527,
+      "loss": 1.2954,
+      "step": 2012
+    },
+    {
+      "epoch": 0.09538024164889837,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019558378886492387,
+      "loss": 0.9051,
+      "step": 2013
+    },
+    {
+      "epoch": 0.09542762378583274,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001955794109547484,
+      "loss": 0.9611,
+      "step": 2014
+    },
+    {
+      "epoch": 0.09547500592276711,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001955750309247234,
+      "loss": 1.608,
+      "step": 2015
+    },
+    {
+      "epoch": 0.0955223880597015,
+      "grad_norm": 0.0086669921875,
+      "learning_rate": 0.000195570648774946,
+      "loss": 0.0008,
+      "step": 2016
+    },
+    {
+      "epoch": 0.09556977019663587,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019556626450551343,
+      "loss": 1.1494,
+      "step": 2017
+    },
+    {
+      "epoch": 0.09561715233357024,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001955618781165229,
+      "loss": 1.3059,
+      "step": 2018
+    },
+    {
+      "epoch": 0.09566453447050462,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001955574896080717,
+      "loss": 1.0875,
+      "step": 2019
+    },
+    {
+      "epoch": 0.095711916607439,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019555309898025718,
+      "loss": 0.706,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09575929874437338,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019554870623317668,
+      "loss": 0.9968,
+      "step": 2021
+    },
+    {
+      "epoch": 0.09580668088130774,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019554431136692765,
+      "loss": 0.8222,
+      "step": 2022
+    },
+    {
+      "epoch": 0.09585406301824212,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001955399143816076,
+      "loss": 1.1134,
+      "step": 2023
+    },
+    {
+      "epoch": 0.0959014451551765,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019553551527731397,
+      "loss": 0.9769,
+      "step": 2024
+    },
+    {
+      "epoch": 0.09594882729211088,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001955311140541444,
+      "loss": 0.0852,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09599620942904526,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019552671071219644,
+      "loss": 1.1401,
+      "step": 2026
+    },
+    {
+      "epoch": 0.09604359156597962,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019552230525156784,
+      "loss": 0.9021,
+      "step": 2027
+    },
+    {
+      "epoch": 0.096090973702914,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001955178976723562,
+      "loss": 0.5925,
+      "step": 2028
+    },
+    {
+      "epoch": 0.09613835583984838,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00019551348797465935,
+      "loss": 1.123,
+      "step": 2029
+    },
+    {
+      "epoch": 0.09618573797678276,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019550907615857507,
+      "loss": 0.0715,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09623312011371712,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019550466222420125,
+      "loss": 1.0904,
+      "step": 2031
+    },
+    {
+      "epoch": 0.0962805022506515,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001955002461716357,
+      "loss": 1.1416,
+      "step": 2032
+    },
+    {
+      "epoch": 0.09632788438758588,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019549582800097644,
+      "loss": 1.2338,
+      "step": 2033
+    },
+    {
+      "epoch": 0.09637526652452026,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019549140771232143,
+      "loss": 0.8136,
+      "step": 2034
+    },
+    {
+      "epoch": 0.09642264866145463,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001954869853057687,
+      "loss": 0.857,
+      "step": 2035
+    },
+    {
+      "epoch": 0.096470030798389,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019548256078141636,
+      "loss": 1.0993,
+      "step": 2036
+    },
+    {
+      "epoch": 0.09651741293532339,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001954781341393625,
+      "loss": 0.6322,
+      "step": 2037
+    },
+    {
+      "epoch": 0.09656479507225776,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019547370537970532,
+      "loss": 0.1113,
+      "step": 2038
+    },
+    {
+      "epoch": 0.09661217720919213,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001954692745025431,
+      "loss": 0.9889,
+      "step": 2039
+    },
+    {
+      "epoch": 0.09665955934612651,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019546484150797402,
+      "loss": 1.0365,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09670694148306089,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019546040639609644,
+      "loss": 1.5881,
+      "step": 2041
+    },
+    {
+      "epoch": 0.09675432361999527,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019545596916700872,
+      "loss": 1.3077,
+      "step": 2042
+    },
+    {
+      "epoch": 0.09680170575692963,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019545152982080932,
+      "loss": 1.132,
+      "step": 2043
+    },
+    {
+      "epoch": 0.09684908789386401,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019544708835759662,
+      "loss": 1.2448,
+      "step": 2044
+    },
+    {
+      "epoch": 0.09689647003079839,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001954426447774692,
+      "loss": 1.0699,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09694385216773277,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019543819908052554,
+      "loss": 0.4107,
+      "step": 2046
+    },
+    {
+      "epoch": 0.09699123430466713,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0001954337512668643,
+      "loss": 0.0083,
+      "step": 2047
+    },
+    {
+      "epoch": 0.09703861644160151,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019542930133658408,
+      "loss": 0.2417,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09708599857853589,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019542484928978363,
+      "loss": 1.1712,
+      "step": 2049
+    },
+    {
+      "epoch": 0.09713338071547027,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019542039512656167,
+      "loss": 1.1497,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09718076285240464,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019541593884701697,
+      "loss": 0.7131,
+      "step": 2051
+    },
+    {
+      "epoch": 0.09722814498933902,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001954114804512484,
+      "loss": 0.7727,
+      "step": 2052
+    },
+    {
+      "epoch": 0.0972755271262734,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001954070199393548,
+      "loss": 0.9297,
+      "step": 2053
+    },
+    {
+      "epoch": 0.09732290926320777,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019540255731143513,
+      "loss": 0.1139,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09737029140014215,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00019539809256758836,
+      "loss": 0.0037,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09741767353707652,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019539362570791352,
+      "loss": 0.208,
+      "step": 2056
+    },
+    {
+      "epoch": 0.0974650556740109,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019538915673250964,
+      "loss": 1.1762,
+      "step": 2057
+    },
+    {
+      "epoch": 0.09751243781094528,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019538468564147588,
+      "loss": 0.9383,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09755981994787966,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001953802124349114,
+      "loss": 1.2892,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09760720208481402,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001953757371129154,
+      "loss": 0.3101,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0976545842217484,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019537125967558714,
+      "loss": 0.0716,
+      "step": 2061
+    },
+    {
+      "epoch": 0.09770196635868278,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019536678012302592,
+      "loss": 1.0844,
+      "step": 2062
+    },
+    {
+      "epoch": 0.09774934849561716,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001953622984553311,
+      "loss": 0.9733,
+      "step": 2063
+    },
+    {
+      "epoch": 0.09779673063255152,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019535781467260206,
+      "loss": 0.0277,
+      "step": 2064
+    },
+    {
+      "epoch": 0.0978441127694859,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001953533287749383,
+      "loss": 0.0644,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09789149490642028,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019534884076243922,
+      "loss": 0.8489,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09793887704335466,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019534435063520446,
+      "loss": 0.8202,
+      "step": 2067
+    },
+    {
+      "epoch": 0.09798625918028903,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019533985839333354,
+      "loss": 0.9201,
+      "step": 2068
+    },
+    {
+      "epoch": 0.0980336413172234,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001953353640369261,
+      "loss": 0.0525,
+      "step": 2069
+    },
+    {
+      "epoch": 0.09808102345415778,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019533086756608187,
+      "loss": 0.9008,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09812840559109216,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001953263689809005,
+      "loss": 1.6623,
+      "step": 2071
+    },
+    {
+      "epoch": 0.09817578772802653,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019532186828148183,
+      "loss": 1.2798,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09822316986496091,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019531736546792562,
+      "loss": 0.8696,
+      "step": 2073
+    },
+    {
+      "epoch": 0.09827055200189529,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001953128605403318,
+      "loss": 1.2022,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09831793413882967,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019530835349880027,
+      "loss": 0.8575,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09836531627576403,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019530384434343097,
+      "loss": 1.1681,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09841269841269841,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001952993330743239,
+      "loss": 1.0974,
+      "step": 2077
+    },
+    {
+      "epoch": 0.09846008054963279,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019529481969157912,
+      "loss": 0.135,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09850746268656717,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019529030419529675,
+      "loss": 1.3588,
+      "step": 2079
+    },
+    {
+      "epoch": 0.09855484482350153,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019528578658557696,
+      "loss": 1.0843,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09860222696043591,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019528126686251988,
+      "loss": 0.9951,
+      "step": 2081
+    },
+    {
+      "epoch": 0.09864960909737029,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001952767450262258,
+      "loss": 0.3888,
+      "step": 2082
+    },
+    {
+      "epoch": 0.09869699123430467,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019527222107679502,
+      "loss": 1.2278,
+      "step": 2083
+    },
+    {
+      "epoch": 0.09874437337123905,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019526769501432785,
+      "loss": 0.0112,
+      "step": 2084
+    },
+    {
+      "epoch": 0.09879175550817342,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019526316683892464,
+      "loss": 0.9511,
+      "step": 2085
+    },
+    {
+      "epoch": 0.0988391376451078,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019525863655068593,
+      "loss": 1.0531,
+      "step": 2086
+    },
+    {
+      "epoch": 0.09888651978204217,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019525410414971206,
+      "loss": 1.47,
+      "step": 2087
+    },
+    {
+      "epoch": 0.09893390191897655,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019524956963610365,
+      "loss": 0.2093,
+      "step": 2088
+    },
+    {
+      "epoch": 0.09898128405591092,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019524503300996125,
+      "loss": 0.4769,
+      "step": 2089
+    },
+    {
+      "epoch": 0.0990286661928453,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019524049427138544,
+      "loss": 1.1945,
+      "step": 2090
+    },
+    {
+      "epoch": 0.09907604832977968,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019523595342047694,
+      "loss": 0.7347,
+      "step": 2091
+    },
+    {
+      "epoch": 0.09912343046671405,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019523141045733642,
+      "loss": 0.0159,
+      "step": 2092
+    },
+    {
+      "epoch": 0.09917081260364842,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019522686538206465,
+      "loss": 1.2201,
+      "step": 2093
+    },
+    {
+      "epoch": 0.0992181947405828,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019522231819476244,
+      "loss": 0.8225,
+      "step": 2094
+    },
+    {
+      "epoch": 0.09926557687751718,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019521776889553066,
+      "loss": 0.9317,
+      "step": 2095
+    },
+    {
+      "epoch": 0.09931295901445156,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00019521321748447015,
+      "loss": 0.1652,
+      "step": 2096
+    },
+    {
+      "epoch": 0.09936034115138592,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001952086639616819,
+      "loss": 0.2254,
+      "step": 2097
+    },
+    {
+      "epoch": 0.0994077232883203,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00019520410832726693,
+      "loss": 0.0149,
+      "step": 2098
+    },
+    {
+      "epoch": 0.09945510542525468,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001951995505813262,
+      "loss": 1.0392,
+      "step": 2099
+    },
+    {
+      "epoch": 0.09950248756218906,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019519499072396087,
+      "loss": 0.2113,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09954986969912342,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019519042875527202,
+      "loss": 0.7009,
+      "step": 2101
+    },
+    {
+      "epoch": 0.0995972518360578,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019518586467536086,
+      "loss": 1.1118,
+      "step": 2102
+    },
+    {
+      "epoch": 0.09964463397299218,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001951812984843286,
+      "loss": 0.8469,
+      "step": 2103
+    },
+    {
+      "epoch": 0.09969201610992656,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00019517673018227654,
+      "loss": 0.0265,
+      "step": 2104
+    },
+    {
+      "epoch": 0.09973939824686093,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000195172159769306,
+      "loss": 0.7084,
+      "step": 2105
+    },
+    {
+      "epoch": 0.0997867803837953,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001951675872455183,
+      "loss": 1.1953,
+      "step": 2106
+    },
+    {
+      "epoch": 0.09983416252072969,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001951630126110149,
+      "loss": 1.233,
+      "step": 2107
+    },
+    {
+      "epoch": 0.09988154465766406,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001951584358658972,
+      "loss": 1.1104,
+      "step": 2108
+    },
+    {
+      "epoch": 0.09992892679459843,
+      "grad_norm": 0.00634765625,
+      "learning_rate": 0.0001951538570102668,
+      "loss": 0.0004,
+      "step": 2109
+    },
+    {
+      "epoch": 0.09997630893153281,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001951492760442252,
+      "loss": 0.917,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10002369106846719,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.000195144692967874,
+      "loss": 0.4697,
+      "step": 2111
+    },
+    {
+      "epoch": 0.10007107320540157,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019514010778131483,
+      "loss": 1.0373,
+      "step": 2112
+    },
+    {
+      "epoch": 0.10011845534233593,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019513552048464942,
+      "loss": 0.9711,
+      "step": 2113
+    },
+    {
+      "epoch": 0.10016583747927031,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001951309310779795,
+      "loss": 1.4549,
+      "step": 2114
+    },
+    {
+      "epoch": 0.10021321961620469,
+      "grad_norm": 0.0244140625,
+      "learning_rate": 0.00019512633956140688,
+      "loss": 0.002,
+      "step": 2115
+    },
+    {
+      "epoch": 0.10026060175313907,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019512174593503336,
+      "loss": 0.0166,
+      "step": 2116
+    },
+    {
+      "epoch": 0.10030798389007345,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019511715019896082,
+      "loss": 1.2642,
+      "step": 2117
+    },
+    {
+      "epoch": 0.10035536602700781,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001951125523532912,
+      "loss": 1.1373,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1004027481639422,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001951079523981265,
+      "loss": 0.2338,
+      "step": 2119
+    },
+    {
+      "epoch": 0.10045013030087657,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019510335033356873,
+      "loss": 0.5619,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10049751243781095,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019509874615971992,
+      "loss": 0.018,
+      "step": 2121
+    },
+    {
+      "epoch": 0.10054489457474532,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019509413987668222,
+      "loss": 0.939,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1005922767116797,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001950895314845578,
+      "loss": 1.0705,
+      "step": 2123
+    },
+    {
+      "epoch": 0.10063965884861407,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019508492098344884,
+      "loss": 1.0411,
+      "step": 2124
+    },
+    {
+      "epoch": 0.10068704098554845,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019508030837345764,
+      "loss": 1.0597,
+      "step": 2125
+    },
+    {
+      "epoch": 0.10073442312248282,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019507569365468644,
+      "loss": 0.0604,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1007818052594172,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019507107682723764,
+      "loss": 1.1313,
+      "step": 2127
+    },
+    {
+      "epoch": 0.10082918739635158,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00019506645789121364,
+      "loss": 0.0049,
+      "step": 2128
+    },
+    {
+      "epoch": 0.10087656953328596,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019506183684671685,
+      "loss": 0.6684,
+      "step": 2129
+    },
+    {
+      "epoch": 0.10092395167022032,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019505721369384975,
+      "loss": 0.1914,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1009713338071547,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001950525884327149,
+      "loss": 0.2069,
+      "step": 2131
+    },
+    {
+      "epoch": 0.10101871594408908,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019504796106341492,
+      "loss": 0.9878,
+      "step": 2132
+    },
+    {
+      "epoch": 0.10106609808102346,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001950433315860524,
+      "loss": 1.1931,
+      "step": 2133
+    },
+    {
+      "epoch": 0.10111348021795782,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019503870000073001,
+      "loss": 1.2078,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1011608623548922,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019503406630755048,
+      "loss": 0.0691,
+      "step": 2135
+    },
+    {
+      "epoch": 0.10120824449182658,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001950294305066166,
+      "loss": 1.4365,
+      "step": 2136
+    },
+    {
+      "epoch": 0.10125562662876096,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019502479259803117,
+      "loss": 1.1509,
+      "step": 2137
+    },
+    {
+      "epoch": 0.10130300876569533,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019502015258189703,
+      "loss": 0.3573,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1013503909026297,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019501551045831715,
+      "loss": 1.3825,
+      "step": 2139
+    },
+    {
+      "epoch": 0.10139777303956408,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019501086622739444,
+      "loss": 0.8767,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10144515517649846,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019500621988923191,
+      "loss": 0.9961,
+      "step": 2141
+    },
+    {
+      "epoch": 0.10149253731343283,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001950015714439326,
+      "loss": 0.7829,
+      "step": 2142
+    },
+    {
+      "epoch": 0.10153991945036721,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001949969208915997,
+      "loss": 0.8525,
+      "step": 2143
+    },
+    {
+      "epoch": 0.10158730158730159,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001949922682323362,
+      "loss": 1.2719,
+      "step": 2144
+    },
+    {
+      "epoch": 0.10163468372423597,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019498761346624542,
+      "loss": 0.6178,
+      "step": 2145
+    },
+    {
+      "epoch": 0.10168206586117035,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001949829565934305,
+      "loss": 0.7717,
+      "step": 2146
+    },
+    {
+      "epoch": 0.10172944799810471,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019497829761399484,
+      "loss": 0.7072,
+      "step": 2147
+    },
+    {
+      "epoch": 0.10177683013503909,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019497363652804168,
+      "loss": 0.8323,
+      "step": 2148
+    },
+    {
+      "epoch": 0.10182421227197347,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001949689733356744,
+      "loss": 0.1043,
+      "step": 2149
+    },
+    {
+      "epoch": 0.10187159440890785,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019496430803699646,
+      "loss": 1.0638,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10191897654584221,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019495964063211135,
+      "loss": 0.8882,
+      "step": 2151
+    },
+    {
+      "epoch": 0.10196635868277659,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019495497112112254,
+      "loss": 0.146,
+      "step": 2152
+    },
+    {
+      "epoch": 0.10201374081971097,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001949502995041336,
+      "loss": 0.4443,
+      "step": 2153
+    },
+    {
+      "epoch": 0.10206112295664535,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019494562578124813,
+      "loss": 0.9381,
+      "step": 2154
+    },
+    {
+      "epoch": 0.10210850509357972,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019494094995256984,
+      "loss": 1.013,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1021558872305141,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019493627201820242,
+      "loss": 0.9014,
+      "step": 2156
+    },
+    {
+      "epoch": 0.10220326936744847,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001949315919782496,
+      "loss": 0.9222,
+      "step": 2157
+    },
+    {
+      "epoch": 0.10225065150438285,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019492690983281515,
+      "loss": 0.0349,
+      "step": 2158
+    },
+    {
+      "epoch": 0.10229803364131722,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000194922225582003,
+      "loss": 0.1476,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1023454157782516,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019491753922591695,
+      "loss": 0.1474,
+      "step": 2160
+    },
+    {
+      "epoch": 0.10239279791518598,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019491285076466102,
+      "loss": 0.107,
+      "step": 2161
+    },
+    {
+      "epoch": 0.10244018005212036,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00019490816019833914,
+      "loss": 0.0058,
+      "step": 2162
+    },
+    {
+      "epoch": 0.10248756218905472,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019490346752705536,
+      "loss": 0.2579,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1025349443259891,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019489877275091378,
+      "loss": 1.0398,
+      "step": 2164
+    },
+    {
+      "epoch": 0.10258232646292348,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001948940758700185,
+      "loss": 1.6905,
+      "step": 2165
+    },
+    {
+      "epoch": 0.10262970859985786,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00019488937688447365,
+      "loss": 0.2404,
+      "step": 2166
+    },
+    {
+      "epoch": 0.10267709073679222,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001948846757943835,
+      "loss": 0.8703,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1027244728737266,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019487997259985234,
+      "loss": 1.3456,
+      "step": 2168
+    },
+    {
+      "epoch": 0.10277185501066098,
+      "grad_norm": 0.00750732421875,
+      "learning_rate": 0.00019487526730098441,
+      "loss": 0.0007,
+      "step": 2169
+    },
+    {
+      "epoch": 0.10281923714759536,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019487055989788417,
+      "loss": 1.2684,
+      "step": 2170
+    },
+    {
+      "epoch": 0.10286661928452973,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001948658503906559,
+      "loss": 0.1757,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1029140014214641,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019486113877940415,
+      "loss": 0.1336,
+      "step": 2172
+    },
+    {
+      "epoch": 0.10296138355839848,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019485642506423338,
+      "loss": 0.6852,
+      "step": 2173
+    },
+    {
+      "epoch": 0.10300876569533286,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019485170924524813,
+      "loss": 1.1825,
+      "step": 2174
+    },
+    {
+      "epoch": 0.10305614783226724,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000194846991322553,
+      "loss": 0.2251,
+      "step": 2175
+    },
+    {
+      "epoch": 0.10310352996920161,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019484227129625264,
+      "loss": 0.3324,
+      "step": 2176
+    },
+    {
+      "epoch": 0.10315091210613599,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019483754916645168,
+      "loss": 0.0443,
+      "step": 2177
+    },
+    {
+      "epoch": 0.10319829424307037,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019483282493325493,
+      "loss": 0.9885,
+      "step": 2178
+    },
+    {
+      "epoch": 0.10324567638000474,
+      "grad_norm": 0.03125,
+      "learning_rate": 0.00019482809859676717,
+      "loss": 0.0012,
+      "step": 2179
+    },
+    {
+      "epoch": 0.10329305851693911,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019482337015709313,
+      "loss": 0.7013,
+      "step": 2180
+    },
+    {
+      "epoch": 0.10334044065387349,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019481863961433776,
+      "loss": 1.396,
+      "step": 2181
+    },
+    {
+      "epoch": 0.10338782279080787,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019481390696860596,
+      "loss": 0.5346,
+      "step": 2182
+    },
+    {
+      "epoch": 0.10343520492774225,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019480917222000272,
+      "loss": 1.4599,
+      "step": 2183
+    },
+    {
+      "epoch": 0.10348258706467661,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.000194804435368633,
+      "loss": 0.0461,
+      "step": 2184
+    },
+    {
+      "epoch": 0.10352996920161099,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001947996964146019,
+      "loss": 1.3128,
+      "step": 2185
+    },
+    {
+      "epoch": 0.10357735133854537,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001947949553580145,
+      "loss": 1.5236,
+      "step": 2186
+    },
+    {
+      "epoch": 0.10362473347547975,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019479021219897594,
+      "loss": 0.6163,
+      "step": 2187
+    },
+    {
+      "epoch": 0.10367211561241411,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019478546693759148,
+      "loss": 0.9992,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1037194977493485,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019478071957396628,
+      "loss": 1.2316,
+      "step": 2189
+    },
+    {
+      "epoch": 0.10376687988628287,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001947759701082057,
+      "loss": 0.0197,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10381426202321725,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001947712185404151,
+      "loss": 0.1965,
+      "step": 2191
+    },
+    {
+      "epoch": 0.10386164416015162,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019476646487069977,
+      "loss": 0.1305,
+      "step": 2192
+    },
+    {
+      "epoch": 0.103909026297086,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019476170909916518,
+      "loss": 0.6922,
+      "step": 2193
+    },
+    {
+      "epoch": 0.10395640843402038,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001947569512259168,
+      "loss": 1.1491,
+      "step": 2194
+    },
+    {
+      "epoch": 0.10400379057095475,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019475219125106023,
+      "loss": 1.0979,
+      "step": 2195
+    },
+    {
+      "epoch": 0.10405117270788912,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019474742917470096,
+      "loss": 0.2263,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1040985548448235,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019474266499694464,
+      "loss": 0.2181,
+      "step": 2197
+    },
+    {
+      "epoch": 0.10414593698175788,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001947378987178969,
+      "loss": 0.9503,
+      "step": 2198
+    },
+    {
+      "epoch": 0.10419331911869226,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001947331303376635,
+      "loss": 1.2979,
+      "step": 2199
+    },
+    {
+      "epoch": 0.10424070125562662,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019472835985635017,
+      "loss": 1.5636,
+      "step": 2200
+    },
+    {
+      "epoch": 0.104288083392561,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001947235872740627,
+      "loss": 0.915,
+      "step": 2201
+    },
+    {
+      "epoch": 0.10433546552949538,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00019471881259090697,
+      "loss": 0.8837,
+      "step": 2202
+    },
+    {
+      "epoch": 0.10438284766642976,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019471403580698885,
+      "loss": 1.2649,
+      "step": 2203
+    },
+    {
+      "epoch": 0.10443022980336414,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00019470925692241436,
+      "loss": 0.0166,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1044776119402985,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001947044759372894,
+      "loss": 0.2473,
+      "step": 2205
+    },
+    {
+      "epoch": 0.10452499407723288,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019469969285172,
+      "loss": 1.3336,
+      "step": 2206
+    },
+    {
+      "epoch": 0.10457237621416726,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001946949076658123,
+      "loss": 0.089,
+      "step": 2207
+    },
+    {
+      "epoch": 0.10461975835110164,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019469012037967245,
+      "loss": 0.2336,
+      "step": 2208
+    },
+    {
+      "epoch": 0.104667140488036,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019468533099340656,
+      "loss": 0.5127,
+      "step": 2209
+    },
+    {
+      "epoch": 0.10471452262497039,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019468053950712086,
+      "loss": 1.1701,
+      "step": 2210
+    },
+    {
+      "epoch": 0.10476190476190476,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019467574592092168,
+      "loss": 0.7627,
+      "step": 2211
+    },
+    {
+      "epoch": 0.10480928689883914,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019467095023491528,
+      "loss": 1.283,
+      "step": 2212
+    },
+    {
+      "epoch": 0.10485666903577351,
+      "grad_norm": 0.02294921875,
+      "learning_rate": 0.00019466615244920807,
+      "loss": 0.0013,
+      "step": 2213
+    },
+    {
+      "epoch": 0.10490405117270789,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001946613525639064,
+      "loss": 0.0062,
+      "step": 2214
+    },
+    {
+      "epoch": 0.10495143330964227,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019465655057911678,
+      "loss": 0.9722,
+      "step": 2215
+    },
+    {
+      "epoch": 0.10499881544657665,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001946517464949457,
+      "loss": 1.1721,
+      "step": 2216
+    },
+    {
+      "epoch": 0.10504619758351101,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019464694031149968,
+      "loss": 0.5812,
+      "step": 2217
+    },
+    {
+      "epoch": 0.10509357972044539,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019464213202888535,
+      "loss": 0.8201,
+      "step": 2218
+    },
+    {
+      "epoch": 0.10514096185737977,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001946373216472093,
+      "loss": 1.0211,
+      "step": 2219
+    },
+    {
+      "epoch": 0.10518834399431415,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019463250916657834,
+      "loss": 1.2849,
+      "step": 2220
+    },
+    {
+      "epoch": 0.10523572613124851,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00019462769458709905,
+      "loss": 0.2717,
+      "step": 2221
+    },
+    {
+      "epoch": 0.10528310826818289,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019462287790887833,
+      "loss": 1.056,
+      "step": 2222
+    },
+    {
+      "epoch": 0.10533049040511727,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019461805913202293,
+      "loss": 0.1818,
+      "step": 2223
+    },
+    {
+      "epoch": 0.10537787254205165,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001946132382566398,
+      "loss": 1.3601,
+      "step": 2224
+    },
+    {
+      "epoch": 0.10542525467898602,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001946084152828358,
+      "loss": 1.411,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1054726368159204,
+      "grad_norm": 0.00494384765625,
+      "learning_rate": 0.00019460359021071793,
+      "loss": 0.0003,
+      "step": 2226
+    },
+    {
+      "epoch": 0.10552001895285477,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001945987630403932,
+      "loss": 0.1902,
+      "step": 2227
+    },
+    {
+      "epoch": 0.10556740108978915,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019459393377196864,
+      "loss": 1.4867,
+      "step": 2228
+    },
+    {
+      "epoch": 0.10561478322672352,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001945891024055514,
+      "loss": 0.1368,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1056621653636579,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019458426894124864,
+      "loss": 0.4235,
+      "step": 2230
+    },
+    {
+      "epoch": 0.10570954750059228,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001945794333791675,
+      "loss": 1.5729,
+      "step": 2231
+    },
+    {
+      "epoch": 0.10575692963752666,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019457459571941532,
+      "loss": 0.5732,
+      "step": 2232
+    },
+    {
+      "epoch": 0.10580431177446104,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019456975596209933,
+      "loss": 1.2299,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1058516939113954,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019456491410732688,
+      "loss": 0.8259,
+      "step": 2234
+    },
+    {
+      "epoch": 0.10589907604832978,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019456007015520533,
+      "loss": 0.0567,
+      "step": 2235
+    },
+    {
+      "epoch": 0.10594645818526416,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019455522410584218,
+      "loss": 0.9271,
+      "step": 2236
+    },
+    {
+      "epoch": 0.10599384032219854,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019455037595934486,
+      "loss": 1.4491,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1060412224591329,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001945455257158209,
+      "loss": 0.2037,
+      "step": 2238
+    },
+    {
+      "epoch": 0.10608860459606728,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001945406733753779,
+      "loss": 1.2239,
+      "step": 2239
+    },
+    {
+      "epoch": 0.10613598673300166,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001945358189381235,
+      "loss": 0.7899,
+      "step": 2240
+    },
+    {
+      "epoch": 0.10618336886993604,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019453096240416528,
+      "loss": 0.9769,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1062307510068704,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019452610377361103,
+      "loss": 1.6305,
+      "step": 2242
+    },
+    {
+      "epoch": 0.10627813314380478,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019452124304656846,
+      "loss": 0.6719,
+      "step": 2243
+    },
+    {
+      "epoch": 0.10632551528073916,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019451638022314541,
+      "loss": 0.8991,
+      "step": 2244
+    },
+    {
+      "epoch": 0.10637289741767354,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019451151530344973,
+      "loss": 1.1208,
+      "step": 2245
+    },
+    {
+      "epoch": 0.10642027955460791,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001945066482875893,
+      "loss": 0.0206,
+      "step": 2246
+    },
+    {
+      "epoch": 0.10646766169154229,
+      "grad_norm": 0.0026397705078125,
+      "learning_rate": 0.00019450177917567206,
+      "loss": 0.0002,
+      "step": 2247
+    },
+    {
+      "epoch": 0.10651504382847667,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019449690796780606,
+      "loss": 0.8428,
+      "step": 2248
+    },
+    {
+      "epoch": 0.10656242596541105,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019449203466409928,
+      "loss": 1.1472,
+      "step": 2249
+    },
+    {
+      "epoch": 0.10660980810234541,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019448715926465978,
+      "loss": 0.9232,
+      "step": 2250
+    },
+    {
+      "epoch": 0.10665719023927979,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019448228176959577,
+      "loss": 0.5728,
+      "step": 2251
+    },
+    {
+      "epoch": 0.10670457237621417,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001944774021790154,
+      "loss": 0.9843,
+      "step": 2252
+    },
+    {
+      "epoch": 0.10675195451314855,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019447252049302686,
+      "loss": 0.2289,
+      "step": 2253
+    },
+    {
+      "epoch": 0.10679933665008291,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019446763671173843,
+      "loss": 1.2277,
+      "step": 2254
+    },
+    {
+      "epoch": 0.10684671878701729,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019446275083525848,
+      "loss": 0.5088,
+      "step": 2255
+    },
+    {
+      "epoch": 0.10689410092395167,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019445786286369527,
+      "loss": 0.1824,
+      "step": 2256
+    },
+    {
+      "epoch": 0.10694148306088605,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001944529727971573,
+      "loss": 0.7188,
+      "step": 2257
+    },
+    {
+      "epoch": 0.10698886519782042,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019444808063575302,
+      "loss": 1.2987,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1070362473347548,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019444318637959091,
+      "loss": 0.9129,
+      "step": 2259
+    },
+    {
+      "epoch": 0.10708362947168917,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019443829002877951,
+      "loss": 0.5481,
+      "step": 2260
+    },
+    {
+      "epoch": 0.10713101160862355,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019443339158342745,
+      "loss": 0.6164,
+      "step": 2261
+    },
+    {
+      "epoch": 0.10717839374555792,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019442849104364334,
+      "loss": 0.0807,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1072257758824923,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019442358840953588,
+      "loss": 1.028,
+      "step": 2263
+    },
+    {
+      "epoch": 0.10727315801942668,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001944186836812138,
+      "loss": 1.1778,
+      "step": 2264
+    },
+    {
+      "epoch": 0.10732054015636106,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019441377685878587,
+      "loss": 1.3525,
+      "step": 2265
+    },
+    {
+      "epoch": 0.10736792229329543,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019440886794236097,
+      "loss": 0.7213,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1074153044302298,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001944039569320479,
+      "loss": 0.8769,
+      "step": 2267
+    },
+    {
+      "epoch": 0.10746268656716418,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019439904382795564,
+      "loss": 0.8678,
+      "step": 2268
+    },
+    {
+      "epoch": 0.10751006870409856,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019439412863019314,
+      "loss": 0.1564,
+      "step": 2269
+    },
+    {
+      "epoch": 0.10755745084103294,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001943892113388694,
+      "loss": 0.0198,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1076048329779673,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019438429195409352,
+      "loss": 0.0609,
+      "step": 2271
+    },
+    {
+      "epoch": 0.10765221511490168,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019437937047597455,
+      "loss": 0.508,
+      "step": 2272
+    },
+    {
+      "epoch": 0.10769959725183606,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001943744469046217,
+      "loss": 0.3306,
+      "step": 2273
+    },
+    {
+      "epoch": 0.10774697938877044,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001943695212401441,
+      "loss": 0.0205,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1077943615257048,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019436459348265106,
+      "loss": 0.569,
+      "step": 2275
+    },
+    {
+      "epoch": 0.10784174366263918,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001943596636322518,
+      "loss": 0.0908,
+      "step": 2276
+    },
+    {
+      "epoch": 0.10788912579957356,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019435473168905577,
+      "loss": 1.1299,
+      "step": 2277
+    },
+    {
+      "epoch": 0.10793650793650794,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001943497976531723,
+      "loss": 0.0519,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1079838900734423,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019434486152471075,
+      "loss": 0.9495,
+      "step": 2279
+    },
+    {
+      "epoch": 0.10803127221037669,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019433992330378073,
+      "loss": 0.0155,
+      "step": 2280
+    },
+    {
+      "epoch": 0.10807865434731107,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019433498299049168,
+      "loss": 0.94,
+      "step": 2281
+    },
+    {
+      "epoch": 0.10812603648424544,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019433004058495317,
+      "loss": 0.8623,
+      "step": 2282
+    },
+    {
+      "epoch": 0.10817341862117981,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019432509608727485,
+      "loss": 1.4621,
+      "step": 2283
+    },
+    {
+      "epoch": 0.10822080075811419,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00019432014949756637,
+      "loss": 0.0021,
+      "step": 2284
+    },
+    {
+      "epoch": 0.10826818289504857,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019431520081593742,
+      "loss": 1.0244,
+      "step": 2285
+    },
+    {
+      "epoch": 0.10831556503198295,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001943102500424978,
+      "loss": 1.214,
+      "step": 2286
+    },
+    {
+      "epoch": 0.10836294716891731,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019430529717735727,
+      "loss": 1.6764,
+      "step": 2287
+    },
+    {
+      "epoch": 0.10841032930585169,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019430034222062573,
+      "loss": 0.8461,
+      "step": 2288
+    },
+    {
+      "epoch": 0.10845771144278607,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019429538517241302,
+      "loss": 1.0771,
+      "step": 2289
+    },
+    {
+      "epoch": 0.10850509357972045,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001942904260328291,
+      "loss": 0.1528,
+      "step": 2290
+    },
+    {
+      "epoch": 0.10855247571665481,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019428546480198397,
+      "loss": 1.1033,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1085998578535892,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019428050147998765,
+      "loss": 1.2995,
+      "step": 2292
+    },
+    {
+      "epoch": 0.10864723999052357,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019427553606695024,
+      "loss": 0.656,
+      "step": 2293
+    },
+    {
+      "epoch": 0.10869462212745795,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019427056856298185,
+      "loss": 0.9073,
+      "step": 2294
+    },
+    {
+      "epoch": 0.10874200426439233,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001942655989681927,
+      "loss": 0.1786,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1087893864013267,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001942606272826929,
+      "loss": 0.4551,
+      "step": 2296
+    },
+    {
+      "epoch": 0.10883676853826108,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019425565350659286,
+      "loss": 0.9024,
+      "step": 2297
+    },
+    {
+      "epoch": 0.10888415067519545,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019425067764000276,
+      "loss": 0.028,
+      "step": 2298
+    },
+    {
+      "epoch": 0.10893153281212983,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001942456996830331,
+      "loss": 0.9282,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1089789149490642,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019424071963579414,
+      "loss": 1.1809,
+      "step": 2300
+    },
+    {
+      "epoch": 0.10902629708599858,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019423573749839643,
+      "loss": 0.8025,
+      "step": 2301
+    },
+    {
+      "epoch": 0.10907367922293296,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001942307532709504,
+      "loss": 1.4061,
+      "step": 2302
+    },
+    {
+      "epoch": 0.10912106135986734,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019422576695356667,
+      "loss": 1.2611,
+      "step": 2303
+    },
+    {
+      "epoch": 0.1091684434968017,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001942207785463558,
+      "loss": 1.0934,
+      "step": 2304
+    },
+    {
+      "epoch": 0.10921582563373608,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019421578804942842,
+      "loss": 1.3257,
+      "step": 2305
+    },
+    {
+      "epoch": 0.10926320777067046,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019421079546289518,
+      "loss": 0.4124,
+      "step": 2306
+    },
+    {
+      "epoch": 0.10931058990760484,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019420580078686689,
+      "loss": 1.3487,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1093579720445392,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019420080402145424,
+      "loss": 0.8191,
+      "step": 2308
+    },
+    {
+      "epoch": 0.10940535418147358,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001941958051667681,
+      "loss": 1.0877,
+      "step": 2309
+    },
+    {
+      "epoch": 0.10945273631840796,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019419080422291936,
+      "loss": 1.1972,
+      "step": 2310
+    },
+    {
+      "epoch": 0.10950011845534234,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019418580119001888,
+      "loss": 0.0473,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1095475005922767,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019418079606817767,
+      "loss": 0.0514,
+      "step": 2312
+    },
+    {
+      "epoch": 0.10959488272921108,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019417578885750673,
+      "loss": 0.1785,
+      "step": 2313
+    },
+    {
+      "epoch": 0.10964226486614546,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019417077955811708,
+      "loss": 1.6825,
+      "step": 2314
+    },
+    {
+      "epoch": 0.10968964700307984,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019416576817011988,
+      "loss": 1.0818,
+      "step": 2315
+    },
+    {
+      "epoch": 0.10973702914001421,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001941607546936262,
+      "loss": 1.2398,
+      "step": 2316
+    },
+    {
+      "epoch": 0.10978441127694859,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019415573912874733,
+      "loss": 0.0085,
+      "step": 2317
+    },
+    {
+      "epoch": 0.10983179341388297,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001941507214755944,
+      "loss": 0.0102,
+      "step": 2318
+    },
+    {
+      "epoch": 0.10987917555081735,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001941457017342788,
+      "loss": 0.8173,
+      "step": 2319
+    },
+    {
+      "epoch": 0.10992655768775171,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019414067990491178,
+      "loss": 0.8912,
+      "step": 2320
+    },
+    {
+      "epoch": 0.10997393982468609,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019413565598760477,
+      "loss": 0.9078,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11002132196162047,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019413062998246917,
+      "loss": 0.1027,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11006870409855485,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019412560188961648,
+      "loss": 0.0374,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11011608623548923,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001941205717091582,
+      "loss": 0.0314,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11016346837242359,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019411553944120584,
+      "loss": 1.018,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11021085050935797,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001941105050858711,
+      "loss": 1.2728,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11025823264629235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001941054686432656,
+      "loss": 0.3018,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11030561478322673,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019410043011350102,
+      "loss": 1.0628,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1103529969201611,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019409538949668916,
+      "loss": 0.2092,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11040037905709547,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001940903467929418,
+      "loss": 0.3697,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11044776119402985,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019408530200237074,
+      "loss": 1.355,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11049514333096423,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001940802551250879,
+      "loss": 1.1421,
+      "step": 2332
+    },
+    {
+      "epoch": 0.1105425254678986,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019407520616120523,
+      "loss": 0.7896,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11058990760483298,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019407015511083465,
+      "loss": 1.1293,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11063728974176736,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019406510197408826,
+      "loss": 0.8314,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11068467187870173,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019406004675107813,
+      "loss": 1.0182,
+      "step": 2336
+    },
+    {
+      "epoch": 0.1107320540156361,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001940549894419163,
+      "loss": 0.9839,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11077943615257048,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000194049930046715,
+      "loss": 1.1501,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11082681828950486,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019404486856558644,
+      "loss": 0.4599,
+      "step": 2339
+    },
+    {
+      "epoch": 0.11087420042643924,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019403980499864285,
+      "loss": 1.2662,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1109215825633736,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019403473934599655,
+      "loss": 0.0025,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11096896470030798,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001940296716077599,
+      "loss": 0.6028,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11101634683724236,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019402460178404534,
+      "loss": 0.8875,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11106372897417674,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001940195298749652,
+      "loss": 1.0461,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00019401445588063205,
+      "loss": 0.0401,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11115849324804548,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001940093798011584,
+      "loss": 0.9364,
+      "step": 2346
+    },
+    {
+      "epoch": 0.11120587538497986,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019400430163665685,
+      "loss": 0.7033,
+      "step": 2347
+    },
+    {
+      "epoch": 0.11125325752191424,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019399922138724004,
+      "loss": 1.3435,
+      "step": 2348
+    },
+    {
+      "epoch": 0.11130063965884861,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001939941390530206,
+      "loss": 0.2642,
+      "step": 2349
+    },
+    {
+      "epoch": 0.11134802179578299,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019398905463411124,
+      "loss": 0.9699,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11139540393271737,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019398396813062482,
+      "loss": 1.213,
+      "step": 2351
+    },
+    {
+      "epoch": 0.11144278606965174,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019397887954267408,
+      "loss": 0.1429,
+      "step": 2352
+    },
+    {
+      "epoch": 0.11149016820658612,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019397378887037187,
+      "loss": 1.3877,
+      "step": 2353
+    },
+    {
+      "epoch": 0.11153755034352049,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019396869611383114,
+      "loss": 0.1812,
+      "step": 2354
+    },
+    {
+      "epoch": 0.11158493248045487,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019396360127316482,
+      "loss": 0.0505,
+      "step": 2355
+    },
+    {
+      "epoch": 0.11163231461738925,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0001939585043484859,
+      "loss": 0.0035,
+      "step": 2356
+    },
+    {
+      "epoch": 0.11167969675432363,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019395340533990744,
+      "loss": 0.2011,
+      "step": 2357
+    },
+    {
+      "epoch": 0.11172707889125799,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019394830424754252,
+      "loss": 0.8884,
+      "step": 2358
+    },
+    {
+      "epoch": 0.11177446102819237,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019394320107150428,
+      "loss": 0.7974,
+      "step": 2359
+    },
+    {
+      "epoch": 0.11182184316512675,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001939380958119059,
+      "loss": 0.2071,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11186922530206113,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019393298846886062,
+      "loss": 0.1647,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1119166074389955,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001939278790424817,
+      "loss": 0.9444,
+      "step": 2362
+    },
+    {
+      "epoch": 0.11196398957592987,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019392276753288248,
+      "loss": 0.9224,
+      "step": 2363
+    },
+    {
+      "epoch": 0.11201137171286425,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001939176539401763,
+      "loss": 0.7519,
+      "step": 2364
+    },
+    {
+      "epoch": 0.11205875384979863,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019391253826447663,
+      "loss": 0.1332,
+      "step": 2365
+    },
+    {
+      "epoch": 0.112106135986733,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019390742050589687,
+      "loss": 0.548,
+      "step": 2366
+    },
+    {
+      "epoch": 0.11215351812366738,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019390230066455058,
+      "loss": 1.065,
+      "step": 2367
+    },
+    {
+      "epoch": 0.11220090026060175,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019389717874055125,
+      "loss": 0.1695,
+      "step": 2368
+    },
+    {
+      "epoch": 0.11224828239753613,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001938920547340125,
+      "loss": 0.9457,
+      "step": 2369
+    },
+    {
+      "epoch": 0.1122956645344705,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019388692864504798,
+      "loss": 0.2065,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11234304667140488,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019388180047377144,
+      "loss": 0.761,
+      "step": 2371
+    },
+    {
+      "epoch": 0.11239042880833926,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001938766702202965,
+      "loss": 0.9179,
+      "step": 2372
+    },
+    {
+      "epoch": 0.11243781094527364,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019387153788473705,
+      "loss": 1.2718,
+      "step": 2373
+    },
+    {
+      "epoch": 0.112485193082208,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019386640346720686,
+      "loss": 0.8863,
+      "step": 2374
+    },
+    {
+      "epoch": 0.11253257521914238,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019386126696781982,
+      "loss": 0.0903,
+      "step": 2375
+    },
+    {
+      "epoch": 0.11257995735607676,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019385612838668985,
+      "loss": 0.3469,
+      "step": 2376
+    },
+    {
+      "epoch": 0.11262733949301114,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019385098772393096,
+      "loss": 0.8277,
+      "step": 2377
+    },
+    {
+      "epoch": 0.1126747216299455,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001938458449796571,
+      "loss": 0.7903,
+      "step": 2378
+    },
+    {
+      "epoch": 0.11272210376687988,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019384070015398235,
+      "loss": 0.6248,
+      "step": 2379
+    },
+    {
+      "epoch": 0.11276948590381426,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019383555324702082,
+      "loss": 0.5589,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11281686804074864,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019383040425888668,
+      "loss": 0.4301,
+      "step": 2381
+    },
+    {
+      "epoch": 0.11286425017768302,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019382525318969413,
+      "loss": 1.2382,
+      "step": 2382
+    },
+    {
+      "epoch": 0.11291163231461739,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001938201000395574,
+      "loss": 0.9135,
+      "step": 2383
+    },
+    {
+      "epoch": 0.11295901445155176,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019381494480859076,
+      "loss": 0.4733,
+      "step": 2384
+    },
+    {
+      "epoch": 0.11300639658848614,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001938097874969086,
+      "loss": 1.1656,
+      "step": 2385
+    },
+    {
+      "epoch": 0.11305377872542052,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019380462810462525,
+      "loss": 0.0332,
+      "step": 2386
+    },
+    {
+      "epoch": 0.11310116086235489,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001937994666318552,
+      "loss": 1.0793,
+      "step": 2387
+    },
+    {
+      "epoch": 0.11314854299928927,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019379430307871288,
+      "loss": 0.7872,
+      "step": 2388
+    },
+    {
+      "epoch": 0.11319592513622365,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001937891374453128,
+      "loss": 0.0647,
+      "step": 2389
+    },
+    {
+      "epoch": 0.11324330727315803,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019378396973176955,
+      "loss": 0.4569,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11329068941009239,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019377879993819777,
+      "loss": 0.8047,
+      "step": 2391
+    },
+    {
+      "epoch": 0.11333807154702677,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019377362806471208,
+      "loss": 1.0586,
+      "step": 2392
+    },
+    {
+      "epoch": 0.11338545368396115,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001937684541114272,
+      "loss": 1.0862,
+      "step": 2393
+    },
+    {
+      "epoch": 0.11343283582089553,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019376327807845792,
+      "loss": 1.7054,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1134802179578299,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019375809996591896,
+      "loss": 1.0919,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11352760009476427,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019375291977392523,
+      "loss": 0.983,
+      "step": 2396
+    },
+    {
+      "epoch": 0.11357498223169865,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001937477375025916,
+      "loss": 0.7082,
+      "step": 2397
+    },
+    {
+      "epoch": 0.11362236436863303,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000193742553152033,
+      "loss": 0.8129,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1136697465055674,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019373736672236445,
+      "loss": 1.3653,
+      "step": 2399
+    },
+    {
+      "epoch": 0.11371712864250177,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001937321782137009,
+      "loss": 1.09,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11376451077943615,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001937269876261575,
+      "loss": 0.521,
+      "step": 2401
+    },
+    {
+      "epoch": 0.11381189291637053,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019372179495984936,
+      "loss": 0.9942,
+      "step": 2402
+    },
+    {
+      "epoch": 0.1138592750533049,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019371660021489162,
+      "loss": 1.1217,
+      "step": 2403
+    },
+    {
+      "epoch": 0.11390665719023928,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019371140339139952,
+      "loss": 1.177,
+      "step": 2404
+    },
+    {
+      "epoch": 0.11395403932717366,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001937062044894883,
+      "loss": 0.0618,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11400142146410804,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019370100350927328,
+      "loss": 1.4012,
+      "step": 2406
+    },
+    {
+      "epoch": 0.1140488036010424,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001936958004508698,
+      "loss": 0.2863,
+      "step": 2407
+    },
+    {
+      "epoch": 0.11409618573797678,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00019369059531439332,
+      "loss": 1.1873,
+      "step": 2408
+    },
+    {
+      "epoch": 0.11414356787491116,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019368538809995918,
+      "loss": 1.0045,
+      "step": 2409
+    },
+    {
+      "epoch": 0.11419095001184554,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001936801788076829,
+      "loss": 0.8343,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1142383321487799,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019367496743768009,
+      "loss": 0.5111,
+      "step": 2411
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00019366975399006626,
+      "loss": 0.1618,
+      "step": 2412
+    },
+    {
+      "epoch": 0.11433309642264866,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019366453846495705,
+      "loss": 0.8744,
+      "step": 2413
+    },
+    {
+      "epoch": 0.11438047855958304,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00019365932086246813,
+      "loss": 0.7495,
+      "step": 2414
+    },
+    {
+      "epoch": 0.11442786069651742,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019365410118271528,
+      "loss": 0.179,
+      "step": 2415
+    },
+    {
+      "epoch": 0.11447524283345178,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019364887942581417,
+      "loss": 0.8303,
+      "step": 2416
+    },
+    {
+      "epoch": 0.11452262497038616,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001936436555918807,
+      "loss": 0.5321,
+      "step": 2417
+    },
+    {
+      "epoch": 0.11457000710732054,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001936384296810307,
+      "loss": 0.6623,
+      "step": 2418
+    },
+    {
+      "epoch": 0.11461738924425492,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019363320169338004,
+      "loss": 0.2194,
+      "step": 2419
+    },
+    {
+      "epoch": 0.11466477138118929,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001936279716290447,
+      "loss": 0.5661,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11471215351812367,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019362273948814068,
+      "loss": 1.2285,
+      "step": 2421
+    },
+    {
+      "epoch": 0.11475953565505805,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019361750527078405,
+      "loss": 1.3478,
+      "step": 2422
+    },
+    {
+      "epoch": 0.11480691779199242,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019361226897709086,
+      "loss": 1.6165,
+      "step": 2423
+    },
+    {
+      "epoch": 0.11485429992892679,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019360703060717724,
+      "loss": 0.9164,
+      "step": 2424
+    },
+    {
+      "epoch": 0.11490168206586117,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001936017901611594,
+      "loss": 0.9924,
+      "step": 2425
+    },
+    {
+      "epoch": 0.11494906420279555,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019359654763915354,
+      "loss": 0.7366,
+      "step": 2426
+    },
+    {
+      "epoch": 0.11499644633972993,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019359130304127595,
+      "loss": 0.0893,
+      "step": 2427
+    },
+    {
+      "epoch": 0.11504382847666429,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019358605636764296,
+      "loss": 0.8715,
+      "step": 2428
+    },
+    {
+      "epoch": 0.11509121061359867,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001935808076183709,
+      "loss": 1.0793,
+      "step": 2429
+    },
+    {
+      "epoch": 0.11513859275053305,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019357555679357623,
+      "loss": 0.9252,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11518597488746743,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001935703038933754,
+      "loss": 0.8074,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1152333570244018,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019356504891788486,
+      "loss": 1.7112,
+      "step": 2432
+    },
+    {
+      "epoch": 0.11528073916133617,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019355979186722118,
+      "loss": 1.0211,
+      "step": 2433
+    },
+    {
+      "epoch": 0.11532812129827055,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000193554532741501,
+      "loss": 0.1586,
+      "step": 2434
+    },
+    {
+      "epoch": 0.11537550343520493,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019354927154084095,
+      "loss": 0.1016,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1154228855721393,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019354400826535767,
+      "loss": 1.0657,
+      "step": 2436
+    },
+    {
+      "epoch": 0.11547026770907368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019353874291516793,
+      "loss": 0.4119,
+      "step": 2437
+    },
+    {
+      "epoch": 0.11551764984600806,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001935334754903885,
+      "loss": 0.0556,
+      "step": 2438
+    },
+    {
+      "epoch": 0.11556503198294243,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00019352820599113622,
+      "loss": 0.0062,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1156124141198768,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019352293441752798,
+      "loss": 1.1129,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11565979625681118,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019351766076968065,
+      "loss": 1.6857,
+      "step": 2441
+    },
+    {
+      "epoch": 0.11570717839374556,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001935123850477112,
+      "loss": 1.5279,
+      "step": 2442
+    },
+    {
+      "epoch": 0.11575456053067994,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019350710725173667,
+      "loss": 0.855,
+      "step": 2443
+    },
+    {
+      "epoch": 0.11580194266761432,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001935018273818741,
+      "loss": 1.1485,
+      "step": 2444
+    },
+    {
+      "epoch": 0.11584932480454868,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019349654543824059,
+      "loss": 0.5004,
+      "step": 2445
+    },
+    {
+      "epoch": 0.11589670694148306,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019349126142095328,
+      "loss": 0.9367,
+      "step": 2446
+    },
+    {
+      "epoch": 0.11594408907841744,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019348597533012937,
+      "loss": 0.3459,
+      "step": 2447
+    },
+    {
+      "epoch": 0.11599147121535182,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019348068716588615,
+      "loss": 0.1398,
+      "step": 2448
+    },
+    {
+      "epoch": 0.11603885335228618,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001934753969283408,
+      "loss": 0.9748,
+      "step": 2449
+    },
+    {
+      "epoch": 0.11608623548922056,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019347010461761075,
+      "loss": 1.0431,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11613361762615494,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019346481023381333,
+      "loss": 1.1636,
+      "step": 2451
+    },
+    {
+      "epoch": 0.11618099976308932,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019345951377706597,
+      "loss": 0.1638,
+      "step": 2452
+    },
+    {
+      "epoch": 0.11622838190002369,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019345421524748614,
+      "loss": 1.0393,
+      "step": 2453
+    },
+    {
+      "epoch": 0.11627576403695807,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019344891464519138,
+      "loss": 0.4738,
+      "step": 2454
+    },
+    {
+      "epoch": 0.11632314617389244,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019344361197029918,
+      "loss": 0.9189,
+      "step": 2455
+    },
+    {
+      "epoch": 0.11637052831082682,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019343830722292726,
+      "loss": 0.9371,
+      "step": 2456
+    },
+    {
+      "epoch": 0.11641791044776119,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019343300040319317,
+      "loss": 0.3964,
+      "step": 2457
+    },
+    {
+      "epoch": 0.11646529258469557,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019342769151121467,
+      "loss": 0.8681,
+      "step": 2458
+    },
+    {
+      "epoch": 0.11651267472162995,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001934223805471095,
+      "loss": 0.8003,
+      "step": 2459
+    },
+    {
+      "epoch": 0.11656005685856433,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019341706751099542,
+      "loss": 0.0212,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11660743899549869,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.00019341175240299028,
+      "loss": 0.0065,
+      "step": 2461
+    },
+    {
+      "epoch": 0.11665482113243307,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000193406435223212,
+      "loss": 1.2321,
+      "step": 2462
+    },
+    {
+      "epoch": 0.11670220326936745,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019340111597177843,
+      "loss": 1.1462,
+      "step": 2463
+    },
+    {
+      "epoch": 0.11674958540630183,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019339579464880763,
+      "loss": 0.9835,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1167969675432362,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019339047125441756,
+      "loss": 0.8235,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11684434968017057,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019338514578872633,
+      "loss": 0.8367,
+      "step": 2466
+    },
+    {
+      "epoch": 0.11689173181710495,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019337981825185202,
+      "loss": 1.147,
+      "step": 2467
+    },
+    {
+      "epoch": 0.11693911395403933,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019337448864391283,
+      "loss": 0.4575,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1169864960909737,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00019336915696502693,
+      "loss": 0.0063,
+      "step": 2469
+    },
+    {
+      "epoch": 0.11703387822790808,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001933638232153126,
+      "loss": 1.6739,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11708126036484245,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.00019335848739488807,
+      "loss": 0.0051,
+      "step": 2471
+    },
+    {
+      "epoch": 0.11712864250177683,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019335314950387174,
+      "loss": 0.0252,
+      "step": 2472
+    },
+    {
+      "epoch": 0.11717602463871121,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019334780954238204,
+      "loss": 1.0257,
+      "step": 2473
+    },
+    {
+      "epoch": 0.11722340677564558,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001933424675105373,
+      "loss": 0.0673,
+      "step": 2474
+    },
+    {
+      "epoch": 0.11727078891257996,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019333712340845608,
+      "loss": 1.181,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11731817104951434,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001933317772362569,
+      "loss": 1.1553,
+      "step": 2476
+    },
+    {
+      "epoch": 0.11736555318644872,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019332642899405825,
+      "loss": 0.8072,
+      "step": 2477
+    },
+    {
+      "epoch": 0.11741293532338308,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019332107868197886,
+      "loss": 0.0409,
+      "step": 2478
+    },
+    {
+      "epoch": 0.11746031746031746,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019331572630013736,
+      "loss": 0.2116,
+      "step": 2479
+    },
+    {
+      "epoch": 0.11750769959725184,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001933103718486524,
+      "loss": 0.1154,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11755508173418622,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019330501532764283,
+      "loss": 0.9118,
+      "step": 2481
+    },
+    {
+      "epoch": 0.11760246387112058,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019329965673722737,
+      "loss": 0.5453,
+      "step": 2482
+    },
+    {
+      "epoch": 0.11764984600805496,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001932942960775249,
+      "loss": 0.026,
+      "step": 2483
+    },
+    {
+      "epoch": 0.11769722814498934,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019328893334865431,
+      "loss": 1.1812,
+      "step": 2484
+    },
+    {
+      "epoch": 0.11774461028192372,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001932835685507346,
+      "loss": 1.4514,
+      "step": 2485
+    },
+    {
+      "epoch": 0.11779199241885809,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019327820168388464,
+      "loss": 0.6865,
+      "step": 2486
+    },
+    {
+      "epoch": 0.11783937455579246,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019327283274822357,
+      "loss": 1.2399,
+      "step": 2487
+    },
+    {
+      "epoch": 0.11788675669272684,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019326746174387038,
+      "loss": 0.2101,
+      "step": 2488
+    },
+    {
+      "epoch": 0.11793413882966122,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019326208867094424,
+      "loss": 0.0275,
+      "step": 2489
+    },
+    {
+      "epoch": 0.11798152096659559,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001932567135295643,
+      "loss": 0.8127,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11802890310352997,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019325133631984981,
+      "loss": 0.7025,
+      "step": 2491
+    },
+    {
+      "epoch": 0.11807628524046435,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019324595704192,
+      "loss": 0.1387,
+      "step": 2492
+    },
+    {
+      "epoch": 0.11812366737739873,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001932405756958942,
+      "loss": 1.3282,
+      "step": 2493
+    },
+    {
+      "epoch": 0.11817104951433309,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019323519228189173,
+      "loss": 0.1715,
+      "step": 2494
+    },
+    {
+      "epoch": 0.11821843165126747,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000193229806800032,
+      "loss": 0.1249,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11826581378820185,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019322441925043448,
+      "loss": 1.3299,
+      "step": 2496
+    },
+    {
+      "epoch": 0.11831319592513623,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019321902963321863,
+      "loss": 1.0653,
+      "step": 2497
+    },
+    {
+      "epoch": 0.11836057806207059,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019321363794850397,
+      "loss": 0.9062,
+      "step": 2498
+    },
+    {
+      "epoch": 0.11840796019900497,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019320824419641016,
+      "loss": 1.5567,
+      "step": 2499
+    },
+    {
+      "epoch": 0.11845534233593935,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019320284837705671,
+      "loss": 0.9094,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11850272447287373,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001931974504905634,
+      "loss": 1.2802,
+      "step": 2501
+    },
+    {
+      "epoch": 0.11855010660980811,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019319205053704993,
+      "loss": 1.1052,
+      "step": 2502
+    },
+    {
+      "epoch": 0.11859748874674247,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019318664851663599,
+      "loss": 0.3962,
+      "step": 2503
+    },
+    {
+      "epoch": 0.11864487088367685,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019318124442944146,
+      "loss": 1.1065,
+      "step": 2504
+    },
+    {
+      "epoch": 0.11869225302061123,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001931758382755862,
+      "loss": 1.4303,
+      "step": 2505
+    },
+    {
+      "epoch": 0.11873963515754561,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019317043005519008,
+      "loss": 0.8571,
+      "step": 2506
+    },
+    {
+      "epoch": 0.11878701729447998,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019316501976837308,
+      "loss": 1.3223,
+      "step": 2507
+    },
+    {
+      "epoch": 0.11883439943141436,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00019315960741525512,
+      "loss": 0.8867,
+      "step": 2508
+    },
+    {
+      "epoch": 0.11888178156834874,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019315419299595633,
+      "loss": 0.8417,
+      "step": 2509
+    },
+    {
+      "epoch": 0.11892916370528311,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019314877651059676,
+      "loss": 0.8918,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11897654584221748,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001931433579592966,
+      "loss": 0.7333,
+      "step": 2511
+    },
+    {
+      "epoch": 0.11902392797915186,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001931379373421759,
+      "loss": 1.0359,
+      "step": 2512
+    },
+    {
+      "epoch": 0.11907131011608624,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019313251465935498,
+      "loss": 1.1578,
+      "step": 2513
+    },
+    {
+      "epoch": 0.11911869225302062,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019312708991095408,
+      "loss": 1.3204,
+      "step": 2514
+    },
+    {
+      "epoch": 0.11916607438995498,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019312166309709352,
+      "loss": 0.3033,
+      "step": 2515
+    },
+    {
+      "epoch": 0.11921345652688936,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019311623421789368,
+      "loss": 1.3375,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11926083866382374,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019311080327347492,
+      "loss": 1.2571,
+      "step": 2517
+    },
+    {
+      "epoch": 0.11930822080075812,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019310537026395773,
+      "loss": 0.9547,
+      "step": 2518
+    },
+    {
+      "epoch": 0.11935560293769248,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019309993518946264,
+      "loss": 0.0299,
+      "step": 2519
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001930944980501101,
+      "loss": 0.7248,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11945036721156124,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019308905884602074,
+      "loss": 0.7792,
+      "step": 2521
+    },
+    {
+      "epoch": 0.11949774934849562,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019308361757731528,
+      "loss": 1.0552,
+      "step": 2522
+    },
+    {
+      "epoch": 0.11954513148542999,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019307817424411426,
+      "loss": 0.7193,
+      "step": 2523
+    },
+    {
+      "epoch": 0.11959251362236437,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001930727288465385,
+      "loss": 1.0973,
+      "step": 2524
+    },
+    {
+      "epoch": 0.11963989575929874,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019306728138470877,
+      "loss": 0.8714,
+      "step": 2525
+    },
+    {
+      "epoch": 0.11968727789623312,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019306183185874585,
+      "loss": 0.0116,
+      "step": 2526
+    },
+    {
+      "epoch": 0.11973466003316749,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019305638026877064,
+      "loss": 0.1626,
+      "step": 2527
+    },
+    {
+      "epoch": 0.11978204217010187,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019305092661490406,
+      "loss": 0.0571,
+      "step": 2528
+    },
+    {
+      "epoch": 0.11982942430703625,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000193045470897267,
+      "loss": 0.9813,
+      "step": 2529
+    },
+    {
+      "epoch": 0.11987680644397063,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001930400131159805,
+      "loss": 1.069,
+      "step": 2530
+    },
+    {
+      "epoch": 0.119924188580905,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001930345532711656,
+      "loss": 1.0806,
+      "step": 2531
+    },
+    {
+      "epoch": 0.11997157071783937,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019302909136294344,
+      "loss": 0.8771,
+      "step": 2532
+    },
+    {
+      "epoch": 0.12001895285477375,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019302362739143512,
+      "loss": 0.1197,
+      "step": 2533
+    },
+    {
+      "epoch": 0.12006633499170813,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019301816135676182,
+      "loss": 1.5407,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12011371712864251,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019301269325904476,
+      "loss": 0.9555,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12016109926557687,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.00019300722309840526,
+      "loss": 0.0034,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12020848140251125,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019300175087496463,
+      "loss": 0.8896,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12025586353944563,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019299627658884421,
+      "loss": 0.0083,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12030324567638001,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019299080024016543,
+      "loss": 0.7692,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12035062781331438,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019298532182904975,
+      "loss": 0.9235,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12039800995024875,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019297984135561866,
+      "loss": 1.2387,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12044539208718313,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019297435881999376,
+      "loss": 0.1512,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12049277422411751,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001929688742222966,
+      "loss": 0.6726,
+      "step": 2543
+    },
+    {
+      "epoch": 0.12054015636105188,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019296338756264882,
+      "loss": 1.4926,
+      "step": 2544
+    },
+    {
+      "epoch": 0.12058753849798626,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019295789884117212,
+      "loss": 1.0649,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12063492063492064,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019295240805798826,
+      "loss": 0.0948,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12068230277185502,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000192946915213219,
+      "loss": 0.7311,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12072968490878938,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019294142030698615,
+      "loss": 0.1531,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12077706704572376,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019293592333941158,
+      "loss": 0.6935,
+      "step": 2549
+    },
+    {
+      "epoch": 0.12082444918265814,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019293042431061725,
+      "loss": 1.1029,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12087183131959252,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019292492322072507,
+      "loss": 1.3596,
+      "step": 2551
+    },
+    {
+      "epoch": 0.12091921345652688,
+      "grad_norm": 0.019775390625,
+      "learning_rate": 0.0001929194200698571,
+      "loss": 0.0017,
+      "step": 2552
+    },
+    {
+      "epoch": 0.12096659559346126,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00019291391485813533,
+      "loss": 0.8052,
+      "step": 2553
+    },
+    {
+      "epoch": 0.12101397773039564,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019290840758568194,
+      "loss": 0.0633,
+      "step": 2554
+    },
+    {
+      "epoch": 0.12106135986733002,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000192902898252619,
+      "loss": 0.4026,
+      "step": 2555
+    },
+    {
+      "epoch": 0.12110874200426439,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019289738685906874,
+      "loss": 1.0573,
+      "step": 2556
+    },
+    {
+      "epoch": 0.12115612414119876,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0001928918734051534,
+      "loss": 0.008,
+      "step": 2557
+    },
+    {
+      "epoch": 0.12120350627813314,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00019288635789099524,
+      "loss": 1.0692,
+      "step": 2558
+    },
+    {
+      "epoch": 0.12125088841506752,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001928808403167166,
+      "loss": 1.4574,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1212982705520019,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019287532068243984,
+      "loss": 0.7272,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12134565268893627,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00019286979898828742,
+      "loss": 0.0126,
+      "step": 2561
+    },
+    {
+      "epoch": 0.12139303482587065,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019286427523438178,
+      "loss": 1.0641,
+      "step": 2562
+    },
+    {
+      "epoch": 0.12144041696280503,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001928587494208454,
+      "loss": 0.1086,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1214877990997394,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001928532215478009,
+      "loss": 1.4355,
+      "step": 2564
+    },
+    {
+      "epoch": 0.12153518123667377,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019284769161537083,
+      "loss": 0.7239,
+      "step": 2565
+    },
+    {
+      "epoch": 0.12158256337360815,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019284215962367786,
+      "loss": 1.2765,
+      "step": 2566
+    },
+    {
+      "epoch": 0.12162994551054253,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001928366255728447,
+      "loss": 0.2377,
+      "step": 2567
+    },
+    {
+      "epoch": 0.12167732764747691,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019283108946299403,
+      "loss": 0.0398,
+      "step": 2568
+    },
+    {
+      "epoch": 0.12172470978441127,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001928255512942487,
+      "loss": 0.9694,
+      "step": 2569
+    },
+    {
+      "epoch": 0.12177209192134565,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019282001106673153,
+      "loss": 1.048,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12181947405828003,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019281446878056534,
+      "loss": 0.5022,
+      "step": 2571
+    },
+    {
+      "epoch": 0.12186685619521441,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00019280892443587316,
+      "loss": 0.0089,
+      "step": 2572
+    },
+    {
+      "epoch": 0.12191423833214877,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019280337803277781,
+      "loss": 0.6116,
+      "step": 2573
+    },
+    {
+      "epoch": 0.12196162046908315,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019279782957140243,
+      "loss": 0.2022,
+      "step": 2574
+    },
+    {
+      "epoch": 0.12200900260601753,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00019279227905187005,
+      "loss": 0.0068,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12205638474295191,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019278672647430375,
+      "loss": 0.4393,
+      "step": 2576
+    },
+    {
+      "epoch": 0.12210376687988628,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019278117183882664,
+      "loss": 1.1396,
+      "step": 2577
+    },
+    {
+      "epoch": 0.12215114901682066,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019277561514556202,
+      "loss": 0.3896,
+      "step": 2578
+    },
+    {
+      "epoch": 0.12219853115375504,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019277005639463304,
+      "loss": 1.1313,
+      "step": 2579
+    },
+    {
+      "epoch": 0.12224591329068941,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019276449558616306,
+      "loss": 0.6474,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12229329542762378,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019275893272027535,
+      "loss": 0.0849,
+      "step": 2581
+    },
+    {
+      "epoch": 0.12234067756455816,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001927533677970933,
+      "loss": 1.3711,
+      "step": 2582
+    },
+    {
+      "epoch": 0.12238805970149254,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001927478008167404,
+      "loss": 0.8757,
+      "step": 2583
+    },
+    {
+      "epoch": 0.12243544183842692,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019274223177934,
+      "loss": 0.7932,
+      "step": 2584
+    },
+    {
+      "epoch": 0.12248282397536128,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019273666068501575,
+      "loss": 0.963,
+      "step": 2585
+    },
+    {
+      "epoch": 0.12253020611229566,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0001927310875338911,
+      "loss": 1.1819,
+      "step": 2586
+    },
+    {
+      "epoch": 0.12257758824923004,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019272551232608974,
+      "loss": 0.968,
+      "step": 2587
+    },
+    {
+      "epoch": 0.12262497038616442,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019271993506173526,
+      "loss": 0.5281,
+      "step": 2588
+    },
+    {
+      "epoch": 0.12267235252309878,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019271435574095137,
+      "loss": 0.8354,
+      "step": 2589
+    },
+    {
+      "epoch": 0.12271973466003316,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019270877436386186,
+      "loss": 1.0137,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12276711679696754,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019270319093059044,
+      "loss": 0.9656,
+      "step": 2591
+    },
+    {
+      "epoch": 0.12281449893390192,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.000192697605441261,
+      "loss": 0.6587,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1228618810708363,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019269201789599743,
+      "loss": 0.191,
+      "step": 2593
+    },
+    {
+      "epoch": 0.12290926320777067,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019268642829492363,
+      "loss": 0.1004,
+      "step": 2594
+    },
+    {
+      "epoch": 0.12295664534470505,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00019268083663816357,
+      "loss": 0.0405,
+      "step": 2595
+    },
+    {
+      "epoch": 0.12300402748163942,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00019267524292584126,
+      "loss": 0.0198,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1230514096185738,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00019266964715808078,
+      "loss": 0.0101,
+      "step": 2597
+    },
+    {
+      "epoch": 0.12309879175550817,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019266404933500624,
+      "loss": 0.7614,
+      "step": 2598
+    },
+    {
+      "epoch": 0.12314617389244255,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019265844945674177,
+      "loss": 0.6976,
+      "step": 2599
+    },
+    {
+      "epoch": 0.12319355602937693,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019265284752341158,
+      "loss": 0.7968,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1232409381663113,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019264724353514,
+      "loss": 0.6219,
+      "step": 2601
+    },
+    {
+      "epoch": 0.12328832030324567,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019264163749205116,
+      "loss": 1.4453,
+      "step": 2602
+    },
+    {
+      "epoch": 0.12333570244018005,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001926360293942695,
+      "loss": 1.453,
+      "step": 2603
+    },
+    {
+      "epoch": 0.12338308457711443,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00019263041924191937,
+      "loss": 0.0269,
+      "step": 2604
+    },
+    {
+      "epoch": 0.12343046671404881,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019262480703512521,
+      "loss": 0.6972,
+      "step": 2605
+    },
+    {
+      "epoch": 0.12347784885098317,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019261919277401154,
+      "loss": 0.6406,
+      "step": 2606
+    },
+    {
+      "epoch": 0.12352523098791755,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00019261357645870275,
+      "loss": 0.2155,
+      "step": 2607
+    },
+    {
+      "epoch": 0.12357261312485193,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019260795808932353,
+      "loss": 0.0741,
+      "step": 2608
+    },
+    {
+      "epoch": 0.12361999526178631,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019260233766599843,
+      "loss": 1.0198,
+      "step": 2609
+    },
+    {
+      "epoch": 0.12366737739872068,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001925967151888521,
+      "loss": 1.0583,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12371475953565506,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001925910906580093,
+      "loss": 1.1802,
+      "step": 2611
+    },
+    {
+      "epoch": 0.12376214167258943,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019258546407359468,
+      "loss": 0.5909,
+      "step": 2612
+    },
+    {
+      "epoch": 0.12380952380952381,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001925798354357331,
+      "loss": 1.1352,
+      "step": 2613
+    },
+    {
+      "epoch": 0.12385690594645818,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001925742047445494,
+      "loss": 0.2152,
+      "step": 2614
+    },
+    {
+      "epoch": 0.12390428808339256,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019256857200016845,
+      "loss": 0.5009,
+      "step": 2615
+    },
+    {
+      "epoch": 0.12395167022032694,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019256293720271514,
+      "loss": 0.6322,
+      "step": 2616
+    },
+    {
+      "epoch": 0.12399905235726132,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001925573003523145,
+      "loss": 0.7375,
+      "step": 2617
+    },
+    {
+      "epoch": 0.12404643449419568,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019255166144909152,
+      "loss": 0.8515,
+      "step": 2618
+    },
+    {
+      "epoch": 0.12409381663113006,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019254602049317127,
+      "loss": 0.8582,
+      "step": 2619
+    },
+    {
+      "epoch": 0.12414119876806444,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019254037748467885,
+      "loss": 0.7536,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12418858090499882,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019253473242373945,
+      "loss": 1.3561,
+      "step": 2621
+    },
+    {
+      "epoch": 0.1242359630419332,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019252908531047823,
+      "loss": 1.0946,
+      "step": 2622
+    },
+    {
+      "epoch": 0.12428334517886756,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019252343614502046,
+      "loss": 1.1829,
+      "step": 2623
+    },
+    {
+      "epoch": 0.12433072731580194,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00019251778492749145,
+      "loss": 0.5284,
+      "step": 2624
+    },
+    {
+      "epoch": 0.12437810945273632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019251213165801648,
+      "loss": 0.7203,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1244254915896707,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000192506476336721,
+      "loss": 1.3022,
+      "step": 2626
+    },
+    {
+      "epoch": 0.12447287372660507,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019250081896373037,
+      "loss": 1.3903,
+      "step": 2627
+    },
+    {
+      "epoch": 0.12452025586353944,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019249515953917014,
+      "loss": 1.207,
+      "step": 2628
+    },
+    {
+      "epoch": 0.12456763800047382,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019248949806316578,
+      "loss": 0.473,
+      "step": 2629
+    },
+    {
+      "epoch": 0.1246150201374082,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019248383453584288,
+      "loss": 0.1061,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12466240227434257,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000192478168957327,
+      "loss": 0.4016,
+      "step": 2631
+    },
+    {
+      "epoch": 0.12470978441127695,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001924725013277439,
+      "loss": 0.8508,
+      "step": 2632
+    },
+    {
+      "epoch": 0.12475716654821133,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019246683164721916,
+      "loss": 0.8059,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1248045486851457,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019246115991587858,
+      "loss": 1.0854,
+      "step": 2634
+    },
+    {
+      "epoch": 0.12485193082208007,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000192455486133848,
+      "loss": 1.3106,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12489931295901445,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001924498103012532,
+      "loss": 1.2176,
+      "step": 2636
+    },
+    {
+      "epoch": 0.12494669509594883,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019244413241822008,
+      "loss": 0.8235,
+      "step": 2637
+    },
+    {
+      "epoch": 0.12499407723288321,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019243845248487456,
+      "loss": 1.1107,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1250414593698176,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019243277050134266,
+      "loss": 0.9661,
+      "step": 2639
+    },
+    {
+      "epoch": 0.12508884150675195,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019242708646775034,
+      "loss": 0.8563,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12513622364368632,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001924214003842237,
+      "loss": 1.1063,
+      "step": 2641
+    },
+    {
+      "epoch": 0.1251836057806207,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001924157122508888,
+      "loss": 0.0833,
+      "step": 2642
+    },
+    {
+      "epoch": 0.12523098791755508,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001924100220678719,
+      "loss": 1.1375,
+      "step": 2643
+    },
+    {
+      "epoch": 0.12527837005448947,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001924043298352991,
+      "loss": 0.0631,
+      "step": 2644
+    },
+    {
+      "epoch": 0.12532575219142383,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019239863555329671,
+      "loss": 1.0201,
+      "step": 2645
+    },
+    {
+      "epoch": 0.1253731343283582,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.000192392939221991,
+      "loss": 0.2895,
+      "step": 2646
+    },
+    {
+      "epoch": 0.1254205164652926,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001923872408415083,
+      "loss": 0.9183,
+      "step": 2647
+    },
+    {
+      "epoch": 0.12546789860222696,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000192381540411975,
+      "loss": 0.0157,
+      "step": 2648
+    },
+    {
+      "epoch": 0.12551528073916132,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019237583793351758,
+      "loss": 0.6224,
+      "step": 2649
+    },
+    {
+      "epoch": 0.12556266287609572,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001923701334062624,
+      "loss": 0.5273,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12561004501303008,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001923644268303361,
+      "loss": 0.8104,
+      "step": 2651
+    },
+    {
+      "epoch": 0.12565742714996447,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019235871820586517,
+      "loss": 0.8764,
+      "step": 2652
+    },
+    {
+      "epoch": 0.12570480928689884,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019235300753297626,
+      "loss": 0.928,
+      "step": 2653
+    },
+    {
+      "epoch": 0.1257521914238332,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000192347294811796,
+      "loss": 0.51,
+      "step": 2654
+    },
+    {
+      "epoch": 0.1257995735607676,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019234158004245112,
+      "loss": 1.4853,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12584695569770196,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019233586322506832,
+      "loss": 1.1049,
+      "step": 2656
+    },
+    {
+      "epoch": 0.12589433783463636,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019233014435977448,
+      "loss": 0.4638,
+      "step": 2657
+    },
+    {
+      "epoch": 0.12594171997157072,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019232442344669634,
+      "loss": 0.6833,
+      "step": 2658
+    },
+    {
+      "epoch": 0.12598910210850509,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019231870048596085,
+      "loss": 0.0987,
+      "step": 2659
+    },
+    {
+      "epoch": 0.12603648424543948,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019231297547769492,
+      "loss": 0.349,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12608386638237384,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001923072484220255,
+      "loss": 1.1312,
+      "step": 2661
+    },
+    {
+      "epoch": 0.1261312485193082,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019230151931907963,
+      "loss": 1.0402,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1261786306562426,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001922957881689844,
+      "loss": 1.4235,
+      "step": 2663
+    },
+    {
+      "epoch": 0.12622601279317697,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019229005497186687,
+      "loss": 1.2282,
+      "step": 2664
+    },
+    {
+      "epoch": 0.12627339493011136,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019228431972785425,
+      "loss": 0.1748,
+      "step": 2665
+    },
+    {
+      "epoch": 0.12632077706704573,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019227858243707371,
+      "loss": 0.4302,
+      "step": 2666
+    },
+    {
+      "epoch": 0.1263681592039801,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001922728430996525,
+      "loss": 0.7094,
+      "step": 2667
+    },
+    {
+      "epoch": 0.12641554134091448,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019226710171571791,
+      "loss": 0.7875,
+      "step": 2668
+    },
+    {
+      "epoch": 0.12646292347784885,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019226135828539728,
+      "loss": 0.8099,
+      "step": 2669
+    },
+    {
+      "epoch": 0.12651030561478321,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.000192255612808818,
+      "loss": 0.1927,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1265576877517176,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001922498652861075,
+      "loss": 0.7767,
+      "step": 2671
+    },
+    {
+      "epoch": 0.12660506988865197,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001922441157173932,
+      "loss": 0.8704,
+      "step": 2672
+    },
+    {
+      "epoch": 0.12665245202558637,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019223836410280274,
+      "loss": 1.3345,
+      "step": 2673
+    },
+    {
+      "epoch": 0.12669983416252073,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019223261044246356,
+      "loss": 0.2152,
+      "step": 2674
+    },
+    {
+      "epoch": 0.1267472162994551,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019222685473650338,
+      "loss": 0.6204,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1267945984363895,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019222109698504975,
+      "loss": 0.3218,
+      "step": 2676
+    },
+    {
+      "epoch": 0.12684198057332385,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019221533718823044,
+      "loss": 1.485,
+      "step": 2677
+    },
+    {
+      "epoch": 0.12688936271025822,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019220957534617314,
+      "loss": 0.0332,
+      "step": 2678
+    },
+    {
+      "epoch": 0.1269367448471926,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001922038114590057,
+      "loss": 1.5897,
+      "step": 2679
+    },
+    {
+      "epoch": 0.12698412698412698,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019219804552685595,
+      "loss": 0.7048,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12703150912106137,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019219227754985175,
+      "loss": 0.0491,
+      "step": 2681
+    },
+    {
+      "epoch": 0.12707889125799574,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.000192186507528121,
+      "loss": 0.8848,
+      "step": 2682
+    },
+    {
+      "epoch": 0.1271262733949301,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019218073546179172,
+      "loss": 1.0105,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1271736555318645,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001921749613509919,
+      "loss": 0.1505,
+      "step": 2684
+    },
+    {
+      "epoch": 0.12722103766879886,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019216918519584966,
+      "loss": 0.5415,
+      "step": 2685
+    },
+    {
+      "epoch": 0.12726841980573325,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019216340699649304,
+      "loss": 0.145,
+      "step": 2686
+    },
+    {
+      "epoch": 0.12731580194266762,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001921576267530502,
+      "loss": 1.0475,
+      "step": 2687
+    },
+    {
+      "epoch": 0.12736318407960198,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019215184446564942,
+      "loss": 0.1473,
+      "step": 2688
+    },
+    {
+      "epoch": 0.12741056621653638,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019214606013441885,
+      "loss": 0.7821,
+      "step": 2689
+    },
+    {
+      "epoch": 0.12745794835347074,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001921402737594868,
+      "loss": 0.5792,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1275053304904051,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019213448534098165,
+      "loss": 0.8598,
+      "step": 2691
+    },
+    {
+      "epoch": 0.1275527126273395,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019212869487903174,
+      "loss": 0.1843,
+      "step": 2692
+    },
+    {
+      "epoch": 0.12760009476427386,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001921229023737655,
+      "loss": 0.3087,
+      "step": 2693
+    },
+    {
+      "epoch": 0.12764747690120826,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019211710782531143,
+      "loss": 0.8627,
+      "step": 2694
+    },
+    {
+      "epoch": 0.12769485903814262,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.00019211131123379803,
+      "loss": 0.0027,
+      "step": 2695
+    },
+    {
+      "epoch": 0.127742241175077,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019210551259935384,
+      "loss": 0.7263,
+      "step": 2696
+    },
+    {
+      "epoch": 0.12778962331201138,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019209971192210747,
+      "loss": 1.5114,
+      "step": 2697
+    },
+    {
+      "epoch": 0.12783700544894575,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001920939092021876,
+      "loss": 1.0862,
+      "step": 2698
+    },
+    {
+      "epoch": 0.1278843875858801,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019208810443972296,
+      "loss": 1.0623,
+      "step": 2699
+    },
+    {
+      "epoch": 0.1279317697228145,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019208229763484222,
+      "loss": 1.2576,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12797915185974887,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019207648878767422,
+      "loss": 0.7411,
+      "step": 2701
+    },
+    {
+      "epoch": 0.12802653399668326,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019207067789834775,
+      "loss": 1.1092,
+      "step": 2702
+    },
+    {
+      "epoch": 0.12807391613361763,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001920648649669917,
+      "loss": 1.7907,
+      "step": 2703
+    },
+    {
+      "epoch": 0.128121298270552,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019205904999373502,
+      "loss": 1.2354,
+      "step": 2704
+    },
+    {
+      "epoch": 0.12816868040748639,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00019205323297870667,
+      "loss": 0.4706,
+      "step": 2705
+    },
+    {
+      "epoch": 0.12821606254442075,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019204741392203568,
+      "loss": 1.2389,
+      "step": 2706
+    },
+    {
+      "epoch": 0.12826344468135512,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019204159282385107,
+      "loss": 1.2217,
+      "step": 2707
+    },
+    {
+      "epoch": 0.1283108268182895,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.000192035769684282,
+      "loss": 0.4758,
+      "step": 2708
+    },
+    {
+      "epoch": 0.12835820895522387,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019202994450345757,
+      "loss": 1.1666,
+      "step": 2709
+    },
+    {
+      "epoch": 0.12840559109215827,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019202411728150702,
+      "loss": 1.061,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12845297322909263,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001920182880185595,
+      "loss": 1.5762,
+      "step": 2711
+    },
+    {
+      "epoch": 0.128500355366027,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019201245671474445,
+      "loss": 1.1611,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1285477375029614,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019200662337019107,
+      "loss": 1.3035,
+      "step": 2713
+    },
+    {
+      "epoch": 0.12859511963989576,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001920007879850288,
+      "loss": 0.1655,
+      "step": 2714
+    },
+    {
+      "epoch": 0.12864250177683015,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019199495055938703,
+      "loss": 0.8907,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1286898839137645,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019198911109339525,
+      "loss": 0.8073,
+      "step": 2716
+    },
+    {
+      "epoch": 0.12873726605069888,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000191983269587183,
+      "loss": 0.2654,
+      "step": 2717
+    },
+    {
+      "epoch": 0.12878464818763327,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001919774260408798,
+      "loss": 0.3346,
+      "step": 2718
+    },
+    {
+      "epoch": 0.12883203032456764,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019197158045461524,
+      "loss": 0.1698,
+      "step": 2719
+    },
+    {
+      "epoch": 0.128879412461502,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019196573282851897,
+      "loss": 0.0602,
+      "step": 2720
+    },
+    {
+      "epoch": 0.1289267945984364,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00019195988316272073,
+      "loss": 0.1667,
+      "step": 2721
+    },
+    {
+      "epoch": 0.12897417673537076,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019195403145735025,
+      "loss": 0.5652,
+      "step": 2722
+    },
+    {
+      "epoch": 0.12902155887230515,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019194817771253726,
+      "loss": 0.7821,
+      "step": 2723
+    },
+    {
+      "epoch": 0.12906894100923952,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019194232192841165,
+      "loss": 0.9453,
+      "step": 2724
+    },
+    {
+      "epoch": 0.12911632314617388,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00019193646410510325,
+      "loss": 0.2774,
+      "step": 2725
+    },
+    {
+      "epoch": 0.12916370528310828,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019193060424274204,
+      "loss": 0.0525,
+      "step": 2726
+    },
+    {
+      "epoch": 0.12921108742004264,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0001919247423414579,
+      "loss": 0.1739,
+      "step": 2727
+    },
+    {
+      "epoch": 0.129258469556977,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019191887840138092,
+      "loss": 0.6484,
+      "step": 2728
+    },
+    {
+      "epoch": 0.1293058516939114,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001919130124226411,
+      "loss": 1.1583,
+      "step": 2729
+    },
+    {
+      "epoch": 0.12935323383084577,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001919071444053686,
+      "loss": 0.7938,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12940061596778016,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001919012743496935,
+      "loss": 0.8336,
+      "step": 2731
+    },
+    {
+      "epoch": 0.12944799810471452,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019189540225574608,
+      "loss": 0.5599,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1294953802416489,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019188952812365645,
+      "loss": 1.0495,
+      "step": 2733
+    },
+    {
+      "epoch": 0.12954276237858328,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000191883651953555,
+      "loss": 0.0177,
+      "step": 2734
+    },
+    {
+      "epoch": 0.12959014451551765,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000191877773745572,
+      "loss": 0.4574,
+      "step": 2735
+    },
+    {
+      "epoch": 0.129637526652452,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019187189349983787,
+      "loss": 0.2015,
+      "step": 2736
+    },
+    {
+      "epoch": 0.1296849087893864,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000191866011216483,
+      "loss": 1.0366,
+      "step": 2737
+    },
+    {
+      "epoch": 0.12973229092632077,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019186012689563785,
+      "loss": 0.2846,
+      "step": 2738
+    },
+    {
+      "epoch": 0.12977967306325516,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001918542405374329,
+      "loss": 1.2374,
+      "step": 2739
+    },
+    {
+      "epoch": 0.12982705520018953,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00019184835214199874,
+      "loss": 0.2086,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1298744373371239,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019184246170946597,
+      "loss": 1.0787,
+      "step": 2741
+    },
+    {
+      "epoch": 0.1299218194740583,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019183656923996525,
+      "loss": 0.8791,
+      "step": 2742
+    },
+    {
+      "epoch": 0.12996920161099265,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019183067473362722,
+      "loss": 0.7641,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13001658374792704,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019182477819058264,
+      "loss": 0.735,
+      "step": 2744
+    },
+    {
+      "epoch": 0.1300639658848614,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001918188796109623,
+      "loss": 0.598,
+      "step": 2745
+    },
+    {
+      "epoch": 0.13011134802179578,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019181297899489698,
+      "loss": 0.1809,
+      "step": 2746
+    },
+    {
+      "epoch": 0.13015873015873017,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019180707634251762,
+      "loss": 1.3001,
+      "step": 2747
+    },
+    {
+      "epoch": 0.13020611229566453,
+      "grad_norm": 0.0234375,
+      "learning_rate": 0.0001918011716539551,
+      "loss": 0.0017,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1302534944325989,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00019179526492934032,
+      "loss": 0.2991,
+      "step": 2749
+    },
+    {
+      "epoch": 0.1303008765695333,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019178935616880435,
+      "loss": 0.6896,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13034825870646766,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019178344537247827,
+      "loss": 1.1772,
+      "step": 2751
+    },
+    {
+      "epoch": 0.13039564084340205,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001917775325404931,
+      "loss": 0.8827,
+      "step": 2752
+    },
+    {
+      "epoch": 0.13044302298033641,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00019177161767298004,
+      "loss": 0.3459,
+      "step": 2753
+    },
+    {
+      "epoch": 0.13049040511727078,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019176570077007025,
+      "loss": 0.0461,
+      "step": 2754
+    },
+    {
+      "epoch": 0.13053778725420517,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019175978183189494,
+      "loss": 1.1708,
+      "step": 2755
+    },
+    {
+      "epoch": 0.13058516939113954,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019175386085858542,
+      "loss": 0.1037,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1306325515280739,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000191747937850273,
+      "loss": 1.2245,
+      "step": 2757
+    },
+    {
+      "epoch": 0.1306799336650083,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019174201280708903,
+      "loss": 0.968,
+      "step": 2758
+    },
+    {
+      "epoch": 0.13072731580194266,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019173608572916497,
+      "loss": 0.7292,
+      "step": 2759
+    },
+    {
+      "epoch": 0.13077469793887705,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019173015661663222,
+      "loss": 1.3887,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13082208007581142,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001917242254696223,
+      "loss": 1.1602,
+      "step": 2761
+    },
+    {
+      "epoch": 0.13086946221274579,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019171829228826676,
+      "loss": 1.1013,
+      "step": 2762
+    },
+    {
+      "epoch": 0.13091684434968018,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019171235707269717,
+      "loss": 0.5124,
+      "step": 2763
+    },
+    {
+      "epoch": 0.13096422648661454,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00019170641982304524,
+      "loss": 0.2174,
+      "step": 2764
+    },
+    {
+      "epoch": 0.1310116086235489,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019170048053944255,
+      "loss": 0.9312,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1310589907604833,
+      "grad_norm": 0.0203857421875,
+      "learning_rate": 0.0001916945392220209,
+      "loss": 0.0015,
+      "step": 2766
+    },
+    {
+      "epoch": 0.13110637289741767,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019168859587091205,
+      "loss": 0.5787,
+      "step": 2767
+    },
+    {
+      "epoch": 0.13115375503435206,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019168265048624775,
+      "loss": 0.7287,
+      "step": 2768
+    },
+    {
+      "epoch": 0.13120113717128642,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019167670306815998,
+      "loss": 0.0285,
+      "step": 2769
+    },
+    {
+      "epoch": 0.1312485193082208,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019167075361678054,
+      "loss": 0.8246,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13129590144515518,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019166480213224146,
+      "loss": 0.5416,
+      "step": 2771
+    },
+    {
+      "epoch": 0.13134328358208955,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019165884861467472,
+      "loss": 0.0086,
+      "step": 2772
+    },
+    {
+      "epoch": 0.13139066571902394,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019165289306421232,
+      "loss": 1.7462,
+      "step": 2773
+    },
+    {
+      "epoch": 0.1314380478559583,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001916469354809864,
+      "loss": 1.0287,
+      "step": 2774
+    },
+    {
+      "epoch": 0.13148542999289267,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00019164097586512902,
+      "loss": 0.0097,
+      "step": 2775
+    },
+    {
+      "epoch": 0.13153281212982706,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019163501421677243,
+      "loss": 1.4429,
+      "step": 2776
+    },
+    {
+      "epoch": 0.13158019426676143,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019162905053604886,
+      "loss": 1.0607,
+      "step": 2777
+    },
+    {
+      "epoch": 0.1316275764036958,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019162308482309053,
+      "loss": 0.9388,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1316749585406302,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019161711707802977,
+      "loss": 0.1465,
+      "step": 2779
+    },
+    {
+      "epoch": 0.13172234067756455,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019161114730099893,
+      "loss": 1.1465,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13176972281449895,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00019160517549213042,
+      "loss": 0.1622,
+      "step": 2781
+    },
+    {
+      "epoch": 0.1318171049514333,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019159920165155668,
+      "loss": 1.3999,
+      "step": 2782
+    },
+    {
+      "epoch": 0.13186448708836768,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00019159322577941024,
+      "loss": 0.3647,
+      "step": 2783
+    },
+    {
+      "epoch": 0.13191186922530207,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019158724787582363,
+      "loss": 1.1849,
+      "step": 2784
+    },
+    {
+      "epoch": 0.13195925136223643,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001915812679409294,
+      "loss": 0.7259,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1320066334991708,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019157528597486017,
+      "loss": 0.9717,
+      "step": 2786
+    },
+    {
+      "epoch": 0.1320540156361052,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019156930197774866,
+      "loss": 0.0891,
+      "step": 2787
+    },
+    {
+      "epoch": 0.13210139777303956,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019156331594972757,
+      "loss": 0.2622,
+      "step": 2788
+    },
+    {
+      "epoch": 0.13214877990997395,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019155732789092965,
+      "loss": 1.1889,
+      "step": 2789
+    },
+    {
+      "epoch": 0.13219616204690832,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019155133780148772,
+      "loss": 0.9256,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13224354418384268,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019154534568153464,
+      "loss": 0.9825,
+      "step": 2791
+    },
+    {
+      "epoch": 0.13229092632077707,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001915393515312033,
+      "loss": 1.0119,
+      "step": 2792
+    },
+    {
+      "epoch": 0.13233830845771144,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019153335535062666,
+      "loss": 0.9351,
+      "step": 2793
+    },
+    {
+      "epoch": 0.1323856905946458,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019152735713993767,
+      "loss": 0.7316,
+      "step": 2794
+    },
+    {
+      "epoch": 0.1324330727315802,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001915213568992694,
+      "loss": 0.9324,
+      "step": 2795
+    },
+    {
+      "epoch": 0.13248045486851456,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019151535462875495,
+      "loss": 0.7162,
+      "step": 2796
+    },
+    {
+      "epoch": 0.13252783700544896,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019150935032852736,
+      "loss": 0.7289,
+      "step": 2797
+    },
+    {
+      "epoch": 0.13257521914238332,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001915033439987199,
+      "loss": 1.1233,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1326226012793177,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019149733563946568,
+      "loss": 0.9656,
+      "step": 2799
+    },
+    {
+      "epoch": 0.13266998341625208,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019149132525089806,
+      "loss": 1.158,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13271736555318644,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019148531283315028,
+      "loss": 0.8042,
+      "step": 2801
+    },
+    {
+      "epoch": 0.13276474769012084,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019147929838635573,
+      "loss": 0.2657,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1328121298270552,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019147328191064774,
+      "loss": 0.6868,
+      "step": 2803
+    },
+    {
+      "epoch": 0.13285951196398957,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001914672634061598,
+      "loss": 0.9452,
+      "step": 2804
+    },
+    {
+      "epoch": 0.13290689410092396,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001914612428730254,
+      "loss": 0.5108,
+      "step": 2805
+    },
+    {
+      "epoch": 0.13295427623785833,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019145522031137807,
+      "loss": 0.863,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1330016583747927,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019144919572135135,
+      "loss": 0.0661,
+      "step": 2807
+    },
+    {
+      "epoch": 0.13304904051172708,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019144316910307883,
+      "loss": 0.8761,
+      "step": 2808
+    },
+    {
+      "epoch": 0.13309642264866145,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019143714045669427,
+      "loss": 0.0686,
+      "step": 2809
+    },
+    {
+      "epoch": 0.13314380478559584,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019143110978233133,
+      "loss": 0.0148,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1331911869225302,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019142507708012376,
+      "loss": 0.0215,
+      "step": 2811
+    },
+    {
+      "epoch": 0.13323856905946457,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00019141904235020537,
+      "loss": 0.0066,
+      "step": 2812
+    },
+    {
+      "epoch": 0.13328595119639897,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019141300559270995,
+      "loss": 0.6714,
+      "step": 2813
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019140696680777146,
+      "loss": 0.7004,
+      "step": 2814
+    },
+    {
+      "epoch": 0.1333807154702677,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00019140092599552385,
+      "loss": 1.147,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1334280976072021,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000191394883156101,
+      "loss": 0.0055,
+      "step": 2816
+    },
+    {
+      "epoch": 0.13347547974413645,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.000191388838289637,
+      "loss": 0.4192,
+      "step": 2817
+    },
+    {
+      "epoch": 0.13352286188107085,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019138279139626594,
+      "loss": 0.714,
+      "step": 2818
+    },
+    {
+      "epoch": 0.1335702440180052,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019137674247612186,
+      "loss": 0.732,
+      "step": 2819
+    },
+    {
+      "epoch": 0.13361762615493958,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000191370691529339,
+      "loss": 0.4689,
+      "step": 2820
+    },
+    {
+      "epoch": 0.13366500829187397,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001913646385560515,
+      "loss": 0.087,
+      "step": 2821
+    },
+    {
+      "epoch": 0.13371239042880834,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019135858355639367,
+      "loss": 0.7453,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1337597725657427,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019135252653049975,
+      "loss": 0.992,
+      "step": 2823
+    },
+    {
+      "epoch": 0.1338071547026771,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019134646747850408,
+      "loss": 0.8448,
+      "step": 2824
+    },
+    {
+      "epoch": 0.13385453683961146,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019134040640054106,
+      "loss": 0.6233,
+      "step": 2825
+    },
+    {
+      "epoch": 0.13390191897654585,
+      "grad_norm": 0.01019287109375,
+      "learning_rate": 0.00019133434329674512,
+      "loss": 0.0008,
+      "step": 2826
+    },
+    {
+      "epoch": 0.13394930111348022,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019132827816725073,
+      "loss": 1.0276,
+      "step": 2827
+    },
+    {
+      "epoch": 0.13399668325041458,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019132221101219243,
+      "loss": 1.1445,
+      "step": 2828
+    },
+    {
+      "epoch": 0.13404406538734898,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019131614183170477,
+      "loss": 0.0918,
+      "step": 2829
+    },
+    {
+      "epoch": 0.13409144752428334,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019131007062592233,
+      "loss": 0.8945,
+      "step": 2830
+    },
+    {
+      "epoch": 0.13413882966121773,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019130399739497977,
+      "loss": 0.3067,
+      "step": 2831
+    },
+    {
+      "epoch": 0.1341862117981521,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019129792213901182,
+      "loss": 1.005,
+      "step": 2832
+    },
+    {
+      "epoch": 0.13423359393508646,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0001912918448581532,
+      "loss": 0.0211,
+      "step": 2833
+    },
+    {
+      "epoch": 0.13428097607202086,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00019128576555253868,
+      "loss": 1.0393,
+      "step": 2834
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019127968422230315,
+      "loss": 0.6882,
+      "step": 2835
+    },
+    {
+      "epoch": 0.1343757403458896,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0001912736008675814,
+      "loss": 0.0056,
+      "step": 2836
+    },
+    {
+      "epoch": 0.13442312248282398,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019126751548850844,
+      "loss": 1.6895,
+      "step": 2837
+    },
+    {
+      "epoch": 0.13447050461975835,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019126142808521918,
+      "loss": 0.7938,
+      "step": 2838
+    },
+    {
+      "epoch": 0.13451788675669274,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019125533865784868,
+      "loss": 1.026,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1345652688936271,
+      "grad_norm": 0.0155029296875,
+      "learning_rate": 0.00019124924720653196,
+      "loss": 0.0009,
+      "step": 2840
+    },
+    {
+      "epoch": 0.13461265103056147,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001912431537314041,
+      "loss": 0.8351,
+      "step": 2841
+    },
+    {
+      "epoch": 0.13466003316749586,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019123705823260033,
+      "loss": 0.245,
+      "step": 2842
+    },
+    {
+      "epoch": 0.13470741530443023,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019123096071025576,
+      "loss": 0.8535,
+      "step": 2843
+    },
+    {
+      "epoch": 0.1347547974413646,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019122486116450566,
+      "loss": 0.469,
+      "step": 2844
+    },
+    {
+      "epoch": 0.134802179578299,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001912187595954853,
+      "loss": 0.3835,
+      "step": 2845
+    },
+    {
+      "epoch": 0.13484956171523335,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019121265600333004,
+      "loss": 0.022,
+      "step": 2846
+    },
+    {
+      "epoch": 0.13489694385216774,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001912065503881752,
+      "loss": 0.1573,
+      "step": 2847
+    },
+    {
+      "epoch": 0.1349443259891021,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019120044275015624,
+      "loss": 0.1412,
+      "step": 2848
+    },
+    {
+      "epoch": 0.13499170812603647,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019119433308940857,
+      "loss": 1.1661,
+      "step": 2849
+    },
+    {
+      "epoch": 0.13503909026297087,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019118822140606776,
+      "loss": 1.1511,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13508647239990523,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001911821077002693,
+      "loss": 0.3723,
+      "step": 2851
+    },
+    {
+      "epoch": 0.1351338545368396,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00019117599197214884,
+      "loss": 0.1053,
+      "step": 2852
+    },
+    {
+      "epoch": 0.135181236673774,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00019116987422184198,
+      "loss": 0.1539,
+      "step": 2853
+    },
+    {
+      "epoch": 0.13522861881070836,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001911637544494844,
+      "loss": 0.8968,
+      "step": 2854
+    },
+    {
+      "epoch": 0.13527600094764275,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001911576326552119,
+      "loss": 0.7788,
+      "step": 2855
+    },
+    {
+      "epoch": 0.13532338308457711,
+      "grad_norm": 0.002777099609375,
+      "learning_rate": 0.00019115150883916018,
+      "loss": 0.0002,
+      "step": 2856
+    },
+    {
+      "epoch": 0.13537076522151148,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019114538300146508,
+      "loss": 0.1204,
+      "step": 2857
+    },
+    {
+      "epoch": 0.13541814735844587,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019113925514226246,
+      "loss": 0.0465,
+      "step": 2858
+    },
+    {
+      "epoch": 0.13546552949538024,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019113312526168827,
+      "loss": 0.9245,
+      "step": 2859
+    },
+    {
+      "epoch": 0.13551291163231463,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019112699335987842,
+      "loss": 0.3414,
+      "step": 2860
+    },
+    {
+      "epoch": 0.135560293769249,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019112085943696888,
+      "loss": 0.7549,
+      "step": 2861
+    },
+    {
+      "epoch": 0.13560767590618336,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019111472349309577,
+      "loss": 0.6086,
+      "step": 2862
+    },
+    {
+      "epoch": 0.13565505804311775,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019110858552839517,
+      "loss": 0.5047,
+      "step": 2863
+    },
+    {
+      "epoch": 0.13570244018005212,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019110244554300314,
+      "loss": 0.9906,
+      "step": 2864
+    },
+    {
+      "epoch": 0.13574982231698648,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00019109630353705592,
+      "loss": 0.1466,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13579720445392088,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019109015951068975,
+      "loss": 0.0174,
+      "step": 2866
+    },
+    {
+      "epoch": 0.13584458659085524,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019108401346404084,
+      "loss": 0.0408,
+      "step": 2867
+    },
+    {
+      "epoch": 0.13589196872778964,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019107786539724554,
+      "loss": 1.2381,
+      "step": 2868
+    },
+    {
+      "epoch": 0.135939350864724,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019107171531044018,
+      "loss": 0.967,
+      "step": 2869
+    },
+    {
+      "epoch": 0.13598673300165837,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001910655632037612,
+      "loss": 0.0303,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13603411513859276,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019105940907734498,
+      "loss": 0.6367,
+      "step": 2871
+    },
+    {
+      "epoch": 0.13608149727552712,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019105325293132812,
+      "loss": 0.944,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1361288794124615,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00019104709476584707,
+      "loss": 1.4161,
+      "step": 2873
+    },
+    {
+      "epoch": 0.13617626154939588,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019104093458103841,
+      "loss": 0.6082,
+      "step": 2874
+    },
+    {
+      "epoch": 0.13622364368633025,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019103477237703885,
+      "loss": 1.1064,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13627102582326464,
+      "grad_norm": 0.02099609375,
+      "learning_rate": 0.00019102860815398495,
+      "loss": 0.0011,
+      "step": 2876
+    },
+    {
+      "epoch": 0.136318407960199,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001910224419120135,
+      "loss": 0.8008,
+      "step": 2877
+    },
+    {
+      "epoch": 0.13636579009713337,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019101627365126125,
+      "loss": 0.0692,
+      "step": 2878
+    },
+    {
+      "epoch": 0.13641317223406776,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.000191010103371865,
+      "loss": 0.0145,
+      "step": 2879
+    },
+    {
+      "epoch": 0.13646055437100213,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019100393107396158,
+      "loss": 0.2141,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1365079365079365,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00019099775675768791,
+      "loss": 0.0824,
+      "step": 2881
+    },
+    {
+      "epoch": 0.1365553186448709,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019099158042318096,
+      "loss": 0.1288,
+      "step": 2882
+    },
+    {
+      "epoch": 0.13660270078180525,
+      "grad_norm": 0.00469970703125,
+      "learning_rate": 0.0001909854020705776,
+      "loss": 0.0003,
+      "step": 2883
+    },
+    {
+      "epoch": 0.13665008291873965,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019097922170001498,
+      "loss": 0.8375,
+      "step": 2884
+    },
+    {
+      "epoch": 0.136697465055674,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019097303931163014,
+      "loss": 1.0394,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13674484719260838,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019096685490556018,
+      "loss": 0.7349,
+      "step": 2886
+    },
+    {
+      "epoch": 0.13679222932954277,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019096066848194225,
+      "loss": 1.3313,
+      "step": 2887
+    },
+    {
+      "epoch": 0.13683961146647713,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019095448004091358,
+      "loss": 1.1952,
+      "step": 2888
+    },
+    {
+      "epoch": 0.13688699360341153,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.00019094828958261145,
+      "loss": 0.002,
+      "step": 2889
+    },
+    {
+      "epoch": 0.1369343757403459,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001909420971071731,
+      "loss": 1.0378,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13698175787728026,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019093590261473592,
+      "loss": 1.2131,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13702914001421465,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001909297061054373,
+      "loss": 0.2519,
+      "step": 2892
+    },
+    {
+      "epoch": 0.13707652215114902,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019092350757941463,
+      "loss": 0.1355,
+      "step": 2893
+    },
+    {
+      "epoch": 0.13712390428808338,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001909173070368054,
+      "loss": 0.0684,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13717128642501777,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001909111044777472,
+      "loss": 1.0816,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13721866856195214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019090489990237746,
+      "loss": 0.1926,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13726605069888653,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001908986933108339,
+      "loss": 0.7542,
+      "step": 2897
+    },
+    {
+      "epoch": 0.1373134328358209,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019089248470325414,
+      "loss": 1.1729,
+      "step": 2898
+    },
+    {
+      "epoch": 0.13736081497275526,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019088627407977588,
+      "loss": 1.0235,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13740819710968966,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019088006144053686,
+      "loss": 1.096,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13745557924662402,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001908738467856749,
+      "loss": 0.1479,
+      "step": 2901
+    },
+    {
+      "epoch": 0.1375029613835584,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001908676301153278,
+      "loss": 0.6035,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13755034352049278,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001908614114296335,
+      "loss": 0.0202,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13759772565742714,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0001908551907287298,
+      "loss": 0.1763,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13764510779436154,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019084896801275478,
+      "loss": 1.3041,
+      "step": 2905
+    },
+    {
+      "epoch": 0.1376924899312959,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001908427432818464,
+      "loss": 1.2016,
+      "step": 2906
+    },
+    {
+      "epoch": 0.13773987206823027,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019083651653614277,
+      "loss": 1.1691,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13778725420516466,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019083028777578192,
+      "loss": 0.8341,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13783463634209903,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019082405700090207,
+      "loss": 0.1651,
+      "step": 2909
+    },
+    {
+      "epoch": 0.1378820184790334,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019081782421164137,
+      "loss": 1.4576,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13792940061596778,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019081158940813806,
+      "loss": 0.9515,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13797678275290215,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001908053525905304,
+      "loss": 0.2158,
+      "step": 2912
+    },
+    {
+      "epoch": 0.13802416488983654,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019079911375895675,
+      "loss": 1.3892,
+      "step": 2913
+    },
+    {
+      "epoch": 0.1380715470267709,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001907928729135555,
+      "loss": 0.9338,
+      "step": 2914
+    },
+    {
+      "epoch": 0.13811892916370527,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.000190786630054465,
+      "loss": 0.0649,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13816631130063967,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019078038518182376,
+      "loss": 0.9949,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13821369343757403,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001907741382957703,
+      "loss": 1.1079,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13826107557450842,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019076788939644313,
+      "loss": 0.1373,
+      "step": 2918
+    },
+    {
+      "epoch": 0.1383084577114428,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001907616384839808,
+      "loss": 0.8246,
+      "step": 2919
+    },
+    {
+      "epoch": 0.13835583984837715,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019075538555852207,
+      "loss": 1.201,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13840322198531155,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019074913062020553,
+      "loss": 0.1198,
+      "step": 2921
+    },
+    {
+      "epoch": 0.1384506041222459,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019074287366916995,
+      "loss": 0.9153,
+      "step": 2922
+    },
+    {
+      "epoch": 0.13849798625918028,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019073661470555406,
+      "loss": 1.2473,
+      "step": 2923
+    },
+    {
+      "epoch": 0.13854536839611467,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019073035372949671,
+      "loss": 0.9549,
+      "step": 2924
+    },
+    {
+      "epoch": 0.13859275053304904,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00019072409074113677,
+      "loss": 0.7476,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13864013266998343,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019071782574061312,
+      "loss": 0.6304,
+      "step": 2926
+    },
+    {
+      "epoch": 0.1386875148069178,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019071155872806475,
+      "loss": 1.4026,
+      "step": 2927
+    },
+    {
+      "epoch": 0.13873489694385216,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001907052897036306,
+      "loss": 0.1602,
+      "step": 2928
+    },
+    {
+      "epoch": 0.13878227908078655,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019069901866744976,
+      "loss": 0.7066,
+      "step": 2929
+    },
+    {
+      "epoch": 0.13882966121772092,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019069274561966127,
+      "loss": 0.9224,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13887704335465528,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001906864705604043,
+      "loss": 1.169,
+      "step": 2931
+    },
+    {
+      "epoch": 0.13892442549158968,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.000190680193489818,
+      "loss": 0.5324,
+      "step": 2932
+    },
+    {
+      "epoch": 0.13897180762852404,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019067391440804162,
+      "loss": 0.424,
+      "step": 2933
+    },
+    {
+      "epoch": 0.13901918976545843,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 0.00019066763331521437,
+      "loss": 0.0024,
+      "step": 2934
+    },
+    {
+      "epoch": 0.1390665719023928,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001906613502114756,
+      "loss": 0.7701,
+      "step": 2935
+    },
+    {
+      "epoch": 0.13911395403932716,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019065506509696465,
+      "loss": 1.1735,
+      "step": 2936
+    },
+    {
+      "epoch": 0.13916133617626156,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001906487779718209,
+      "loss": 1.383,
+      "step": 2937
+    },
+    {
+      "epoch": 0.13920871831319592,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019064248883618383,
+      "loss": 0.4393,
+      "step": 2938
+    },
+    {
+      "epoch": 0.1392561004501303,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001906361976901929,
+      "loss": 0.4204,
+      "step": 2939
+    },
+    {
+      "epoch": 0.13930348258706468,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019062990453398766,
+      "loss": 1.2174,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13935086472399905,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019062360936770767,
+      "loss": 0.6505,
+      "step": 2941
+    },
+    {
+      "epoch": 0.13939824686093344,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00019061731219149251,
+      "loss": 1.3342,
+      "step": 2942
+    },
+    {
+      "epoch": 0.1394456289978678,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019061101300548195,
+      "loss": 1.5339,
+      "step": 2943
+    },
+    {
+      "epoch": 0.13949301113480217,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00019060471180981558,
+      "loss": 1.3235,
+      "step": 2944
+    },
+    {
+      "epoch": 0.13954039327173656,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019059840860463325,
+      "loss": 1.2126,
+      "step": 2945
+    },
+    {
+      "epoch": 0.13958777540867093,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001905921033900747,
+      "loss": 1.1159,
+      "step": 2946
+    },
+    {
+      "epoch": 0.13963515754560532,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001905857961662798,
+      "loss": 0.1903,
+      "step": 2947
+    },
+    {
+      "epoch": 0.13968253968253969,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019057948693338843,
+      "loss": 0.8658,
+      "step": 2948
+    },
+    {
+      "epoch": 0.13972992181947405,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001905731756915405,
+      "loss": 0.0776,
+      "step": 2949
+    },
+    {
+      "epoch": 0.13977730395640844,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019056686244087602,
+      "loss": 0.8222,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1398246860933428,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.000190560547181535,
+      "loss": 1.0425,
+      "step": 2951
+    },
+    {
+      "epoch": 0.13987206823027717,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019055422991365754,
+      "loss": 0.9886,
+      "step": 2952
+    },
+    {
+      "epoch": 0.13991945036721157,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019054791063738365,
+      "loss": 0.6058,
+      "step": 2953
+    },
+    {
+      "epoch": 0.13996683250414593,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019054158935285357,
+      "loss": 0.2055,
+      "step": 2954
+    },
+    {
+      "epoch": 0.14001421464108033,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001905352660602075,
+      "loss": 1.2486,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1400615967780147,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00019052894075958566,
+      "loss": 0.1538,
+      "step": 2956
+    },
+    {
+      "epoch": 0.14010897891494906,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001905226134511283,
+      "loss": 0.6826,
+      "step": 2957
+    },
+    {
+      "epoch": 0.14015636105188345,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019051628413497583,
+      "loss": 1.2466,
+      "step": 2958
+    },
+    {
+      "epoch": 0.14020374318881781,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001905099528112686,
+      "loss": 0.7137,
+      "step": 2959
+    },
+    {
+      "epoch": 0.14025112532575218,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019050361948014702,
+      "loss": 0.0606,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14029850746268657,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019049728414175153,
+      "loss": 0.8559,
+      "step": 2961
+    },
+    {
+      "epoch": 0.14034588959962094,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019049094679622267,
+      "loss": 0.6058,
+      "step": 2962
+    },
+    {
+      "epoch": 0.14039327173655533,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019048460744370104,
+      "loss": 0.1514,
+      "step": 2963
+    },
+    {
+      "epoch": 0.1404406538734897,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019047826608432718,
+      "loss": 1.0782,
+      "step": 2964
+    },
+    {
+      "epoch": 0.14048803601042406,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019047192271824174,
+      "loss": 0.1652,
+      "step": 2965
+    },
+    {
+      "epoch": 0.14053541814735845,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019046557734558542,
+      "loss": 0.8274,
+      "step": 2966
+    },
+    {
+      "epoch": 0.14058280028429282,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019045922996649897,
+      "loss": 0.9063,
+      "step": 2967
+    },
+    {
+      "epoch": 0.14063018242122718,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019045288058112313,
+      "loss": 1.4254,
+      "step": 2968
+    },
+    {
+      "epoch": 0.14067756455816158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019044652918959876,
+      "loss": 0.1689,
+      "step": 2969
+    },
+    {
+      "epoch": 0.14072494669509594,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019044017579206672,
+      "loss": 0.1385,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14077232883203034,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019043382038866789,
+      "loss": 0.9092,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1408197109689647,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00019042746297954328,
+      "loss": 0.1566,
+      "step": 2972
+    },
+    {
+      "epoch": 0.14086709310589907,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019042110356483382,
+      "loss": 1.2021,
+      "step": 2973
+    },
+    {
+      "epoch": 0.14091447524283346,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001904147421446806,
+      "loss": 1.2113,
+      "step": 2974
+    },
+    {
+      "epoch": 0.14096185737976782,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019040837871922475,
+      "loss": 1.3793,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1410092395167022,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00019040201328860733,
+      "loss": 0.2159,
+      "step": 2976
+    },
+    {
+      "epoch": 0.14105662165363658,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019039564585296957,
+      "loss": 0.1552,
+      "step": 2977
+    },
+    {
+      "epoch": 0.14110400379057095,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019038927641245264,
+      "loss": 1.4,
+      "step": 2978
+    },
+    {
+      "epoch": 0.14115138592750534,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001903829049671978,
+      "loss": 0.9316,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1411987680644397,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019037653151734646,
+      "loss": 0.8791,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14124615020137407,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019037015606303991,
+      "loss": 1.2623,
+      "step": 2981
+    },
+    {
+      "epoch": 0.14129353233830846,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019036377860441955,
+      "loss": 0.7812,
+      "step": 2982
+    },
+    {
+      "epoch": 0.14134091447524283,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019035739914162683,
+      "loss": 0.2049,
+      "step": 2983
+    },
+    {
+      "epoch": 0.14138829661217722,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019035101767480322,
+      "loss": 1.0255,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1414356787491116,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001903446342040903,
+      "loss": 0.916,
+      "step": 2985
+    },
+    {
+      "epoch": 0.14148306088604595,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00019033824872962963,
+      "loss": 1.4644,
+      "step": 2986
+    },
+    {
+      "epoch": 0.14153044302298035,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019033186125156282,
+      "loss": 1.173,
+      "step": 2987
+    },
+    {
+      "epoch": 0.1415778251599147,
+      "grad_norm": 0.016845703125,
+      "learning_rate": 0.00019032547177003152,
+      "loss": 0.0009,
+      "step": 2988
+    },
+    {
+      "epoch": 0.14162520729684908,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001903190802851775,
+      "loss": 0.8895,
+      "step": 2989
+    },
+    {
+      "epoch": 0.14167258943378347,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001903126867971425,
+      "loss": 0.7037,
+      "step": 2990
+    },
+    {
+      "epoch": 0.14171997157071783,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019030629130606825,
+      "loss": 0.5518,
+      "step": 2991
+    },
+    {
+      "epoch": 0.14176735370765223,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019029989381209669,
+      "loss": 0.968,
+      "step": 2992
+    },
+    {
+      "epoch": 0.1418147358445866,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019029349431536963,
+      "loss": 0.8648,
+      "step": 2993
+    },
+    {
+      "epoch": 0.14186211798152096,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001902870928160291,
+      "loss": 0.2046,
+      "step": 2994
+    },
+    {
+      "epoch": 0.14190950011845535,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019028068931421695,
+      "loss": 1.3475,
+      "step": 2995
+    },
+    {
+      "epoch": 0.14195688225538972,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019027428381007534,
+      "loss": 0.3664,
+      "step": 2996
+    },
+    {
+      "epoch": 0.14200426439232408,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019026787630374621,
+      "loss": 1.3609,
+      "step": 2997
+    },
+    {
+      "epoch": 0.14205164652925847,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019026146679537175,
+      "loss": 1.3,
+      "step": 2998
+    },
+    {
+      "epoch": 0.14209902866619284,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019025505528509414,
+      "loss": 0.9183,
+      "step": 2999
+    },
+    {
+      "epoch": 0.14214641080312723,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019024864177305547,
+      "loss": 0.1665,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1421937929400616,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019024222625939812,
+      "loss": 0.9524,
+      "step": 3001
+    },
+    {
+      "epoch": 0.14224117507699596,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001902358087442643,
+      "loss": 0.9756,
+      "step": 3002
+    },
+    {
+      "epoch": 0.14228855721393036,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019022938922779633,
+      "loss": 0.3226,
+      "step": 3003
+    },
+    {
+      "epoch": 0.14233593935086472,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019022296771013664,
+      "loss": 0.9146,
+      "step": 3004
+    },
+    {
+      "epoch": 0.14238332148779909,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001902165441914276,
+      "loss": 1.1521,
+      "step": 3005
+    },
+    {
+      "epoch": 0.14243070362473348,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001902101186718117,
+      "loss": 0.9258,
+      "step": 3006
+    },
+    {
+      "epoch": 0.14247808576166784,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001902036911514315,
+      "loss": 1.0645,
+      "step": 3007
+    },
+    {
+      "epoch": 0.14252546789860224,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001901972616304295,
+      "loss": 1.4107,
+      "step": 3008
+    },
+    {
+      "epoch": 0.1425728500355366,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001901908301089483,
+      "loss": 1.3563,
+      "step": 3009
+    },
+    {
+      "epoch": 0.14262023217247097,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00019018439658713055,
+      "loss": 0.1836,
+      "step": 3010
+    },
+    {
+      "epoch": 0.14266761430940536,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019017796106511893,
+      "loss": 0.0152,
+      "step": 3011
+    },
+    {
+      "epoch": 0.14271499644633973,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001901715235430562,
+      "loss": 1.1516,
+      "step": 3012
+    },
+    {
+      "epoch": 0.14276237858327412,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019016508402108513,
+      "loss": 1.2557,
+      "step": 3013
+    },
+    {
+      "epoch": 0.14280976072020848,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019015864249934854,
+      "loss": 1.0156,
+      "step": 3014
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019015219897798927,
+      "loss": 1.5077,
+      "step": 3015
+    },
+    {
+      "epoch": 0.14290452499407724,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019014575345715028,
+      "loss": 0.8094,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1429519071310116,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001901393059369745,
+      "loss": 1.1134,
+      "step": 3017
+    },
+    {
+      "epoch": 0.14299928926794597,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001901328564176049,
+      "loss": 0.5888,
+      "step": 3018
+    },
+    {
+      "epoch": 0.14304667140488037,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019012640489918456,
+      "loss": 1.2123,
+      "step": 3019
+    },
+    {
+      "epoch": 0.14309405354181473,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019011995138185656,
+      "loss": 0.0636,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14314143567874912,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019011349586576403,
+      "loss": 0.9244,
+      "step": 3021
+    },
+    {
+      "epoch": 0.1431888178156835,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019010703835105016,
+      "loss": 1.1969,
+      "step": 3022
+    },
+    {
+      "epoch": 0.14323619995261785,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001901005788378581,
+      "loss": 0.0429,
+      "step": 3023
+    },
+    {
+      "epoch": 0.14328358208955225,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00019009411732633124,
+      "loss": 0.128,
+      "step": 3024
+    },
+    {
+      "epoch": 0.1433309642264866,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001900876538166128,
+      "loss": 1.0117,
+      "step": 3025
+    },
+    {
+      "epoch": 0.14337834636342098,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019008118830884615,
+      "loss": 0.1456,
+      "step": 3026
+    },
+    {
+      "epoch": 0.14342572850035537,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001900747208031747,
+      "loss": 1.0432,
+      "step": 3027
+    },
+    {
+      "epoch": 0.14347311063728974,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019006825129974186,
+      "loss": 0.7482,
+      "step": 3028
+    },
+    {
+      "epoch": 0.14352049277422413,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019006177979869116,
+      "loss": 1.1897,
+      "step": 3029
+    },
+    {
+      "epoch": 0.1435678749111585,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019005530630016612,
+      "loss": 0.0816,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14361525704809286,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019004883080431034,
+      "loss": 1.1493,
+      "step": 3031
+    },
+    {
+      "epoch": 0.14366263918502725,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019004235331126736,
+      "loss": 0.0867,
+      "step": 3032
+    },
+    {
+      "epoch": 0.14371002132196162,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019003587382118088,
+      "loss": 1.1148,
+      "step": 3033
+    },
+    {
+      "epoch": 0.14375740345889598,
+      "grad_norm": 0.0164794921875,
+      "learning_rate": 0.00019002939233419465,
+      "loss": 0.0006,
+      "step": 3034
+    },
+    {
+      "epoch": 0.14380478559583038,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019002290885045242,
+      "loss": 0.1768,
+      "step": 3035
+    },
+    {
+      "epoch": 0.14385216773276474,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019001642337009792,
+      "loss": 0.8938,
+      "step": 3036
+    },
+    {
+      "epoch": 0.14389954986969913,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019000993589327503,
+      "loss": 0.9925,
+      "step": 3037
+    },
+    {
+      "epoch": 0.1439469320066335,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019000344642012765,
+      "loss": 0.9377,
+      "step": 3038
+    },
+    {
+      "epoch": 0.14399431414356786,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001899969549507997,
+      "loss": 1.0997,
+      "step": 3039
+    },
+    {
+      "epoch": 0.14404169628050226,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018999046148543514,
+      "loss": 0.2018,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14408907841743662,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000189983966024178,
+      "loss": 0.7018,
+      "step": 3041
+    },
+    {
+      "epoch": 0.14413646055437102,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018997746856717234,
+      "loss": 0.0291,
+      "step": 3042
+    },
+    {
+      "epoch": 0.14418384269130538,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018997096911456228,
+      "loss": 0.0811,
+      "step": 3043
+    },
+    {
+      "epoch": 0.14423122482823975,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018996446766649192,
+      "loss": 0.0217,
+      "step": 3044
+    },
+    {
+      "epoch": 0.14427860696517414,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018995796422310553,
+      "loss": 0.9543,
+      "step": 3045
+    },
+    {
+      "epoch": 0.1443259891021085,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001899514587845473,
+      "loss": 0.0619,
+      "step": 3046
+    },
+    {
+      "epoch": 0.14437337123904287,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018994495135096152,
+      "loss": 0.1531,
+      "step": 3047
+    },
+    {
+      "epoch": 0.14442075337597726,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018993844192249256,
+      "loss": 0.0108,
+      "step": 3048
+    },
+    {
+      "epoch": 0.14446813551291163,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001899319304992847,
+      "loss": 1.1234,
+      "step": 3049
+    },
+    {
+      "epoch": 0.14451551764984602,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018992541708148246,
+      "loss": 1.077,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14456289978678039,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001899189016692302,
+      "loss": 0.5022,
+      "step": 3051
+    },
+    {
+      "epoch": 0.14461028192371475,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018991238426267252,
+      "loss": 0.6068,
+      "step": 3052
+    },
+    {
+      "epoch": 0.14465766406064914,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018990586486195392,
+      "loss": 1.5448,
+      "step": 3053
+    },
+    {
+      "epoch": 0.1447050461975835,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00018989934346721901,
+      "loss": 0.582,
+      "step": 3054
+    },
+    {
+      "epoch": 0.14475242833451787,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001898928200786124,
+      "loss": 0.877,
+      "step": 3055
+    },
+    {
+      "epoch": 0.14479981047145227,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001898862946962788,
+      "loss": 1.1089,
+      "step": 3056
+    },
+    {
+      "epoch": 0.14484719260838663,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018987976732036293,
+      "loss": 0.8184,
+      "step": 3057
+    },
+    {
+      "epoch": 0.14489457474532103,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00018987323795100955,
+      "loss": 0.9716,
+      "step": 3058
+    },
+    {
+      "epoch": 0.1449419568822554,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018986670658836346,
+      "loss": 0.8909,
+      "step": 3059
+    },
+    {
+      "epoch": 0.14498933901918976,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001898601732325696,
+      "loss": 1.1241,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14503672115612415,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018985363788377274,
+      "loss": 0.8768,
+      "step": 3061
+    },
+    {
+      "epoch": 0.14508410329305851,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.00018984710054211794,
+      "loss": 0.0018,
+      "step": 3062
+    },
+    {
+      "epoch": 0.14513148542999288,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018984056120775015,
+      "loss": 1.1582,
+      "step": 3063
+    },
+    {
+      "epoch": 0.14517886756692727,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001898340198808144,
+      "loss": 1.1511,
+      "step": 3064
+    },
+    {
+      "epoch": 0.14522624970386164,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001898274765614558,
+      "loss": 0.1091,
+      "step": 3065
+    },
+    {
+      "epoch": 0.14527363184079603,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018982093124981941,
+      "loss": 0.0181,
+      "step": 3066
+    },
+    {
+      "epoch": 0.1453210139777304,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018981438394605044,
+      "loss": 1.2479,
+      "step": 3067
+    },
+    {
+      "epoch": 0.14536839611466476,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018980783465029412,
+      "loss": 0.6365,
+      "step": 3068
+    },
+    {
+      "epoch": 0.14541577825159915,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001898012833626957,
+      "loss": 1.0946,
+      "step": 3069
+    },
+    {
+      "epoch": 0.14546316038853352,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018979473008340044,
+      "loss": 1.1289,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1455105425254679,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001897881748125537,
+      "loss": 0.1688,
+      "step": 3071
+    },
+    {
+      "epoch": 0.14555792466240228,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018978161755030094,
+      "loss": 1.2244,
+      "step": 3072
+    },
+    {
+      "epoch": 0.14560530679933664,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018977505829678747,
+      "loss": 1.4977,
+      "step": 3073
+    },
+    {
+      "epoch": 0.14565268893627104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018976849705215883,
+      "loss": 1.0672,
+      "step": 3074
+    },
+    {
+      "epoch": 0.1457000710732054,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001897619338165606,
+      "loss": 0.1644,
+      "step": 3075
+    },
+    {
+      "epoch": 0.14574745321013977,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018975536859013826,
+      "loss": 1.2112,
+      "step": 3076
+    },
+    {
+      "epoch": 0.14579483534707416,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018974880137303743,
+      "loss": 1.2029,
+      "step": 3077
+    },
+    {
+      "epoch": 0.14584221748400852,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018974223216540378,
+      "loss": 0.9227,
+      "step": 3078
+    },
+    {
+      "epoch": 0.14588959962094292,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018973566096738305,
+      "loss": 1.0532,
+      "step": 3079
+    },
+    {
+      "epoch": 0.14593698175787728,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001897290877791209,
+      "loss": 0.6377,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14598436389481165,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018972251260076316,
+      "loss": 1.3745,
+      "step": 3081
+    },
+    {
+      "epoch": 0.14603174603174604,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001897159354324557,
+      "loss": 0.8157,
+      "step": 3082
+    },
+    {
+      "epoch": 0.1460791281686804,
+      "grad_norm": 0.01007080078125,
+      "learning_rate": 0.00018970935627434432,
+      "loss": 0.0007,
+      "step": 3083
+    },
+    {
+      "epoch": 0.14612651030561477,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018970277512657497,
+      "loss": 0.8798,
+      "step": 3084
+    },
+    {
+      "epoch": 0.14617389244254916,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018969619198929363,
+      "loss": 0.2542,
+      "step": 3085
+    },
+    {
+      "epoch": 0.14622127457948353,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018968960686264628,
+      "loss": 1.2204,
+      "step": 3086
+    },
+    {
+      "epoch": 0.14626865671641792,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018968301974677898,
+      "loss": 1.1266,
+      "step": 3087
+    },
+    {
+      "epoch": 0.1463160388533523,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018967643064183784,
+      "loss": 0.9517,
+      "step": 3088
+    },
+    {
+      "epoch": 0.14636342099028665,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018966983954796896,
+      "loss": 0.1777,
+      "step": 3089
+    },
+    {
+      "epoch": 0.14641080312722105,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001896632464653186,
+      "loss": 0.0187,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1464581852641554,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001896566513940329,
+      "loss": 1.1612,
+      "step": 3091
+    },
+    {
+      "epoch": 0.14650556740108978,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001896500543342582,
+      "loss": 1.0695,
+      "step": 3092
+    },
+    {
+      "epoch": 0.14655294953802417,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018964345528614077,
+      "loss": 1.1036,
+      "step": 3093
+    },
+    {
+      "epoch": 0.14660033167495853,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000189636854249827,
+      "loss": 1.5296,
+      "step": 3094
+    },
+    {
+      "epoch": 0.14664771381189293,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018963025122546324,
+      "loss": 0.0533,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1466950959488273,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000189623646213196,
+      "loss": 1.1633,
+      "step": 3096
+    },
+    {
+      "epoch": 0.14674247808576166,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018961703921317174,
+      "loss": 0.1973,
+      "step": 3097
+    },
+    {
+      "epoch": 0.14678986022269605,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018961043022553705,
+      "loss": 1.3722,
+      "step": 3098
+    },
+    {
+      "epoch": 0.14683724235963042,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018960381925043842,
+      "loss": 0.1504,
+      "step": 3099
+    },
+    {
+      "epoch": 0.1468846244965648,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018959720628802253,
+      "loss": 0.2417,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14693200663349917,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018959059133843607,
+      "loss": 0.8952,
+      "step": 3101
+    },
+    {
+      "epoch": 0.14697938877043354,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018958397440182567,
+      "loss": 1.1358,
+      "step": 3102
+    },
+    {
+      "epoch": 0.14702677090736793,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018957735547833816,
+      "loss": 1.035,
+      "step": 3103
+    },
+    {
+      "epoch": 0.1470741530443023,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00018957073456812032,
+      "loss": 0.1452,
+      "step": 3104
+    },
+    {
+      "epoch": 0.14712153518123666,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000189564111671319,
+      "loss": 0.9038,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14716891731817106,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018955748678808106,
+      "loss": 0.7211,
+      "step": 3106
+    },
+    {
+      "epoch": 0.14721629945510542,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001895508599185535,
+      "loss": 0.0434,
+      "step": 3107
+    },
+    {
+      "epoch": 0.1472636815920398,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018954423106288322,
+      "loss": 0.996,
+      "step": 3108
+    },
+    {
+      "epoch": 0.14731106372897418,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001895376002212173,
+      "loss": 1.033,
+      "step": 3109
+    },
+    {
+      "epoch": 0.14735844586590854,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018953096739370272,
+      "loss": 1.0574,
+      "step": 3110
+    },
+    {
+      "epoch": 0.14740582800284294,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001895243325804867,
+      "loss": 1.3177,
+      "step": 3111
+    },
+    {
+      "epoch": 0.1474532101397773,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001895176957817163,
+      "loss": 0.1774,
+      "step": 3112
+    },
+    {
+      "epoch": 0.14750059227671167,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001895110569975388,
+      "loss": 0.6744,
+      "step": 3113
+    },
+    {
+      "epoch": 0.14754797441364606,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001895044162281014,
+      "loss": 0.9408,
+      "step": 3114
+    },
+    {
+      "epoch": 0.14759535655058043,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018949777347355138,
+      "loss": 0.8715,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14764273868751482,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018949112873403604,
+      "loss": 0.6799,
+      "step": 3116
+    },
+    {
+      "epoch": 0.14769012082444918,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001894844820097028,
+      "loss": 0.7846,
+      "step": 3117
+    },
+    {
+      "epoch": 0.14773750296138355,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018947783330069908,
+      "loss": 0.7918,
+      "step": 3118
+    },
+    {
+      "epoch": 0.14778488509831794,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001894711826071723,
+      "loss": 1.1212,
+      "step": 3119
+    },
+    {
+      "epoch": 0.1478322672352523,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018946452992927,
+      "loss": 1.0106,
+      "step": 3120
+    },
+    {
+      "epoch": 0.14787964937218667,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018945787526713974,
+      "loss": 0.8408,
+      "step": 3121
+    },
+    {
+      "epoch": 0.14792703150912107,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018945121862092907,
+      "loss": 0.1438,
+      "step": 3122
+    },
+    {
+      "epoch": 0.14797441364605543,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018944455999078566,
+      "loss": 1.0483,
+      "step": 3123
+    },
+    {
+      "epoch": 0.14802179578298982,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018943789937685718,
+      "loss": 0.7446,
+      "step": 3124
+    },
+    {
+      "epoch": 0.1480691779199242,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018943123677929135,
+      "loss": 1.4493,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14811656005685855,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018942457219823594,
+      "loss": 1.049,
+      "step": 3126
+    },
+    {
+      "epoch": 0.14816394219379295,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001894179056338388,
+      "loss": 1.1085,
+      "step": 3127
+    },
+    {
+      "epoch": 0.1482113243307273,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00018941123708624772,
+      "loss": 0.0172,
+      "step": 3128
+    },
+    {
+      "epoch": 0.1482587064676617,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018940456655561064,
+      "loss": 0.0563,
+      "step": 3129
+    },
+    {
+      "epoch": 0.14830608860459607,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00018939789404207556,
+      "loss": 0.3327,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14835347074153044,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00018939121954579036,
+      "loss": 0.8707,
+      "step": 3131
+    },
+    {
+      "epoch": 0.14840085287846483,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018938454306690315,
+      "loss": 0.489,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1484482350153992,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018937786460556196,
+      "loss": 0.1558,
+      "step": 3133
+    },
+    {
+      "epoch": 0.14849561715233356,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018937118416191494,
+      "loss": 1.3409,
+      "step": 3134
+    },
+    {
+      "epoch": 0.14854299928926795,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018936450173611026,
+      "loss": 0.0315,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14859038142620232,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001893578173282961,
+      "loss": 0.1632,
+      "step": 3136
+    },
+    {
+      "epoch": 0.1486377635631367,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00018935113093862074,
+      "loss": 0.031,
+      "step": 3137
+    },
+    {
+      "epoch": 0.14868514570007108,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018934444256723246,
+      "loss": 0.6992,
+      "step": 3138
+    },
+    {
+      "epoch": 0.14873252783700544,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018933775221427964,
+      "loss": 0.453,
+      "step": 3139
+    },
+    {
+      "epoch": 0.14877990997393983,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001893310598799106,
+      "loss": 0.9826,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1488272921108742,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018932436556427383,
+      "loss": 0.1852,
+      "step": 3141
+    },
+    {
+      "epoch": 0.14887467424780856,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018931766926751778,
+      "loss": 1.0186,
+      "step": 3142
+    },
+    {
+      "epoch": 0.14892205638474296,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018931097098979095,
+      "loss": 0.8663,
+      "step": 3143
+    },
+    {
+      "epoch": 0.14896943852167732,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018930427073124187,
+      "loss": 1.1418,
+      "step": 3144
+    },
+    {
+      "epoch": 0.14901682065861172,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018929756849201925,
+      "loss": 0.0588,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14906420279554608,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018929086427227164,
+      "loss": 0.5915,
+      "step": 3146
+    },
+    {
+      "epoch": 0.14911158493248045,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018928415807214778,
+      "loss": 0.9467,
+      "step": 3147
+    },
+    {
+      "epoch": 0.14915896706941484,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001892774498917964,
+      "loss": 0.8641,
+      "step": 3148
+    },
+    {
+      "epoch": 0.1492063492063492,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018927073973136626,
+      "loss": 0.1645,
+      "step": 3149
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018926402759100622,
+      "loss": 1.561,
+      "step": 3150
+    },
+    {
+      "epoch": 0.14930111348021796,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018925731347086512,
+      "loss": 0.0236,
+      "step": 3151
+    },
+    {
+      "epoch": 0.14934849561715233,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018925059737109188,
+      "loss": 0.66,
+      "step": 3152
+    },
+    {
+      "epoch": 0.14939587775408672,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018924387929183546,
+      "loss": 1.2031,
+      "step": 3153
+    },
+    {
+      "epoch": 0.14944325989102109,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018923715923324484,
+      "loss": 0.9212,
+      "step": 3154
+    },
+    {
+      "epoch": 0.14949064202795545,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001892304371954691,
+      "loss": 0.5071,
+      "step": 3155
+    },
+    {
+      "epoch": 0.14953802416488984,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001892237131786573,
+      "loss": 1.53,
+      "step": 3156
+    },
+    {
+      "epoch": 0.1495854063018242,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.00018921698718295856,
+      "loss": 0.003,
+      "step": 3157
+    },
+    {
+      "epoch": 0.1496327884387586,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001892102592085221,
+      "loss": 0.8592,
+      "step": 3158
+    },
+    {
+      "epoch": 0.14968017057569297,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018920352925549708,
+      "loss": 1.1312,
+      "step": 3159
+    },
+    {
+      "epoch": 0.14972755271262733,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018919679732403284,
+      "loss": 0.9752,
+      "step": 3160
+    },
+    {
+      "epoch": 0.14977493484956172,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001891900634142786,
+      "loss": 0.942,
+      "step": 3161
+    },
+    {
+      "epoch": 0.1498223169864961,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001891833275263838,
+      "loss": 1.1579,
+      "step": 3162
+    },
+    {
+      "epoch": 0.14986969912343046,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018917658966049778,
+      "loss": 0.173,
+      "step": 3163
+    },
+    {
+      "epoch": 0.14991708126036485,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018916984981676995,
+      "loss": 1.3253,
+      "step": 3164
+    },
+    {
+      "epoch": 0.1499644633972992,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018916310799534986,
+      "loss": 0.0086,
+      "step": 3165
+    },
+    {
+      "epoch": 0.1500118455342336,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018915636419638703,
+      "loss": 1.1239,
+      "step": 3166
+    },
+    {
+      "epoch": 0.15005922767116797,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018914961842003095,
+      "loss": 1.4062,
+      "step": 3167
+    },
+    {
+      "epoch": 0.15010660980810234,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018914287066643134,
+      "loss": 0.9106,
+      "step": 3168
+    },
+    {
+      "epoch": 0.15015399194503673,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018913612093573778,
+      "loss": 0.9539,
+      "step": 3169
+    },
+    {
+      "epoch": 0.1502013740819711,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001891293692281,
+      "loss": 0.1564,
+      "step": 3170
+    },
+    {
+      "epoch": 0.15024875621890546,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00018912261554366778,
+      "loss": 0.1612,
+      "step": 3171
+    },
+    {
+      "epoch": 0.15029613835583985,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018911585988259084,
+      "loss": 0.8047,
+      "step": 3172
+    },
+    {
+      "epoch": 0.15034352049277422,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018910910224501906,
+      "loss": 1.1706,
+      "step": 3173
+    },
+    {
+      "epoch": 0.1503909026297086,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001891023426311023,
+      "loss": 0.8821,
+      "step": 3174
+    },
+    {
+      "epoch": 0.15043828476664298,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001890955810409905,
+      "loss": 0.4706,
+      "step": 3175
+    },
+    {
+      "epoch": 0.15048566690357734,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018908881747483364,
+      "loss": 0.8414,
+      "step": 3176
+    },
+    {
+      "epoch": 0.15053304904051173,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018908205193278165,
+      "loss": 0.9074,
+      "step": 3177
+    },
+    {
+      "epoch": 0.1505804311774461,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018907528441498465,
+      "loss": 0.1279,
+      "step": 3178
+    },
+    {
+      "epoch": 0.15062781331438047,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018906851492159275,
+      "loss": 0.0764,
+      "step": 3179
+    },
+    {
+      "epoch": 0.15067519545131486,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018906174345275603,
+      "loss": 0.9166,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15072257758824922,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018905497000862474,
+      "loss": 0.9086,
+      "step": 3181
+    },
+    {
+      "epoch": 0.15076995972518362,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000189048194589349,
+      "loss": 0.5789,
+      "step": 3182
+    },
+    {
+      "epoch": 0.15081734186211798,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018904141719507922,
+      "loss": 0.7374,
+      "step": 3183
+    },
+    {
+      "epoch": 0.15086472399905235,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018903463782596563,
+      "loss": 0.9864,
+      "step": 3184
+    },
+    {
+      "epoch": 0.15091210613598674,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018902785648215858,
+      "loss": 0.9219,
+      "step": 3185
+    },
+    {
+      "epoch": 0.1509594882729211,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00018902107316380855,
+      "loss": 0.0079,
+      "step": 3186
+    },
+    {
+      "epoch": 0.1510068704098555,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018901428787106592,
+      "loss": 0.0924,
+      "step": 3187
+    },
+    {
+      "epoch": 0.15105425254678986,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018900750060408118,
+      "loss": 0.8376,
+      "step": 3188
+    },
+    {
+      "epoch": 0.15110163468372423,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018900071136300492,
+      "loss": 1.1438,
+      "step": 3189
+    },
+    {
+      "epoch": 0.15114901682065862,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018899392014798766,
+      "loss": 0.009,
+      "step": 3190
+    },
+    {
+      "epoch": 0.151196398957593,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018898712695918004,
+      "loss": 0.1356,
+      "step": 3191
+    },
+    {
+      "epoch": 0.15124378109452735,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018898033179673274,
+      "loss": 1.1841,
+      "step": 3192
+    },
+    {
+      "epoch": 0.15129116323146174,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018897353466079645,
+      "loss": 1.0475,
+      "step": 3193
+    },
+    {
+      "epoch": 0.1513385453683961,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018896673555152195,
+      "loss": 0.2508,
+      "step": 3194
+    },
+    {
+      "epoch": 0.1513859275053305,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018895993446905998,
+      "loss": 0.8208,
+      "step": 3195
+    },
+    {
+      "epoch": 0.15143330964226487,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00018895313141356143,
+      "loss": 0.0463,
+      "step": 3196
+    },
+    {
+      "epoch": 0.15148069177919923,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018894632638517716,
+      "loss": 0.0668,
+      "step": 3197
+    },
+    {
+      "epoch": 0.15152807391613363,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018893951938405817,
+      "loss": 0.8999,
+      "step": 3198
+    },
+    {
+      "epoch": 0.151575456053068,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018893271041035533,
+      "loss": 1.1003,
+      "step": 3199
+    },
+    {
+      "epoch": 0.15162283819000236,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001889258994642197,
+      "loss": 1.0037,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15167022032693675,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018891908654580238,
+      "loss": 0.8504,
+      "step": 3201
+    },
+    {
+      "epoch": 0.15171760246387112,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018891227165525437,
+      "loss": 0.1001,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1517649846008055,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018890545479272692,
+      "loss": 0.9516,
+      "step": 3203
+    },
+    {
+      "epoch": 0.15181236673773987,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018889863595837118,
+      "loss": 1.1313,
+      "step": 3204
+    },
+    {
+      "epoch": 0.15185974887467424,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018889181515233835,
+      "loss": 1.1998,
+      "step": 3205
+    },
+    {
+      "epoch": 0.15190713101160863,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001888849923747798,
+      "loss": 0.2061,
+      "step": 3206
+    },
+    {
+      "epoch": 0.151954513148543,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018887816762584676,
+      "loss": 0.8141,
+      "step": 3207
+    },
+    {
+      "epoch": 0.15200189528547736,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018887134090569063,
+      "loss": 1.5754,
+      "step": 3208
+    },
+    {
+      "epoch": 0.15204927742241175,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018886451221446283,
+      "loss": 1.2315,
+      "step": 3209
+    },
+    {
+      "epoch": 0.15209665955934612,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018885768155231481,
+      "loss": 1.3574,
+      "step": 3210
+    },
+    {
+      "epoch": 0.1521440416962805,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018885084891939803,
+      "loss": 1.0823,
+      "step": 3211
+    },
+    {
+      "epoch": 0.15219142383321488,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018884401431586408,
+      "loss": 0.7045,
+      "step": 3212
+    },
+    {
+      "epoch": 0.15223880597014924,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018883717774186454,
+      "loss": 0.9636,
+      "step": 3213
+    },
+    {
+      "epoch": 0.15228618810708364,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000188830339197551,
+      "loss": 0.1879,
+      "step": 3214
+    },
+    {
+      "epoch": 0.152333570244018,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018882349868307516,
+      "loss": 0.955,
+      "step": 3215
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018881665619858873,
+      "loss": 0.2655,
+      "step": 3216
+    },
+    {
+      "epoch": 0.15242833451788676,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018880981174424348,
+      "loss": 0.8976,
+      "step": 3217
+    },
+    {
+      "epoch": 0.15247571665482113,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001888029653201912,
+      "loss": 0.0646,
+      "step": 3218
+    },
+    {
+      "epoch": 0.15252309879175552,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00018879611692658373,
+      "loss": 0.2538,
+      "step": 3219
+    },
+    {
+      "epoch": 0.15257048092868988,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018878926656357297,
+      "loss": 1.1655,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15261786306562425,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018878241423131084,
+      "loss": 1.2484,
+      "step": 3221
+    },
+    {
+      "epoch": 0.15266524520255864,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018877555992994935,
+      "loss": 0.7451,
+      "step": 3222
+    },
+    {
+      "epoch": 0.152712627339493,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018876870365964048,
+      "loss": 0.2073,
+      "step": 3223
+    },
+    {
+      "epoch": 0.1527600094764274,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018876184542053633,
+      "loss": 1.1962,
+      "step": 3224
+    },
+    {
+      "epoch": 0.15280739161336176,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000188754985212789,
+      "loss": 1.0728,
+      "step": 3225
+    },
+    {
+      "epoch": 0.15285477375029613,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001887481230365506,
+      "loss": 0.671,
+      "step": 3226
+    },
+    {
+      "epoch": 0.15290215588723052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018874125889197337,
+      "loss": 1.2873,
+      "step": 3227
+    },
+    {
+      "epoch": 0.1529495380241649,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018873439277920957,
+      "loss": 0.7527,
+      "step": 3228
+    },
+    {
+      "epoch": 0.15299692016109925,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018872752469841145,
+      "loss": 0.3701,
+      "step": 3229
+    },
+    {
+      "epoch": 0.15304430229803365,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001887206546497313,
+      "loss": 1.3629,
+      "step": 3230
+    },
+    {
+      "epoch": 0.153091684434968,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001887137826333216,
+      "loss": 0.8888,
+      "step": 3231
+    },
+    {
+      "epoch": 0.1531390665719024,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018870690864933467,
+      "loss": 1.2509,
+      "step": 3232
+    },
+    {
+      "epoch": 0.15318644870883677,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00018870003269792296,
+      "loss": 0.5475,
+      "step": 3233
+    },
+    {
+      "epoch": 0.15323383084577114,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018869315477923904,
+      "loss": 1.0475,
+      "step": 3234
+    },
+    {
+      "epoch": 0.15328121298270553,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00018868627489343538,
+      "loss": 0.1699,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1533285951196399,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00018867939304066463,
+      "loss": 0.0069,
+      "step": 3236
+    },
+    {
+      "epoch": 0.15337597725657426,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001886725092210794,
+      "loss": 0.9158,
+      "step": 3237
+    },
+    {
+      "epoch": 0.15342335939350865,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018866562343483237,
+      "loss": 0.7676,
+      "step": 3238
+    },
+    {
+      "epoch": 0.15347074153044302,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018865873568207623,
+      "loss": 0.9575,
+      "step": 3239
+    },
+    {
+      "epoch": 0.1535181236673774,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001886518459629638,
+      "loss": 0.2172,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15356550580431177,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018864495427764783,
+      "loss": 0.9754,
+      "step": 3241
+    },
+    {
+      "epoch": 0.15361288794124614,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001886380606262812,
+      "loss": 1.0084,
+      "step": 3242
+    },
+    {
+      "epoch": 0.15366027007818053,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018863116500901682,
+      "loss": 1.1972,
+      "step": 3243
+    },
+    {
+      "epoch": 0.1537076522151149,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018862426742600756,
+      "loss": 0.046,
+      "step": 3244
+    },
+    {
+      "epoch": 0.1537550343520493,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001886173678774065,
+      "loss": 0.0177,
+      "step": 3245
+    },
+    {
+      "epoch": 0.15380241648898366,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018861046636336657,
+      "loss": 1.08,
+      "step": 3246
+    },
+    {
+      "epoch": 0.15384979862591802,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001886035628840409,
+      "loss": 0.9977,
+      "step": 3247
+    },
+    {
+      "epoch": 0.15389718076285241,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001885966574395826,
+      "loss": 0.55,
+      "step": 3248
+    },
+    {
+      "epoch": 0.15394456289978678,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018858975003014476,
+      "loss": 1.112,
+      "step": 3249
+    },
+    {
+      "epoch": 0.15399194503672115,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018858284065588065,
+      "loss": 1.4552,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15403932717365554,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018857592931694348,
+      "loss": 1.1829,
+      "step": 3251
+    },
+    {
+      "epoch": 0.1540867093105899,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018856901601348657,
+      "loss": 0.1491,
+      "step": 3252
+    },
+    {
+      "epoch": 0.1541340914475243,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001885621007456632,
+      "loss": 0.7196,
+      "step": 3253
+    },
+    {
+      "epoch": 0.15418147358445866,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018855518351362677,
+      "loss": 1.1786,
+      "step": 3254
+    },
+    {
+      "epoch": 0.15422885572139303,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001885482643175307,
+      "loss": 0.1435,
+      "step": 3255
+    },
+    {
+      "epoch": 0.15427623785832742,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018854134315752846,
+      "loss": 0.7636,
+      "step": 3256
+    },
+    {
+      "epoch": 0.15432361999526178,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018853442003377352,
+      "loss": 0.7926,
+      "step": 3257
+    },
+    {
+      "epoch": 0.15437100213219615,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018852749494641943,
+      "loss": 1.0515,
+      "step": 3258
+    },
+    {
+      "epoch": 0.15441838426913054,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018852056789561982,
+      "loss": 0.7984,
+      "step": 3259
+    },
+    {
+      "epoch": 0.1544657664060649,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018851363888152832,
+      "loss": 1.0246,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1545131485429993,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018850670790429854,
+      "loss": 0.9434,
+      "step": 3261
+    },
+    {
+      "epoch": 0.15456053067993367,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001884997749640843,
+      "loss": 0.8916,
+      "step": 3262
+    },
+    {
+      "epoch": 0.15460791281686803,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001884928400610393,
+      "loss": 0.81,
+      "step": 3263
+    },
+    {
+      "epoch": 0.15465529495380242,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018848590319531735,
+      "loss": 1.3236,
+      "step": 3264
+    },
+    {
+      "epoch": 0.1547026770907368,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018847896436707234,
+      "loss": 1.135,
+      "step": 3265
+    },
+    {
+      "epoch": 0.15475005922767116,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018847202357645814,
+      "loss": 0.0196,
+      "step": 3266
+    },
+    {
+      "epoch": 0.15479744136460555,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018846508082362865,
+      "loss": 0.8033,
+      "step": 3267
+    },
+    {
+      "epoch": 0.1548448235015399,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018845813610873796,
+      "loss": 0.871,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1548922056384743,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018845118943194,
+      "loss": 0.0952,
+      "step": 3269
+    },
+    {
+      "epoch": 0.15493958777540867,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001884442407933889,
+      "loss": 1.1118,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15498696991234304,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001884372901932387,
+      "loss": 0.2118,
+      "step": 3271
+    },
+    {
+      "epoch": 0.15503435204927743,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018843033763164363,
+      "loss": 0.1658,
+      "step": 3272
+    },
+    {
+      "epoch": 0.1550817341862118,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018842338310875786,
+      "loss": 0.8745,
+      "step": 3273
+    },
+    {
+      "epoch": 0.15512911632314616,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018841642662473565,
+      "loss": 0.0557,
+      "step": 3274
+    },
+    {
+      "epoch": 0.15517649846008055,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018840946817973126,
+      "loss": 0.1246,
+      "step": 3275
+    },
+    {
+      "epoch": 0.15522388059701492,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018840250777389902,
+      "loss": 0.8924,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1552712627339493,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018839554540739335,
+      "loss": 0.5771,
+      "step": 3277
+    },
+    {
+      "epoch": 0.15531864487088368,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018838858108036864,
+      "loss": 1.1268,
+      "step": 3278
+    },
+    {
+      "epoch": 0.15536602700781804,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018838161479297933,
+      "loss": 1.1074,
+      "step": 3279
+    },
+    {
+      "epoch": 0.15541340914475243,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018837464654537995,
+      "loss": 1.6613,
+      "step": 3280
+    },
+    {
+      "epoch": 0.1554607912816868,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018836767633772502,
+      "loss": 1.0114,
+      "step": 3281
+    },
+    {
+      "epoch": 0.1555081734186212,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001883607041701692,
+      "loss": 1.2269,
+      "step": 3282
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018835373004286702,
+      "loss": 1.402,
+      "step": 3283
+    },
+    {
+      "epoch": 0.15560293769248992,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018834675395597324,
+      "loss": 1.0193,
+      "step": 3284
+    },
+    {
+      "epoch": 0.15565031982942432,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018833977590964257,
+      "loss": 1.3892,
+      "step": 3285
+    },
+    {
+      "epoch": 0.15569770196635868,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001883327959040298,
+      "loss": 0.4987,
+      "step": 3286
+    },
+    {
+      "epoch": 0.15574508410329305,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018832581393928965,
+      "loss": 1.3572,
+      "step": 3287
+    },
+    {
+      "epoch": 0.15579246624022744,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018831883001557706,
+      "loss": 1.2606,
+      "step": 3288
+    },
+    {
+      "epoch": 0.1558398483771618,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001883118441330469,
+      "loss": 1.7267,
+      "step": 3289
+    },
+    {
+      "epoch": 0.1558872305140962,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018830485629185405,
+      "loss": 1.2203,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15593461265103056,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00018829786649215358,
+      "loss": 0.3889,
+      "step": 3291
+    },
+    {
+      "epoch": 0.15598199478796493,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018829087473410048,
+      "loss": 1.3834,
+      "step": 3292
+    },
+    {
+      "epoch": 0.15602937692489932,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018828388101784985,
+      "loss": 1.2501,
+      "step": 3293
+    },
+    {
+      "epoch": 0.1560767590618337,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001882768853435567,
+      "loss": 0.1653,
+      "step": 3294
+    },
+    {
+      "epoch": 0.15612414119876805,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018826988771137635,
+      "loss": 0.2747,
+      "step": 3295
+    },
+    {
+      "epoch": 0.15617152333570244,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018826288812146387,
+      "loss": 0.7825,
+      "step": 3296
+    },
+    {
+      "epoch": 0.1562189054726368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018825588657397452,
+      "loss": 0.854,
+      "step": 3297
+    },
+    {
+      "epoch": 0.1562662876095712,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018824888306906366,
+      "loss": 1.0428,
+      "step": 3298
+    },
+    {
+      "epoch": 0.15631366974650557,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018824187760688654,
+      "loss": 1.6231,
+      "step": 3299
+    },
+    {
+      "epoch": 0.15636105188343993,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001882348701875986,
+      "loss": 0.2232,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15640843402037433,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018822786081135518,
+      "loss": 0.8423,
+      "step": 3301
+    },
+    {
+      "epoch": 0.1564558161573087,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018822084947831181,
+      "loss": 0.9263,
+      "step": 3302
+    },
+    {
+      "epoch": 0.15650319829424306,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018821383618862397,
+      "loss": 0.1621,
+      "step": 3303
+    },
+    {
+      "epoch": 0.15655058043117745,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018820682094244717,
+      "loss": 1.0773,
+      "step": 3304
+    },
+    {
+      "epoch": 0.15659796256811181,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00018819980373993705,
+      "loss": 0.0241,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1566453447050462,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018819278458124923,
+      "loss": 1.2362,
+      "step": 3306
+    },
+    {
+      "epoch": 0.15669272684198057,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001881857634665394,
+      "loss": 1.1092,
+      "step": 3307
+    },
+    {
+      "epoch": 0.15674010897891494,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018817874039596326,
+      "loss": 0.7047,
+      "step": 3308
+    },
+    {
+      "epoch": 0.15678749111584933,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00018817171536967658,
+      "loss": 0.0124,
+      "step": 3309
+    },
+    {
+      "epoch": 0.1568348732527837,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00018816468838783518,
+      "loss": 0.4429,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1568822553897181,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001881576594505949,
+      "loss": 0.8604,
+      "step": 3311
+    },
+    {
+      "epoch": 0.15692963752665245,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001881506285581116,
+      "loss": 0.155,
+      "step": 3312
+    },
+    {
+      "epoch": 0.15697701966358682,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001881435957105413,
+      "loss": 0.875,
+      "step": 3313
+    },
+    {
+      "epoch": 0.1570244018005212,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00018813656090803992,
+      "loss": 1.6006,
+      "step": 3314
+    },
+    {
+      "epoch": 0.15707178393745558,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018812952415076347,
+      "loss": 0.1606,
+      "step": 3315
+    },
+    {
+      "epoch": 0.15711916607438994,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018812248543886807,
+      "loss": 0.0366,
+      "step": 3316
+    },
+    {
+      "epoch": 0.15716654821132434,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001881154447725098,
+      "loss": 0.5749,
+      "step": 3317
+    },
+    {
+      "epoch": 0.1572139303482587,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018810840215184485,
+      "loss": 0.1889,
+      "step": 3318
+    },
+    {
+      "epoch": 0.1572613124851931,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018810135757702935,
+      "loss": 1.1963,
+      "step": 3319
+    },
+    {
+      "epoch": 0.15730869462212746,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001880943110482196,
+      "loss": 1.172,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15735607675906182,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018808726256557185,
+      "loss": 0.8667,
+      "step": 3321
+    },
+    {
+      "epoch": 0.15740345889599622,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018808021212924245,
+      "loss": 1.3381,
+      "step": 3322
+    },
+    {
+      "epoch": 0.15745084103293058,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018807315973938776,
+      "loss": 1.1065,
+      "step": 3323
+    },
+    {
+      "epoch": 0.15749822316986495,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018806610539616423,
+      "loss": 1.1935,
+      "step": 3324
+    },
+    {
+      "epoch": 0.15754560530679934,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00018805904909972824,
+      "loss": 0.7319,
+      "step": 3325
+    },
+    {
+      "epoch": 0.1575929874437337,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018805199085023637,
+      "loss": 1.0317,
+      "step": 3326
+    },
+    {
+      "epoch": 0.1576403695806681,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018804493064784511,
+      "loss": 0.1413,
+      "step": 3327
+    },
+    {
+      "epoch": 0.15768775171760246,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018803786849271107,
+      "loss": 1.0061,
+      "step": 3328
+    },
+    {
+      "epoch": 0.15773513385453683,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001880308043849909,
+      "loss": 0.4742,
+      "step": 3329
+    },
+    {
+      "epoch": 0.15778251599147122,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018802373832484123,
+      "loss": 0.0569,
+      "step": 3330
+    },
+    {
+      "epoch": 0.1578298981284056,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018801667031241882,
+      "loss": 0.3141,
+      "step": 3331
+    },
+    {
+      "epoch": 0.15787728026533995,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018800960034788043,
+      "loss": 0.4049,
+      "step": 3332
+    },
+    {
+      "epoch": 0.15792466240227435,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018800252843138282,
+      "loss": 1.3762,
+      "step": 3333
+    },
+    {
+      "epoch": 0.1579720445392087,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018799545456308289,
+      "loss": 1.1696,
+      "step": 3334
+    },
+    {
+      "epoch": 0.1580194266761431,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00018798837874313748,
+      "loss": 0.1247,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15806680881307747,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00018798130097170358,
+      "loss": 0.962,
+      "step": 3336
+    },
+    {
+      "epoch": 0.15811419095001183,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018797422124893807,
+      "loss": 0.8894,
+      "step": 3337
+    },
+    {
+      "epoch": 0.15816157308694623,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001879671395749981,
+      "loss": 0.8901,
+      "step": 3338
+    },
+    {
+      "epoch": 0.1582089552238806,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00018796005595004067,
+      "loss": 0.9047,
+      "step": 3339
+    },
+    {
+      "epoch": 0.15825633736081499,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001879529703742229,
+      "loss": 1.0445,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15830371949774935,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018794588284770186,
+      "loss": 1.084,
+      "step": 3341
+    },
+    {
+      "epoch": 0.15835110163468372,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018793879337063488,
+      "loss": 1.3039,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1583984837716181,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018793170194317913,
+      "loss": 1.1952,
+      "step": 3343
+    },
+    {
+      "epoch": 0.15844586590855247,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018792460856549185,
+      "loss": 1.0394,
+      "step": 3344
+    },
+    {
+      "epoch": 0.15849324804548684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018791751323773043,
+      "loss": 0.5739,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15854063018242123,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018791041596005225,
+      "loss": 0.9165,
+      "step": 3346
+    },
+    {
+      "epoch": 0.1585880123193556,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018790331673261464,
+      "loss": 1.1403,
+      "step": 3347
+    },
+    {
+      "epoch": 0.15863539445629,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001878962155555751,
+      "loss": 1.0392,
+      "step": 3348
+    },
+    {
+      "epoch": 0.15868277659322436,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018788911242909113,
+      "loss": 1.1052,
+      "step": 3349
+    },
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018788200735332024,
+      "loss": 1.0922,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15877754086709311,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018787490032842006,
+      "loss": 1.0305,
+      "step": 3351
+    },
+    {
+      "epoch": 0.15882492300402748,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018786779135454817,
+      "loss": 0.1058,
+      "step": 3352
+    },
+    {
+      "epoch": 0.15887230514096184,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018786068043186226,
+      "loss": 0.1585,
+      "step": 3353
+    },
+    {
+      "epoch": 0.15891968727789624,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018785356756052007,
+      "loss": 0.3798,
+      "step": 3354
+    },
+    {
+      "epoch": 0.1589670694148306,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018784645274067931,
+      "loss": 0.9803,
+      "step": 3355
+    },
+    {
+      "epoch": 0.159014451551765,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001878393359724978,
+      "loss": 0.0151,
+      "step": 3356
+    },
+    {
+      "epoch": 0.15906183368869936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018783221725613336,
+      "loss": 0.762,
+      "step": 3357
+    },
+    {
+      "epoch": 0.15910921582563373,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018782509659174393,
+      "loss": 0.0296,
+      "step": 3358
+    },
+    {
+      "epoch": 0.15915659796256812,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001878179739794874,
+      "loss": 0.0517,
+      "step": 3359
+    },
+    {
+      "epoch": 0.15920398009950248,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018781084941952173,
+      "loss": 0.2991,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15925136223643685,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018780372291200497,
+      "loss": 0.8215,
+      "step": 3361
+    },
+    {
+      "epoch": 0.15929874437337124,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018779659445709515,
+      "loss": 0.6526,
+      "step": 3362
+    },
+    {
+      "epoch": 0.1593461265103056,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018778946405495036,
+      "loss": 0.0389,
+      "step": 3363
+    },
+    {
+      "epoch": 0.15939350864724,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018778233170572877,
+      "loss": 0.1111,
+      "step": 3364
+    },
+    {
+      "epoch": 0.15944089078417437,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018777519740958862,
+      "loss": 1.1919,
+      "step": 3365
+    },
+    {
+      "epoch": 0.15948827292110873,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00018776806116668803,
+      "loss": 0.0036,
+      "step": 3366
+    },
+    {
+      "epoch": 0.15953565505804312,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018776092297718535,
+      "loss": 0.8826,
+      "step": 3367
+    },
+    {
+      "epoch": 0.1595830371949775,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018775378284123888,
+      "loss": 0.0293,
+      "step": 3368
+    },
+    {
+      "epoch": 0.15963041933191188,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018774664075900697,
+      "loss": 0.8727,
+      "step": 3369
+    },
+    {
+      "epoch": 0.15967780146884625,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018773949673064804,
+      "loss": 0.9371,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1597251836057806,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018773235075632052,
+      "loss": 1.1879,
+      "step": 3371
+    },
+    {
+      "epoch": 0.159772565742715,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001877252028361829,
+      "loss": 0.7749,
+      "step": 3372
+    },
+    {
+      "epoch": 0.15981994787964937,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018771805297039374,
+      "loss": 0.829,
+      "step": 3373
+    },
+    {
+      "epoch": 0.15986733001658374,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001877109011591116,
+      "loss": 1.4261,
+      "step": 3374
+    },
+    {
+      "epoch": 0.15991471215351813,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018770374740249507,
+      "loss": 0.3003,
+      "step": 3375
+    },
+    {
+      "epoch": 0.1599620942904525,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018769659170070287,
+      "loss": 0.0686,
+      "step": 3376
+    },
+    {
+      "epoch": 0.1600094764273869,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00018768943405389366,
+      "loss": 0.1909,
+      "step": 3377
+    },
+    {
+      "epoch": 0.16005685856432125,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018768227446222622,
+      "loss": 0.1452,
+      "step": 3378
+    },
+    {
+      "epoch": 0.16010424070125562,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001876751129258593,
+      "loss": 0.8432,
+      "step": 3379
+    },
+    {
+      "epoch": 0.16015162283819,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018766794944495178,
+      "loss": 0.8528,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16019900497512438,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001876607840196625,
+      "loss": 0.9679,
+      "step": 3381
+    },
+    {
+      "epoch": 0.16024638711205874,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018765361665015043,
+      "loss": 0.9675,
+      "step": 3382
+    },
+    {
+      "epoch": 0.16029376924899313,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018764644733657452,
+      "loss": 0.1435,
+      "step": 3383
+    },
+    {
+      "epoch": 0.1603411513859275,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018763927607909375,
+      "loss": 0.6409,
+      "step": 3384
+    },
+    {
+      "epoch": 0.1603885335228619,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001876321028778672,
+      "loss": 0.0473,
+      "step": 3385
+    },
+    {
+      "epoch": 0.16043591565979626,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001876249277330539,
+      "loss": 1.3328,
+      "step": 3386
+    },
+    {
+      "epoch": 0.16048329779673062,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018761775064481308,
+      "loss": 0.0951,
+      "step": 3387
+    },
+    {
+      "epoch": 0.16053067993366502,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018761057161330386,
+      "loss": 1.1392,
+      "step": 3388
+    },
+    {
+      "epoch": 0.16057806207059938,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001876033906386855,
+      "loss": 0.037,
+      "step": 3389
+    },
+    {
+      "epoch": 0.16062544420753375,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018759620772111721,
+      "loss": 0.0903,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16067282634446814,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018758902286075837,
+      "loss": 0.7807,
+      "step": 3391
+    },
+    {
+      "epoch": 0.1607202084814025,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018758183605776827,
+      "loss": 1.0448,
+      "step": 3392
+    },
+    {
+      "epoch": 0.1607675906183369,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018757464731230635,
+      "loss": 1.4464,
+      "step": 3393
+    },
+    {
+      "epoch": 0.16081497275527126,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018756745662453205,
+      "loss": 1.3067,
+      "step": 3394
+    },
+    {
+      "epoch": 0.16086235489220563,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001875602639946048,
+      "loss": 1.1364,
+      "step": 3395
+    },
+    {
+      "epoch": 0.16090973702914002,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018755306942268418,
+      "loss": 0.1941,
+      "step": 3396
+    },
+    {
+      "epoch": 0.16095711916607439,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018754587290892974,
+      "loss": 0.7959,
+      "step": 3397
+    },
+    {
+      "epoch": 0.16100450130300878,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018753867445350108,
+      "loss": 1.3102,
+      "step": 3398
+    },
+    {
+      "epoch": 0.16105188343994314,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018753147405655787,
+      "loss": 0.9047,
+      "step": 3399
+    },
+    {
+      "epoch": 0.1610992655768775,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001875242717182598,
+      "loss": 0.0682,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1611466477138119,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001875170674387666,
+      "loss": 1.1625,
+      "step": 3401
+    },
+    {
+      "epoch": 0.16119402985074627,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001875098612182381,
+      "loss": 1.2987,
+      "step": 3402
+    },
+    {
+      "epoch": 0.16124141198768063,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018750265305683404,
+      "loss": 1.1018,
+      "step": 3403
+    },
+    {
+      "epoch": 0.16128879412461503,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018749544295471436,
+      "loss": 1.2829,
+      "step": 3404
+    },
+    {
+      "epoch": 0.1613361762615494,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018748823091203892,
+      "loss": 0.4753,
+      "step": 3405
+    },
+    {
+      "epoch": 0.16138355839848378,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018748101692896775,
+      "loss": 1.0085,
+      "step": 3406
+    },
+    {
+      "epoch": 0.16143094053541815,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001874738010056608,
+      "loss": 0.0389,
+      "step": 3407
+    },
+    {
+      "epoch": 0.16147832267235251,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001874665831422781,
+      "loss": 0.6164,
+      "step": 3408
+    },
+    {
+      "epoch": 0.1615257048092869,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018745936333897976,
+      "loss": 0.8375,
+      "step": 3409
+    },
+    {
+      "epoch": 0.16157308694622127,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001874521415959259,
+      "loss": 0.0557,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16162046908315564,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018744491791327668,
+      "loss": 1.331,
+      "step": 3411
+    },
+    {
+      "epoch": 0.16166785122009003,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018743769229119232,
+      "loss": 0.8699,
+      "step": 3412
+    },
+    {
+      "epoch": 0.1617152333570244,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001874304647298331,
+      "loss": 0.0352,
+      "step": 3413
+    },
+    {
+      "epoch": 0.1617626154939588,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001874232352293593,
+      "loss": 0.0239,
+      "step": 3414
+    },
+    {
+      "epoch": 0.16180999763089315,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018741600378993124,
+      "loss": 0.9394,
+      "step": 3415
+    },
+    {
+      "epoch": 0.16185737976782752,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018740877041170935,
+      "loss": 0.6944,
+      "step": 3416
+    },
+    {
+      "epoch": 0.1619047619047619,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000187401535094854,
+      "loss": 1.0595,
+      "step": 3417
+    },
+    {
+      "epoch": 0.16195214404169628,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018739429783952574,
+      "loss": 0.993,
+      "step": 3418
+    },
+    {
+      "epoch": 0.16199952617863064,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000187387058645885,
+      "loss": 0.6458,
+      "step": 3419
+    },
+    {
+      "epoch": 0.16204690831556504,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018737981751409241,
+      "loss": 0.9488,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1620942904524994,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018737257444430855,
+      "loss": 0.7799,
+      "step": 3421
+    },
+    {
+      "epoch": 0.1621416725894338,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00018736532943669404,
+      "loss": 1.1519,
+      "step": 3422
+    },
+    {
+      "epoch": 0.16218905472636816,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018735808249140953,
+      "loss": 1.179,
+      "step": 3423
+    },
+    {
+      "epoch": 0.16223643686330252,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018735083360861587,
+      "loss": 0.9812,
+      "step": 3424
+    },
+    {
+      "epoch": 0.16228381900023692,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018734358278847376,
+      "loss": 1.1356,
+      "step": 3425
+    },
+    {
+      "epoch": 0.16233120113717128,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018733633003114396,
+      "loss": 1.0417,
+      "step": 3426
+    },
+    {
+      "epoch": 0.16237858327410568,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018732907533678742,
+      "loss": 0.8498,
+      "step": 3427
+    },
+    {
+      "epoch": 0.16242596541104004,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018732181870556502,
+      "loss": 0.1318,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1624733475479744,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001873145601376377,
+      "loss": 0.9391,
+      "step": 3429
+    },
+    {
+      "epoch": 0.1625207296849088,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018730729963316642,
+      "loss": 0.9662,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16256811182184316,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018730003719231224,
+      "loss": 1.0315,
+      "step": 3431
+    },
+    {
+      "epoch": 0.16261549395877753,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001872927728152362,
+      "loss": 1.0004,
+      "step": 3432
+    },
+    {
+      "epoch": 0.16266287609571192,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018728550650209946,
+      "loss": 0.0759,
+      "step": 3433
+    },
+    {
+      "epoch": 0.1627102582326463,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018727823825306317,
+      "loss": 0.9821,
+      "step": 3434
+    },
+    {
+      "epoch": 0.16275764036958068,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018727096806828847,
+      "loss": 0.1027,
+      "step": 3435
+    },
+    {
+      "epoch": 0.16280502250651505,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018726369594793672,
+      "loss": 0.5308,
+      "step": 3436
+    },
+    {
+      "epoch": 0.1628524046434494,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018725642189216908,
+      "loss": 1.4648,
+      "step": 3437
+    },
+    {
+      "epoch": 0.1628997867803838,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000187249145901147,
+      "loss": 1.4506,
+      "step": 3438
+    },
+    {
+      "epoch": 0.16294716891731817,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018724186797503177,
+      "loss": 0.5585,
+      "step": 3439
+    },
+    {
+      "epoch": 0.16299455105425253,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018723458811398483,
+      "loss": 0.1731,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16304193319118693,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018722730631816765,
+      "loss": 1.1493,
+      "step": 3441
+    },
+    {
+      "epoch": 0.1630893153281213,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018722002258774172,
+      "loss": 0.0177,
+      "step": 3442
+    },
+    {
+      "epoch": 0.16313669746505569,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001872127369228686,
+      "loss": 0.3601,
+      "step": 3443
+    },
+    {
+      "epoch": 0.16318407960199005,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018720544932370988,
+      "loss": 0.0731,
+      "step": 3444
+    },
+    {
+      "epoch": 0.16323146173892442,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00018719815979042714,
+      "loss": 0.8687,
+      "step": 3445
+    },
+    {
+      "epoch": 0.1632788438758588,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00018719086832318213,
+      "loss": 0.9003,
+      "step": 3446
+    },
+    {
+      "epoch": 0.16332622601279317,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018718357492213654,
+      "loss": 0.1883,
+      "step": 3447
+    },
+    {
+      "epoch": 0.16337360814972754,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001871762795874521,
+      "loss": 0.9352,
+      "step": 3448
+    },
+    {
+      "epoch": 0.16342099028666193,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018716898231929064,
+      "loss": 1.4955,
+      "step": 3449
+    },
+    {
+      "epoch": 0.1634683724235963,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000187161683117814,
+      "loss": 0.0188,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1635157545605307,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018715438198318407,
+      "loss": 1.0635,
+      "step": 3451
+    },
+    {
+      "epoch": 0.16356313669746506,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001871470789155628,
+      "loss": 0.8331,
+      "step": 3452
+    },
+    {
+      "epoch": 0.16361051883439942,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018713977391511213,
+      "loss": 1.1757,
+      "step": 3453
+    },
+    {
+      "epoch": 0.16365790097133381,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018713246698199407,
+      "loss": 1.1352,
+      "step": 3454
+    },
+    {
+      "epoch": 0.16370528310826818,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018712515811637073,
+      "loss": 0.7567,
+      "step": 3455
+    },
+    {
+      "epoch": 0.16375266524520257,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018711784731840415,
+      "loss": 0.1355,
+      "step": 3456
+    },
+    {
+      "epoch": 0.16380004738213694,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018711053458825655,
+      "loss": 1.2875,
+      "step": 3457
+    },
+    {
+      "epoch": 0.1638474295190713,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018710321992609008,
+      "loss": 0.0153,
+      "step": 3458
+    },
+    {
+      "epoch": 0.1638948116560057,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018709590333206694,
+      "loss": 0.0086,
+      "step": 3459
+    },
+    {
+      "epoch": 0.16394219379294006,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00018708858480634945,
+      "loss": 0.3582,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16398957592987443,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018708126434909992,
+      "loss": 1.1043,
+      "step": 3461
+    },
+    {
+      "epoch": 0.16403695806680882,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001870739419604807,
+      "loss": 0.7051,
+      "step": 3462
+    },
+    {
+      "epoch": 0.16408434020374318,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018706661764065417,
+      "loss": 0.8314,
+      "step": 3463
+    },
+    {
+      "epoch": 0.16413172234067758,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001870592913897828,
+      "loss": 0.0076,
+      "step": 3464
+    },
+    {
+      "epoch": 0.16417910447761194,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001870519632080291,
+      "loss": 0.216,
+      "step": 3465
+    },
+    {
+      "epoch": 0.1642264866145463,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001870446330955556,
+      "loss": 0.1688,
+      "step": 3466
+    },
+    {
+      "epoch": 0.1642738687514807,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018703730105252482,
+      "loss": 1.0041,
+      "step": 3467
+    },
+    {
+      "epoch": 0.16432125088841507,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001870299670790994,
+      "loss": 0.0299,
+      "step": 3468
+    },
+    {
+      "epoch": 0.16436863302534943,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018702263117544203,
+      "loss": 0.1967,
+      "step": 3469
+    },
+    {
+      "epoch": 0.16441601516228382,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001870152933417154,
+      "loss": 0.9879,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1644633972992182,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018700795357808224,
+      "loss": 0.6545,
+      "step": 3471
+    },
+    {
+      "epoch": 0.16451077943615258,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00018700061188470533,
+      "loss": 1.3322,
+      "step": 3472
+    },
+    {
+      "epoch": 0.16455816157308695,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018699326826174755,
+      "loss": 1.2804,
+      "step": 3473
+    },
+    {
+      "epoch": 0.1646055437100213,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018698592270937172,
+      "loss": 1.1649,
+      "step": 3474
+    },
+    {
+      "epoch": 0.1646529258469557,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001869785752277408,
+      "loss": 0.8521,
+      "step": 3475
+    },
+    {
+      "epoch": 0.16470030798389007,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018697122581701767,
+      "loss": 0.6356,
+      "step": 3476
+    },
+    {
+      "epoch": 0.16474769012082444,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018696387447736544,
+      "loss": 0.1956,
+      "step": 3477
+    },
+    {
+      "epoch": 0.16479507225775883,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001869565212089471,
+      "loss": 0.9507,
+      "step": 3478
+    },
+    {
+      "epoch": 0.1648424543946932,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00018694916601192573,
+      "loss": 0.1632,
+      "step": 3479
+    },
+    {
+      "epoch": 0.1648898365316276,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018694180888646447,
+      "loss": 0.953,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16493721866856195,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001869344498327265,
+      "loss": 1.1483,
+      "step": 3481
+    },
+    {
+      "epoch": 0.16498460080549632,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00018692708885087504,
+      "loss": 1.2897,
+      "step": 3482
+    },
+    {
+      "epoch": 0.1650319829424307,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018691972594107333,
+      "loss": 0.8568,
+      "step": 3483
+    },
+    {
+      "epoch": 0.16507936507936508,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018691236110348467,
+      "loss": 0.276,
+      "step": 3484
+    },
+    {
+      "epoch": 0.16512674721629947,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00018690499433827244,
+      "loss": 0.7986,
+      "step": 3485
+    },
+    {
+      "epoch": 0.16517412935323383,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018689762564559997,
+      "loss": 0.0847,
+      "step": 3486
+    },
+    {
+      "epoch": 0.1652215114901682,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018689025502563076,
+      "loss": 0.156,
+      "step": 3487
+    },
+    {
+      "epoch": 0.1652688936271026,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018688288247852822,
+      "loss": 1.1043,
+      "step": 3488
+    },
+    {
+      "epoch": 0.16531627576403696,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018687550800445591,
+      "loss": 1.3915,
+      "step": 3489
+    },
+    {
+      "epoch": 0.16536365790097132,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018686813160357732,
+      "loss": 0.8655,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16541104003790572,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018686075327605615,
+      "loss": 1.0847,
+      "step": 3491
+    },
+    {
+      "epoch": 0.16545842217484008,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000186853373022056,
+      "loss": 1.0577,
+      "step": 3492
+    },
+    {
+      "epoch": 0.16550580431177447,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018684599084174051,
+      "loss": 0.199,
+      "step": 3493
+    },
+    {
+      "epoch": 0.16555318644870884,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018683860673527348,
+      "loss": 1.3364,
+      "step": 3494
+    },
+    {
+      "epoch": 0.1656005685856432,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001868312207028186,
+      "loss": 0.2806,
+      "step": 3495
+    },
+    {
+      "epoch": 0.1656479507225776,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018682383274453977,
+      "loss": 0.9413,
+      "step": 3496
+    },
+    {
+      "epoch": 0.16569533285951196,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018681644286060083,
+      "loss": 0.8934,
+      "step": 3497
+    },
+    {
+      "epoch": 0.16574271499644633,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018680905105116562,
+      "loss": 0.0685,
+      "step": 3498
+    },
+    {
+      "epoch": 0.16579009713338072,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00018680165731639815,
+      "loss": 0.8225,
+      "step": 3499
+    },
+    {
+      "epoch": 0.16583747927031509,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018679426165646237,
+      "loss": 0.9894,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16588486140724948,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018678686407152227,
+      "loss": 1.1245,
+      "step": 3501
+    },
+    {
+      "epoch": 0.16593224354418384,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018677946456174206,
+      "loss": 0.104,
+      "step": 3502
+    },
+    {
+      "epoch": 0.1659796256811182,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018677206312728569,
+      "loss": 0.9713,
+      "step": 3503
+    },
+    {
+      "epoch": 0.1660270078180526,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001867646597683174,
+      "loss": 0.0467,
+      "step": 3504
+    },
+    {
+      "epoch": 0.16607438995498697,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018675725448500136,
+      "loss": 0.1175,
+      "step": 3505
+    },
+    {
+      "epoch": 0.16612177209192133,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018674984727750184,
+      "loss": 0.274,
+      "step": 3506
+    },
+    {
+      "epoch": 0.16616915422885573,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001867424381459831,
+      "loss": 0.8968,
+      "step": 3507
+    },
+    {
+      "epoch": 0.1662165363657901,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001867350270906095,
+      "loss": 1.1854,
+      "step": 3508
+    },
+    {
+      "epoch": 0.16626391850272448,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018672761411154536,
+      "loss": 1.1038,
+      "step": 3509
+    },
+    {
+      "epoch": 0.16631130063965885,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018672019920895513,
+      "loss": 0.5199,
+      "step": 3510
+    },
+    {
+      "epoch": 0.16635868277659321,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018671278238300328,
+      "loss": 1.3755,
+      "step": 3511
+    },
+    {
+      "epoch": 0.1664060649135276,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001867053636338542,
+      "loss": 0.9573,
+      "step": 3512
+    },
+    {
+      "epoch": 0.16645344705046197,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018669794296167258,
+      "loss": 0.0389,
+      "step": 3513
+    },
+    {
+      "epoch": 0.16650082918739637,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0001866905203666229,
+      "loss": 0.0144,
+      "step": 3514
+    },
+    {
+      "epoch": 0.16654821132433073,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018668309584886984,
+      "loss": 1.4951,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1665955934612651,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018667566940857805,
+      "loss": 0.8611,
+      "step": 3516
+    },
+    {
+      "epoch": 0.1666429755981995,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018666824104591218,
+      "loss": 0.9698,
+      "step": 3517
+    },
+    {
+      "epoch": 0.16669035773513385,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001866608107610371,
+      "loss": 1.074,
+      "step": 3518
+    },
+    {
+      "epoch": 0.16673773987206822,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001866533785541175,
+      "loss": 0.5675,
+      "step": 3519
+    },
+    {
+      "epoch": 0.1667851220090026,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00018664594442531832,
+      "loss": 0.0248,
+      "step": 3520
+    },
+    {
+      "epoch": 0.16683250414593698,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001866385083748043,
+      "loss": 0.0185,
+      "step": 3521
+    },
+    {
+      "epoch": 0.16687988628287137,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001866310704027405,
+      "loss": 0.7097,
+      "step": 3522
+    },
+    {
+      "epoch": 0.16692726841980574,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018662363050929182,
+      "loss": 0.1294,
+      "step": 3523
+    },
+    {
+      "epoch": 0.1669746505567401,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018661618869462328,
+      "loss": 0.8745,
+      "step": 3524
+    },
+    {
+      "epoch": 0.1670220326936745,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018660874495889996,
+      "loss": 1.0901,
+      "step": 3525
+    },
+    {
+      "epoch": 0.16706941483060886,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00018660129930228695,
+      "loss": 0.1195,
+      "step": 3526
+    },
+    {
+      "epoch": 0.16711679696754322,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001865938517249493,
+      "loss": 0.0636,
+      "step": 3527
+    },
+    {
+      "epoch": 0.16716417910447762,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001865864022270523,
+      "loss": 0.9913,
+      "step": 3528
+    },
+    {
+      "epoch": 0.16721156124141198,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018657895080876109,
+      "loss": 0.8216,
+      "step": 3529
+    },
+    {
+      "epoch": 0.16725894337834638,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018657149747024099,
+      "loss": 1.1584,
+      "step": 3530
+    },
+    {
+      "epoch": 0.16730632551528074,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018656404221165728,
+      "loss": 1.5047,
+      "step": 3531
+    },
+    {
+      "epoch": 0.1673537076522151,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018655658503317533,
+      "loss": 1.0045,
+      "step": 3532
+    },
+    {
+      "epoch": 0.1674010897891495,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001865491259349605,
+      "loss": 1.0516,
+      "step": 3533
+    },
+    {
+      "epoch": 0.16744847192608386,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018654166491717826,
+      "loss": 0.742,
+      "step": 3534
+    },
+    {
+      "epoch": 0.16749585406301823,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001865342019799941,
+      "loss": 0.2457,
+      "step": 3535
+    },
+    {
+      "epoch": 0.16754323619995262,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018652673712357345,
+      "loss": 1.1417,
+      "step": 3536
+    },
+    {
+      "epoch": 0.167590618336887,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00018651927034808198,
+      "loss": 0.1849,
+      "step": 3537
+    },
+    {
+      "epoch": 0.16763800047382138,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018651180165368524,
+      "loss": 1.0322,
+      "step": 3538
+    },
+    {
+      "epoch": 0.16768538261075575,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018650433104054888,
+      "loss": 1.1652,
+      "step": 3539
+    },
+    {
+      "epoch": 0.1677327647476901,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001864968585088386,
+      "loss": 1.1635,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1677801468846245,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00018648938405872015,
+      "loss": 0.0201,
+      "step": 3541
+    },
+    {
+      "epoch": 0.16782752902155887,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018648190769035927,
+      "loss": 1.3377,
+      "step": 3542
+    },
+    {
+      "epoch": 0.16787491115849326,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001864744294039218,
+      "loss": 0.0092,
+      "step": 3543
+    },
+    {
+      "epoch": 0.16792229329542763,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018646694919957356,
+      "loss": 0.9292,
+      "step": 3544
+    },
+    {
+      "epoch": 0.167969675432362,
+      "grad_norm": 0.0072021484375,
+      "learning_rate": 0.00018645946707748052,
+      "loss": 0.0004,
+      "step": 3545
+    },
+    {
+      "epoch": 0.16801705756929639,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001864519830378086,
+      "loss": 1.5274,
+      "step": 3546
+    },
+    {
+      "epoch": 0.16806443970623075,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018644449708072377,
+      "loss": 0.6784,
+      "step": 3547
+    },
+    {
+      "epoch": 0.16811182184316512,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018643700920639207,
+      "loss": 1.0314,
+      "step": 3548
+    },
+    {
+      "epoch": 0.1681592039800995,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018642951941497956,
+      "loss": 1.3053,
+      "step": 3549
+    },
+    {
+      "epoch": 0.16820658611703387,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018642202770665237,
+      "loss": 1.3362,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16825396825396827,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001864145340815767,
+      "loss": 1.524,
+      "step": 3551
+    },
+    {
+      "epoch": 0.16830135039090263,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018640703853991869,
+      "loss": 0.9737,
+      "step": 3552
+    },
+    {
+      "epoch": 0.168348732527837,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018639954108184457,
+      "loss": 0.5459,
+      "step": 3553
+    },
+    {
+      "epoch": 0.1683961146647714,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018639204170752067,
+      "loss": 0.8134,
+      "step": 3554
+    },
+    {
+      "epoch": 0.16844349680170576,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018638454041711332,
+      "loss": 1.1741,
+      "step": 3555
+    },
+    {
+      "epoch": 0.16849087893864012,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018637703721078886,
+      "loss": 0.6779,
+      "step": 3556
+    },
+    {
+      "epoch": 0.1685382610755745,
+      "grad_norm": 0.0123291015625,
+      "learning_rate": 0.00018636953208871372,
+      "loss": 0.0007,
+      "step": 3557
+    },
+    {
+      "epoch": 0.16858564321250888,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001863620250510544,
+      "loss": 0.7599,
+      "step": 3558
+    },
+    {
+      "epoch": 0.16863302534944327,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001863545160979773,
+      "loss": 0.0232,
+      "step": 3559
+    },
+    {
+      "epoch": 0.16868040748637764,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00018634700522964903,
+      "loss": 0.0911,
+      "step": 3560
+    },
+    {
+      "epoch": 0.168727789623312,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018633949244623615,
+      "loss": 1.1788,
+      "step": 3561
+    },
+    {
+      "epoch": 0.1687751717602464,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00018633197774790535,
+      "loss": 0.0551,
+      "step": 3562
+    },
+    {
+      "epoch": 0.16882255389718076,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018632446113482317,
+      "loss": 0.9242,
+      "step": 3563
+    },
+    {
+      "epoch": 0.16886993603411513,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001863169426071564,
+      "loss": 0.9733,
+      "step": 3564
+    },
+    {
+      "epoch": 0.16891731817104952,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018630942216507182,
+      "loss": 1.0315,
+      "step": 3565
+    },
+    {
+      "epoch": 0.16896470030798388,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018630189980873617,
+      "loss": 1.143,
+      "step": 3566
+    },
+    {
+      "epoch": 0.16901208244491828,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018629437553831628,
+      "loss": 0.8955,
+      "step": 3567
+    },
+    {
+      "epoch": 0.16905946458185264,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00018628684935397907,
+      "loss": 0.0415,
+      "step": 3568
+    },
+    {
+      "epoch": 0.169106846718787,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00018627932125589146,
+      "loss": 0.0113,
+      "step": 3569
+    },
+    {
+      "epoch": 0.1691542288557214,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001862717912442204,
+      "loss": 0.8555,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16920161099265577,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018626425931913293,
+      "loss": 0.1145,
+      "step": 3571
+    },
+    {
+      "epoch": 0.16924899312959013,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018625672548079606,
+      "loss": 0.8483,
+      "step": 3572
+    },
+    {
+      "epoch": 0.16929637526652452,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018624918972937686,
+      "loss": 1.2921,
+      "step": 3573
+    },
+    {
+      "epoch": 0.1693437574034589,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018624165206504253,
+      "loss": 1.0215,
+      "step": 3574
+    },
+    {
+      "epoch": 0.16939113954039328,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0001862341124879602,
+      "loss": 0.0082,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16943852167732765,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018622657099829715,
+      "loss": 1.2434,
+      "step": 3576
+    },
+    {
+      "epoch": 0.169485903814262,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018621902759622056,
+      "loss": 0.1075,
+      "step": 3577
+    },
+    {
+      "epoch": 0.1695332859511964,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018621148228189778,
+      "loss": 0.5215,
+      "step": 3578
+    },
+    {
+      "epoch": 0.16958066808813077,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018620393505549616,
+      "loss": 0.1551,
+      "step": 3579
+    },
+    {
+      "epoch": 0.16962805022506516,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018619638591718307,
+      "loss": 0.0896,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16967543236199953,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018618883486712595,
+      "loss": 0.6216,
+      "step": 3581
+    },
+    {
+      "epoch": 0.1697228144989339,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001861812819054923,
+      "loss": 0.8138,
+      "step": 3582
+    },
+    {
+      "epoch": 0.1697701966358683,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001861737270324496,
+      "loss": 1.1879,
+      "step": 3583
+    },
+    {
+      "epoch": 0.16981757877280265,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001861661702481654,
+      "loss": 1.7843,
+      "step": 3584
+    },
+    {
+      "epoch": 0.16986496090973702,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00018615861155280735,
+      "loss": 0.1884,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1699123430466714,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001861510509465431,
+      "loss": 0.214,
+      "step": 3586
+    },
+    {
+      "epoch": 0.16995972518360578,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018614348842954024,
+      "loss": 0.0472,
+      "step": 3587
+    },
+    {
+      "epoch": 0.17000710732054017,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001861359240019666,
+      "loss": 0.7151,
+      "step": 3588
+    },
+    {
+      "epoch": 0.17005448945747453,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018612835766398987,
+      "loss": 1.2108,
+      "step": 3589
+    },
+    {
+      "epoch": 0.1701018715944089,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00018612078941577799,
+      "loss": 1.048,
+      "step": 3590
+    },
+    {
+      "epoch": 0.1701492537313433,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018611321925749867,
+      "loss": 1.0845,
+      "step": 3591
+    },
+    {
+      "epoch": 0.17019663586827766,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001861056471893199,
+      "loss": 1.0258,
+      "step": 3592
+    },
+    {
+      "epoch": 0.17024401800521202,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018609807321140956,
+      "loss": 1.1531,
+      "step": 3593
+    },
+    {
+      "epoch": 0.17029140014214642,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001860904973239357,
+      "loss": 0.7115,
+      "step": 3594
+    },
+    {
+      "epoch": 0.17033878227908078,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018608291952706626,
+      "loss": 0.7323,
+      "step": 3595
+    },
+    {
+      "epoch": 0.17038616441601517,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001860753398209694,
+      "loss": 1.2005,
+      "step": 3596
+    },
+    {
+      "epoch": 0.17043354655294954,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018606775820581315,
+      "loss": 0.0186,
+      "step": 3597
+    },
+    {
+      "epoch": 0.1704809286898839,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001860601746817657,
+      "loss": 0.8583,
+      "step": 3598
+    },
+    {
+      "epoch": 0.1705283108268183,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018605258924899527,
+      "loss": 0.8947,
+      "step": 3599
+    },
+    {
+      "epoch": 0.17057569296375266,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018604500190767007,
+      "loss": 0.7192,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17062307510068703,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018603741265795835,
+      "loss": 0.0833,
+      "step": 3601
+    },
+    {
+      "epoch": 0.17067045723762142,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001860298215000285,
+      "loss": 0.8278,
+      "step": 3602
+    },
+    {
+      "epoch": 0.17071783937455579,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018602222843404882,
+      "loss": 0.6577,
+      "step": 3603
+    },
+    {
+      "epoch": 0.17076522151149018,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018601463346018776,
+      "loss": 0.9069,
+      "step": 3604
+    },
+    {
+      "epoch": 0.17081260364842454,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018600703657861372,
+      "loss": 0.4135,
+      "step": 3605
+    },
+    {
+      "epoch": 0.1708599857853589,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018599943778949523,
+      "loss": 0.2287,
+      "step": 3606
+    },
+    {
+      "epoch": 0.1709073679222933,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001859918370930008,
+      "loss": 0.5911,
+      "step": 3607
+    },
+    {
+      "epoch": 0.17095475005922767,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00018598423448929906,
+      "loss": 0.0499,
+      "step": 3608
+    },
+    {
+      "epoch": 0.17100213219616206,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018597662997855855,
+      "loss": 1.2342,
+      "step": 3609
+    },
+    {
+      "epoch": 0.17104951433309643,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018596902356094797,
+      "loss": 1.1811,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1710968964700308,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018596141523663601,
+      "loss": 1.1274,
+      "step": 3611
+    },
+    {
+      "epoch": 0.17114427860696518,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018595380500579142,
+      "loss": 0.9245,
+      "step": 3612
+    },
+    {
+      "epoch": 0.17119166074389955,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018594619286858301,
+      "loss": 0.4514,
+      "step": 3613
+    },
+    {
+      "epoch": 0.1712390428808339,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018593857882517957,
+      "loss": 0.0604,
+      "step": 3614
+    },
+    {
+      "epoch": 0.1712864250177683,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018593096287575,
+      "loss": 0.7656,
+      "step": 3615
+    },
+    {
+      "epoch": 0.17133380715470267,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0001859233450204632,
+      "loss": 0.0035,
+      "step": 3616
+    },
+    {
+      "epoch": 0.17138118929163706,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018591572525948814,
+      "loss": 0.8791,
+      "step": 3617
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001859081035929938,
+      "loss": 1.2135,
+      "step": 3618
+    },
+    {
+      "epoch": 0.1714759535655058,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001859004800211492,
+      "loss": 0.5741,
+      "step": 3619
+    },
+    {
+      "epoch": 0.1715233357024402,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018589285454412348,
+      "loss": 0.5292,
+      "step": 3620
+    },
+    {
+      "epoch": 0.17157071783937455,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018588522716208575,
+      "loss": 0.0636,
+      "step": 3621
+    },
+    {
+      "epoch": 0.17161809997630892,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018587759787520514,
+      "loss": 0.921,
+      "step": 3622
+    },
+    {
+      "epoch": 0.1716654821132433,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001858699666836509,
+      "loss": 1.152,
+      "step": 3623
+    },
+    {
+      "epoch": 0.17171286425017768,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018586233358759222,
+      "loss": 0.4444,
+      "step": 3624
+    },
+    {
+      "epoch": 0.17176024638711207,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018585469858719845,
+      "loss": 0.6908,
+      "step": 3625
+    },
+    {
+      "epoch": 0.17180762852404644,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018584706168263895,
+      "loss": 1.046,
+      "step": 3626
+    },
+    {
+      "epoch": 0.1718550106609808,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018583942287408303,
+      "loss": 1.1907,
+      "step": 3627
+    },
+    {
+      "epoch": 0.1719023927979152,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018583178216170017,
+      "loss": 0.9779,
+      "step": 3628
+    },
+    {
+      "epoch": 0.17194977493484956,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001858241395456598,
+      "loss": 1.2669,
+      "step": 3629
+    },
+    {
+      "epoch": 0.17199715707178392,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018581649502613138,
+      "loss": 0.0819,
+      "step": 3630
+    },
+    {
+      "epoch": 0.17204453920871832,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018580884860328455,
+      "loss": 0.0316,
+      "step": 3631
+    },
+    {
+      "epoch": 0.17209192134565268,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00018580120027728887,
+      "loss": 1.0357,
+      "step": 3632
+    },
+    {
+      "epoch": 0.17213930348258707,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018579355004831393,
+      "loss": 1.4687,
+      "step": 3633
+    },
+    {
+      "epoch": 0.17218668561952144,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018578589791652946,
+      "loss": 0.9925,
+      "step": 3634
+    },
+    {
+      "epoch": 0.1722340677564558,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018577824388210515,
+      "loss": 0.9039,
+      "step": 3635
+    },
+    {
+      "epoch": 0.1722814498933902,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00018577058794521075,
+      "loss": 0.0527,
+      "step": 3636
+    },
+    {
+      "epoch": 0.17232883203032456,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018576293010601604,
+      "loss": 0.5258,
+      "step": 3637
+    },
+    {
+      "epoch": 0.17237621416725896,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001857552703646909,
+      "loss": 0.1037,
+      "step": 3638
+    },
+    {
+      "epoch": 0.17242359630419332,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00018574760872140523,
+      "loss": 0.0104,
+      "step": 3639
+    },
+    {
+      "epoch": 0.1724709784411277,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018573994517632892,
+      "loss": 1.4852,
+      "step": 3640
+    },
+    {
+      "epoch": 0.17251836057806208,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018573227972963196,
+      "loss": 1.1529,
+      "step": 3641
+    },
+    {
+      "epoch": 0.17256574271499645,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018572461238148432,
+      "loss": 1.4178,
+      "step": 3642
+    },
+    {
+      "epoch": 0.1726131248519308,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018571694313205614,
+      "loss": 0.4535,
+      "step": 3643
+    },
+    {
+      "epoch": 0.1726605069888652,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018570927198151743,
+      "loss": 0.0695,
+      "step": 3644
+    },
+    {
+      "epoch": 0.17270788912579957,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00018570159893003835,
+      "loss": 0.0868,
+      "step": 3645
+    },
+    {
+      "epoch": 0.17275527126273396,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001856939239777891,
+      "loss": 0.6132,
+      "step": 3646
+    },
+    {
+      "epoch": 0.17280265339966833,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001856862471249399,
+      "loss": 1.2419,
+      "step": 3647
+    },
+    {
+      "epoch": 0.1728500355366027,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018567856837166104,
+      "loss": 0.3735,
+      "step": 3648
+    },
+    {
+      "epoch": 0.17289741767353708,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018567088771812275,
+      "loss": 0.7501,
+      "step": 3649
+    },
+    {
+      "epoch": 0.17294479981047145,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018566320516449545,
+      "loss": 1.2527,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17299218194740582,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018565552071094947,
+      "loss": 1.0334,
+      "step": 3651
+    },
+    {
+      "epoch": 0.1730395640843402,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001856478343576553,
+      "loss": 0.0948,
+      "step": 3652
+    },
+    {
+      "epoch": 0.17308694622127457,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001856401461047834,
+      "loss": 0.1733,
+      "step": 3653
+    },
+    {
+      "epoch": 0.17313432835820897,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018563245595250427,
+      "loss": 1.0906,
+      "step": 3654
+    },
+    {
+      "epoch": 0.17318171049514333,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018562476390098848,
+      "loss": 0.8008,
+      "step": 3655
+    },
+    {
+      "epoch": 0.1732290926320777,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018561706995040661,
+      "loss": 0.0755,
+      "step": 3656
+    },
+    {
+      "epoch": 0.1732764747690121,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018560937410092934,
+      "loss": 1.0517,
+      "step": 3657
+    },
+    {
+      "epoch": 0.17332385690594646,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018560167635272735,
+      "loss": 1.3768,
+      "step": 3658
+    },
+    {
+      "epoch": 0.17337123904288082,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018559397670597135,
+      "loss": 0.0153,
+      "step": 3659
+    },
+    {
+      "epoch": 0.1734186211798152,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018558627516083212,
+      "loss": 1.3332,
+      "step": 3660
+    },
+    {
+      "epoch": 0.17346600331674958,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018557857171748045,
+      "loss": 0.1742,
+      "step": 3661
+    },
+    {
+      "epoch": 0.17351338545368397,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018557086637608722,
+      "loss": 0.9789,
+      "step": 3662
+    },
+    {
+      "epoch": 0.17356076759061834,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00018556315913682335,
+      "loss": 0.1682,
+      "step": 3663
+    },
+    {
+      "epoch": 0.1736081497275527,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018555544999985973,
+      "loss": 1.0887,
+      "step": 3664
+    },
+    {
+      "epoch": 0.1736555318644871,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018554773896536735,
+      "loss": 1.5074,
+      "step": 3665
+    },
+    {
+      "epoch": 0.17370291400142146,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.00018554002603351724,
+      "loss": 0.0026,
+      "step": 3666
+    },
+    {
+      "epoch": 0.17375029613835585,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018553231120448048,
+      "loss": 0.0545,
+      "step": 3667
+    },
+    {
+      "epoch": 0.17379767827529022,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018552459447842813,
+      "loss": 1.0061,
+      "step": 3668
+    },
+    {
+      "epoch": 0.17384506041222458,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001855168758555314,
+      "loss": 1.1792,
+      "step": 3669
+    },
+    {
+      "epoch": 0.17389244254915898,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018550915533596145,
+      "loss": 0.0447,
+      "step": 3670
+    },
+    {
+      "epoch": 0.17393982468609334,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001855014329198895,
+      "loss": 1.3617,
+      "step": 3671
+    },
+    {
+      "epoch": 0.1739872068230277,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018549370860748688,
+      "loss": 0.8235,
+      "step": 3672
+    },
+    {
+      "epoch": 0.1740345889599621,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018548598239892484,
+      "loss": 1.2566,
+      "step": 3673
+    },
+    {
+      "epoch": 0.17408197109689647,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018547825429437474,
+      "loss": 0.8101,
+      "step": 3674
+    },
+    {
+      "epoch": 0.17412935323383086,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018547052429400803,
+      "loss": 1.1601,
+      "step": 3675
+    },
+    {
+      "epoch": 0.17417673537076522,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018546279239799613,
+      "loss": 0.0429,
+      "step": 3676
+    },
+    {
+      "epoch": 0.1742241175076996,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018545505860651055,
+      "loss": 1.0094,
+      "step": 3677
+    },
+    {
+      "epoch": 0.17427149964463398,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018544732291972274,
+      "loss": 1.0445,
+      "step": 3678
+    },
+    {
+      "epoch": 0.17431888178156835,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018543958533780437,
+      "loss": 0.3371,
+      "step": 3679
+    },
+    {
+      "epoch": 0.1743662639185027,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.000185431845860927,
+      "loss": 0.7795,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1744136460554371,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018542410448926227,
+      "loss": 0.0906,
+      "step": 3681
+    },
+    {
+      "epoch": 0.17446102819237147,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018541636122298188,
+      "loss": 0.7835,
+      "step": 3682
+    },
+    {
+      "epoch": 0.17450841032930586,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001854086160622576,
+      "loss": 0.36,
+      "step": 3683
+    },
+    {
+      "epoch": 0.17455579246624023,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018540086900726118,
+      "loss": 0.0148,
+      "step": 3684
+    },
+    {
+      "epoch": 0.1746031746031746,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018539312005816445,
+      "loss": 1.1809,
+      "step": 3685
+    },
+    {
+      "epoch": 0.174650556740109,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018538536921513927,
+      "loss": 0.9764,
+      "step": 3686
+    },
+    {
+      "epoch": 0.17469793887704335,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018537761647835754,
+      "loss": 0.9014,
+      "step": 3687
+    },
+    {
+      "epoch": 0.17474532101397772,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018536986184799123,
+      "loss": 0.2781,
+      "step": 3688
+    },
+    {
+      "epoch": 0.1747927031509121,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018536210532421233,
+      "loss": 0.8563,
+      "step": 3689
+    },
+    {
+      "epoch": 0.17484008528784648,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018535434690719285,
+      "loss": 0.0401,
+      "step": 3690
+    },
+    {
+      "epoch": 0.17488746742478087,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018534658659710487,
+      "loss": 1.1854,
+      "step": 3691
+    },
+    {
+      "epoch": 0.17493484956171523,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001853388243941205,
+      "loss": 1.3527,
+      "step": 3692
+    },
+    {
+      "epoch": 0.1749822316986496,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018533106029841188,
+      "loss": 1.0618,
+      "step": 3693
+    },
+    {
+      "epoch": 0.175029613835584,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018532329431015127,
+      "loss": 0.1632,
+      "step": 3694
+    },
+    {
+      "epoch": 0.17507699597251836,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018531552642951087,
+      "loss": 1.3013,
+      "step": 3695
+    },
+    {
+      "epoch": 0.17512437810945275,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018530775665666295,
+      "loss": 1.2847,
+      "step": 3696
+    },
+    {
+      "epoch": 0.17517176024638711,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001852999849917799,
+      "loss": 0.2259,
+      "step": 3697
+    },
+    {
+      "epoch": 0.17521914238332148,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.000185292211435034,
+      "loss": 1.0169,
+      "step": 3698
+    },
+    {
+      "epoch": 0.17526652452025587,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018528443598659768,
+      "loss": 0.9275,
+      "step": 3699
+    },
+    {
+      "epoch": 0.17531390665719024,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018527665864664344,
+      "loss": 0.1529,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1753612887941246,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018526887941534373,
+      "loss": 1.0,
+      "step": 3701
+    },
+    {
+      "epoch": 0.175408670931059,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018526109829287112,
+      "loss": 0.6171,
+      "step": 3702
+    },
+    {
+      "epoch": 0.17545605306799336,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00018525331527939818,
+      "loss": 0.1511,
+      "step": 3703
+    },
+    {
+      "epoch": 0.17550343520492775,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00018524553037509747,
+      "loss": 0.0093,
+      "step": 3704
+    },
+    {
+      "epoch": 0.17555081734186212,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018523774358014173,
+      "loss": 1.238,
+      "step": 3705
+    },
+    {
+      "epoch": 0.17559819947879649,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001852299548947036,
+      "loss": 0.6388,
+      "step": 3706
+    },
+    {
+      "epoch": 0.17564558161573088,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018522216431895587,
+      "loss": 0.8282,
+      "step": 3707
+    },
+    {
+      "epoch": 0.17569296375266524,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018521437185307132,
+      "loss": 0.5269,
+      "step": 3708
+    },
+    {
+      "epoch": 0.1757403458895996,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00018520657749722275,
+      "loss": 0.666,
+      "step": 3709
+    },
+    {
+      "epoch": 0.175787728026534,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018519878125158306,
+      "loss": 0.1391,
+      "step": 3710
+    },
+    {
+      "epoch": 0.17583511016346837,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018519098311632512,
+      "loss": 0.1767,
+      "step": 3711
+    },
+    {
+      "epoch": 0.17588249230040276,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018518318309162195,
+      "loss": 0.7675,
+      "step": 3712
+    },
+    {
+      "epoch": 0.17592987443733712,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001851753811776465,
+      "loss": 0.9559,
+      "step": 3713
+    },
+    {
+      "epoch": 0.1759772565742715,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018516757737457182,
+      "loss": 1.0426,
+      "step": 3714
+    },
+    {
+      "epoch": 0.17602463871120588,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.000185159771682571,
+      "loss": 1.6759,
+      "step": 3715
+    },
+    {
+      "epoch": 0.17607202084814025,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018515196410181714,
+      "loss": 1.2135,
+      "step": 3716
+    },
+    {
+      "epoch": 0.1761194029850746,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018514415463248343,
+      "loss": 1.1519,
+      "step": 3717
+    },
+    {
+      "epoch": 0.176166785122009,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018513634327474305,
+      "loss": 1.2489,
+      "step": 3718
+    },
+    {
+      "epoch": 0.17621416725894337,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018512853002876928,
+      "loss": 1.2299,
+      "step": 3719
+    },
+    {
+      "epoch": 0.17626154939587776,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018512071489473536,
+      "loss": 0.8664,
+      "step": 3720
+    },
+    {
+      "epoch": 0.17630893153281213,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018511289787281467,
+      "loss": 0.5929,
+      "step": 3721
+    },
+    {
+      "epoch": 0.1763563136697465,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018510507896318056,
+      "loss": 0.1134,
+      "step": 3722
+    },
+    {
+      "epoch": 0.1764036958066809,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018509725816600643,
+      "loss": 0.9311,
+      "step": 3723
+    },
+    {
+      "epoch": 0.17645107794361525,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001850894354814658,
+      "loss": 1.0983,
+      "step": 3724
+    },
+    {
+      "epoch": 0.17649846008054965,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018508161090973206,
+      "loss": 0.3915,
+      "step": 3725
+    },
+    {
+      "epoch": 0.176545842217484,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018507378445097885,
+      "loss": 0.7702,
+      "step": 3726
+    },
+    {
+      "epoch": 0.17659322435441838,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001850659561053797,
+      "loss": 0.9071,
+      "step": 3727
+    },
+    {
+      "epoch": 0.17664060649135277,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018505812587310829,
+      "loss": 1.133,
+      "step": 3728
+    },
+    {
+      "epoch": 0.17668798862828713,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001850502937543382,
+      "loss": 0.0478,
+      "step": 3729
+    },
+    {
+      "epoch": 0.1767353707652215,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018504245974924324,
+      "loss": 0.0362,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1767827529021559,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00018503462385799707,
+      "loss": 1.0033,
+      "step": 3731
+    },
+    {
+      "epoch": 0.17683013503909026,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018502678608077355,
+      "loss": 1.1681,
+      "step": 3732
+    },
+    {
+      "epoch": 0.17687751717602465,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018501894641774643,
+      "loss": 0.0484,
+      "step": 3733
+    },
+    {
+      "epoch": 0.17692489931295902,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018501110486908968,
+      "loss": 0.8182,
+      "step": 3734
+    },
+    {
+      "epoch": 0.17697228144989338,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00018500326143497715,
+      "loss": 0.9496,
+      "step": 3735
+    },
+    {
+      "epoch": 0.17701966358682777,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018499541611558283,
+      "loss": 0.1437,
+      "step": 3736
+    },
+    {
+      "epoch": 0.17706704572376214,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018498756891108072,
+      "loss": 0.1331,
+      "step": 3737
+    },
+    {
+      "epoch": 0.1771144278606965,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018497971982164483,
+      "loss": 1.0338,
+      "step": 3738
+    },
+    {
+      "epoch": 0.1771618099976309,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001849718688474493,
+      "loss": 0.0395,
+      "step": 3739
+    },
+    {
+      "epoch": 0.17720919213456526,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001849640159886682,
+      "loss": 0.9706,
+      "step": 3740
+    },
+    {
+      "epoch": 0.17725657427149966,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018495616124547578,
+      "loss": 0.9039,
+      "step": 3741
+    },
+    {
+      "epoch": 0.17730395640843402,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018494830461804617,
+      "loss": 0.09,
+      "step": 3742
+    },
+    {
+      "epoch": 0.1773513385453684,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018494044610655358,
+      "loss": 0.9534,
+      "step": 3743
+    },
+    {
+      "epoch": 0.17739872068230278,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018493258571117244,
+      "loss": 1.2597,
+      "step": 3744
+    },
+    {
+      "epoch": 0.17744610281923714,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000184924723432077,
+      "loss": 0.0954,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1774934849561715,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018491685926944165,
+      "loss": 0.0738,
+      "step": 3746
+    },
+    {
+      "epoch": 0.1775408670931059,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018490899322344082,
+      "loss": 1.2257,
+      "step": 3747
+    },
+    {
+      "epoch": 0.17758824923004027,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018490112529424897,
+      "loss": 1.2488,
+      "step": 3748
+    },
+    {
+      "epoch": 0.17763563136697466,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018489325548204057,
+      "loss": 1.0829,
+      "step": 3749
+    },
+    {
+      "epoch": 0.17768301350390903,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00018488538378699017,
+      "loss": 1.1709,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1777303956408434,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018487751020927239,
+      "loss": 0.1868,
+      "step": 3751
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018486963474906187,
+      "loss": 0.0744,
+      "step": 3752
+    },
+    {
+      "epoch": 0.17782515991471215,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018486175740653322,
+      "loss": 1.1554,
+      "step": 3753
+    },
+    {
+      "epoch": 0.17787254205164654,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001848538781818612,
+      "loss": 1.1292,
+      "step": 3754
+    },
+    {
+      "epoch": 0.1779199241885809,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018484599707522053,
+      "loss": 0.9446,
+      "step": 3755
+    },
+    {
+      "epoch": 0.17796730632551527,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018483811408678602,
+      "loss": 1.2564,
+      "step": 3756
+    },
+    {
+      "epoch": 0.17801468846244967,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018483022921673249,
+      "loss": 1.0344,
+      "step": 3757
+    },
+    {
+      "epoch": 0.17806207059938403,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00018482234246523487,
+      "loss": 1.0584,
+      "step": 3758
+    },
+    {
+      "epoch": 0.1781094527363184,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018481445383246802,
+      "loss": 0.8575,
+      "step": 3759
+    },
+    {
+      "epoch": 0.1781568348732528,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018480656331860694,
+      "loss": 0.7026,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17820421701018715,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001847986709238266,
+      "loss": 0.1739,
+      "step": 3761
+    },
+    {
+      "epoch": 0.17825159914712155,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018479077664830206,
+      "loss": 1.1396,
+      "step": 3762
+    },
+    {
+      "epoch": 0.1782989812840559,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018478288049220842,
+      "loss": 0.1417,
+      "step": 3763
+    },
+    {
+      "epoch": 0.17834636342099028,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001847749824557208,
+      "loss": 1.2362,
+      "step": 3764
+    },
+    {
+      "epoch": 0.17839374555792467,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018476708253901435,
+      "loss": 0.0794,
+      "step": 3765
+    },
+    {
+      "epoch": 0.17844112769485904,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018475918074226433,
+      "loss": 1.0855,
+      "step": 3766
+    },
+    {
+      "epoch": 0.1784885098317934,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00018475127706564593,
+      "loss": 0.002,
+      "step": 3767
+    },
+    {
+      "epoch": 0.1785358919687278,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001847433715093345,
+      "loss": 0.0393,
+      "step": 3768
+    },
+    {
+      "epoch": 0.17858327410566216,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018473546407350532,
+      "loss": 0.8788,
+      "step": 3769
+    },
+    {
+      "epoch": 0.17863065624259655,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018472755475833384,
+      "loss": 0.5201,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17867803837953092,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018471964356399546,
+      "loss": 0.8121,
+      "step": 3771
+    },
+    {
+      "epoch": 0.17872542051646528,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0001847117304906656,
+      "loss": 1.2936,
+      "step": 3772
+    },
+    {
+      "epoch": 0.17877280265339968,
+      "grad_norm": 0.02099609375,
+      "learning_rate": 0.00018470381553851976,
+      "loss": 0.0016,
+      "step": 3773
+    },
+    {
+      "epoch": 0.17882018479033404,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018469589870773355,
+      "loss": 1.0161,
+      "step": 3774
+    },
+    {
+      "epoch": 0.1788675669272684,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001846879799984825,
+      "loss": 1.019,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1789149490642028,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00018468005941094226,
+      "loss": 0.4715,
+      "step": 3776
+    },
+    {
+      "epoch": 0.17896233120113716,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018467213694528853,
+      "loss": 0.1571,
+      "step": 3777
+    },
+    {
+      "epoch": 0.17900971333807156,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018466421260169695,
+      "loss": 1.4319,
+      "step": 3778
+    },
+    {
+      "epoch": 0.17905709547500592,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018465628638034332,
+      "loss": 1.1197,
+      "step": 3779
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018464835828140347,
+      "loss": 1.3157,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17915185974887468,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018464042830505317,
+      "loss": 1.264,
+      "step": 3781
+    },
+    {
+      "epoch": 0.17919924188580905,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018463249645146834,
+      "loss": 0.2718,
+      "step": 3782
+    },
+    {
+      "epoch": 0.17924662402274344,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018462456272082487,
+      "loss": 0.7724,
+      "step": 3783
+    },
+    {
+      "epoch": 0.1792940061596778,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00018461662711329876,
+      "loss": 0.0419,
+      "step": 3784
+    },
+    {
+      "epoch": 0.17934138829661217,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00018460868962906594,
+      "loss": 0.3372,
+      "step": 3785
+    },
+    {
+      "epoch": 0.17938877043354656,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018460075026830252,
+      "loss": 1.0154,
+      "step": 3786
+    },
+    {
+      "epoch": 0.17943615257048093,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001845928090311846,
+      "loss": 0.8292,
+      "step": 3787
+    },
+    {
+      "epoch": 0.1794835347074153,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018458486591788826,
+      "loss": 0.1723,
+      "step": 3788
+    },
+    {
+      "epoch": 0.17953091684434969,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001845769209285897,
+      "loss": 0.6954,
+      "step": 3789
+    },
+    {
+      "epoch": 0.17957829898128405,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001845689740634651,
+      "loss": 0.8468,
+      "step": 3790
+    },
+    {
+      "epoch": 0.17962568111821844,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018456102532269077,
+      "loss": 1.1859,
+      "step": 3791
+    },
+    {
+      "epoch": 0.1796730632551528,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00018455307470644294,
+      "loss": 0.1165,
+      "step": 3792
+    },
+    {
+      "epoch": 0.17972044539208717,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000184545122214898,
+      "loss": 0.6493,
+      "step": 3793
+    },
+    {
+      "epoch": 0.17976782752902157,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018453716784823227,
+      "loss": 0.1474,
+      "step": 3794
+    },
+    {
+      "epoch": 0.17981520966595593,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018452921160662223,
+      "loss": 0.1255,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1798625918028903,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001845212534902443,
+      "loss": 0.6474,
+      "step": 3796
+    },
+    {
+      "epoch": 0.1799099739398247,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.000184513293499275,
+      "loss": 0.7447,
+      "step": 3797
+    },
+    {
+      "epoch": 0.17995735607675906,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018450533163389085,
+      "loss": 0.9997,
+      "step": 3798
+    },
+    {
+      "epoch": 0.18000473821369345,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018449736789426848,
+      "loss": 1.0661,
+      "step": 3799
+    },
+    {
+      "epoch": 0.18005212035062781,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001844894022805845,
+      "loss": 1.5189,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18009950248756218,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00018448143479301554,
+      "loss": 0.0322,
+      "step": 3801
+    },
+    {
+      "epoch": 0.18014688462449657,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00018447346543173836,
+      "loss": 0.9759,
+      "step": 3802
+    },
+    {
+      "epoch": 0.18019426676143094,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018446549419692972,
+      "loss": 1.4936,
+      "step": 3803
+    },
+    {
+      "epoch": 0.1802416488983653,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018445752108876633,
+      "loss": 1.0951,
+      "step": 3804
+    },
+    {
+      "epoch": 0.1802890310352997,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018444954610742512,
+      "loss": 0.0722,
+      "step": 3805
+    },
+    {
+      "epoch": 0.18033641317223406,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001844415692530829,
+      "loss": 0.0524,
+      "step": 3806
+    },
+    {
+      "epoch": 0.18038379530916845,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018443359052591665,
+      "loss": 0.0333,
+      "step": 3807
+    },
+    {
+      "epoch": 0.18043117744610282,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001844256099261033,
+      "loss": 0.6686,
+      "step": 3808
+    },
+    {
+      "epoch": 0.18047855958303718,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001844176274538198,
+      "loss": 1.0334,
+      "step": 3809
+    },
+    {
+      "epoch": 0.18052594171997158,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018440964310924328,
+      "loss": 0.979,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18057332385690594,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001844016568925508,
+      "loss": 0.6279,
+      "step": 3811
+    },
+    {
+      "epoch": 0.18062070599384034,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018439366880391943,
+      "loss": 0.0123,
+      "step": 3812
+    },
+    {
+      "epoch": 0.1806680881307747,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001843856788435264,
+      "loss": 1.1228,
+      "step": 3813
+    },
+    {
+      "epoch": 0.18071547026770907,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018437768701154891,
+      "loss": 1.2088,
+      "step": 3814
+    },
+    {
+      "epoch": 0.18076285240464346,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018436969330816417,
+      "loss": 0.9791,
+      "step": 3815
+    },
+    {
+      "epoch": 0.18081023454157782,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018436169773354953,
+      "loss": 1.1451,
+      "step": 3816
+    },
+    {
+      "epoch": 0.1808576166785122,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018435370028788226,
+      "loss": 0.9227,
+      "step": 3817
+    },
+    {
+      "epoch": 0.18090499881544658,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018434570097133977,
+      "loss": 1.1372,
+      "step": 3818
+    },
+    {
+      "epoch": 0.18095238095238095,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001843376997840995,
+      "loss": 0.8091,
+      "step": 3819
+    },
+    {
+      "epoch": 0.18099976308931534,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018432969672633887,
+      "loss": 0.1405,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1810471452262497,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018432169179823538,
+      "loss": 0.1652,
+      "step": 3821
+    },
+    {
+      "epoch": 0.18109452736318407,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018431368499996658,
+      "loss": 0.1791,
+      "step": 3822
+    },
+    {
+      "epoch": 0.18114190950011846,
+      "grad_norm": 0.0069580078125,
+      "learning_rate": 0.00018430567633171004,
+      "loss": 0.0005,
+      "step": 3823
+    },
+    {
+      "epoch": 0.18118929163705283,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018429766579364343,
+      "loss": 0.8203,
+      "step": 3824
+    },
+    {
+      "epoch": 0.1812366737739872,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018428965338594436,
+      "loss": 0.1378,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1812840559109216,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018428163910879058,
+      "loss": 1.7155,
+      "step": 3826
+    },
+    {
+      "epoch": 0.18133143804785595,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001842736229623598,
+      "loss": 0.8864,
+      "step": 3827
+    },
+    {
+      "epoch": 0.18137882018479035,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018426560494682982,
+      "loss": 1.245,
+      "step": 3828
+    },
+    {
+      "epoch": 0.1814262023217247,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018425758506237852,
+      "loss": 0.1062,
+      "step": 3829
+    },
+    {
+      "epoch": 0.18147358445865908,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018424956330918369,
+      "loss": 0.1808,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18152096659559347,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001842415396874233,
+      "loss": 0.4874,
+      "step": 3831
+    },
+    {
+      "epoch": 0.18156834873252783,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001842335141972753,
+      "loss": 1.0836,
+      "step": 3832
+    },
+    {
+      "epoch": 0.1816157308694622,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018422548683891764,
+      "loss": 0.5258,
+      "step": 3833
+    },
+    {
+      "epoch": 0.1816631130063966,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001842174576125284,
+      "loss": 0.5478,
+      "step": 3834
+    },
+    {
+      "epoch": 0.18171049514333096,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001842094265182857,
+      "loss": 0.8814,
+      "step": 3835
+    },
+    {
+      "epoch": 0.18175787728026535,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00018420139355636756,
+      "loss": 0.0369,
+      "step": 3836
+    },
+    {
+      "epoch": 0.18180525941719972,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018419335872695225,
+      "loss": 0.8465,
+      "step": 3837
+    },
+    {
+      "epoch": 0.18185264155413408,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018418532203021788,
+      "loss": 0.9397,
+      "step": 3838
+    },
+    {
+      "epoch": 0.18190002369106847,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018417728346634276,
+      "loss": 1.1063,
+      "step": 3839
+    },
+    {
+      "epoch": 0.18194740582800284,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.00018416924303550516,
+      "loss": 0.0021,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18199478796493723,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018416120073788342,
+      "loss": 0.9548,
+      "step": 3841
+    },
+    {
+      "epoch": 0.1820421701018716,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018415315657365584,
+      "loss": 0.0584,
+      "step": 3842
+    },
+    {
+      "epoch": 0.18208955223880596,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001841451105430009,
+      "loss": 0.9115,
+      "step": 3843
+    },
+    {
+      "epoch": 0.18213693437574036,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000184137062646097,
+      "loss": 1.344,
+      "step": 3844
+    },
+    {
+      "epoch": 0.18218431651267472,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018412901288312276,
+      "loss": 0.7186,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1822316986496091,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018412096125425658,
+      "loss": 0.0939,
+      "step": 3846
+    },
+    {
+      "epoch": 0.18227908078654348,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018411290775967708,
+      "loss": 1.3313,
+      "step": 3847
+    },
+    {
+      "epoch": 0.18232646292347784,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018410485239956286,
+      "loss": 0.9501,
+      "step": 3848
+    },
+    {
+      "epoch": 0.18237384506041224,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001840967951740926,
+      "loss": 0.5894,
+      "step": 3849
+    },
+    {
+      "epoch": 0.1824212271973466,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018408873608344504,
+      "loss": 0.3189,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18246860933428097,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018408067512779884,
+      "loss": 0.0554,
+      "step": 3851
+    },
+    {
+      "epoch": 0.18251599147121536,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018407261230733287,
+      "loss": 0.9996,
+      "step": 3852
+    },
+    {
+      "epoch": 0.18256337360814973,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018406454762222584,
+      "loss": 0.1408,
+      "step": 3853
+    },
+    {
+      "epoch": 0.1826107557450841,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018405648107265674,
+      "loss": 0.6393,
+      "step": 3854
+    },
+    {
+      "epoch": 0.18265813788201848,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018404841265880442,
+      "loss": 1.1064,
+      "step": 3855
+    },
+    {
+      "epoch": 0.18270552001895285,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001840403423808478,
+      "loss": 0.1576,
+      "step": 3856
+    },
+    {
+      "epoch": 0.18275290215588724,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018403227023896593,
+      "loss": 1.107,
+      "step": 3857
+    },
+    {
+      "epoch": 0.1828002842928216,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001840241962333378,
+      "loss": 0.6858,
+      "step": 3858
+    },
+    {
+      "epoch": 0.18284766642975597,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001840161203641425,
+      "loss": 0.9427,
+      "step": 3859
+    },
+    {
+      "epoch": 0.18289504856669037,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018400804263155913,
+      "loss": 0.0761,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18294243070362473,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018399996303576687,
+      "loss": 0.9487,
+      "step": 3861
+    },
+    {
+      "epoch": 0.1829898128405591,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001839918815769449,
+      "loss": 1.2183,
+      "step": 3862
+    },
+    {
+      "epoch": 0.1830371949774935,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018398379825527246,
+      "loss": 0.6553,
+      "step": 3863
+    },
+    {
+      "epoch": 0.18308457711442785,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00018397571307092881,
+      "loss": 1.3488,
+      "step": 3864
+    },
+    {
+      "epoch": 0.18313195925136225,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018396762602409332,
+      "loss": 1.0609,
+      "step": 3865
+    },
+    {
+      "epoch": 0.1831793413882966,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018395953711494533,
+      "loss": 0.2589,
+      "step": 3866
+    },
+    {
+      "epoch": 0.18322672352523098,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001839514463436642,
+      "loss": 1.275,
+      "step": 3867
+    },
+    {
+      "epoch": 0.18327410566216537,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00018394335371042943,
+      "loss": 1.2991,
+      "step": 3868
+    },
+    {
+      "epoch": 0.18332148779909974,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018393525921542048,
+      "loss": 1.434,
+      "step": 3869
+    },
+    {
+      "epoch": 0.18336886993603413,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018392716285881692,
+      "loss": 1.0181,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1834162520729685,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018391906464079822,
+      "loss": 0.7575,
+      "step": 3871
+    },
+    {
+      "epoch": 0.18346363420990286,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001839109645615441,
+      "loss": 1.0288,
+      "step": 3872
+    },
+    {
+      "epoch": 0.18351101634683725,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018390286262123416,
+      "loss": 0.3225,
+      "step": 3873
+    },
+    {
+      "epoch": 0.18355839848377162,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001838947588200481,
+      "loss": 1.5614,
+      "step": 3874
+    },
+    {
+      "epoch": 0.18360578062070598,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00018388665315816562,
+      "loss": 1.0404,
+      "step": 3875
+    },
+    {
+      "epoch": 0.18365316275764038,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018387854563576655,
+      "loss": 0.7009,
+      "step": 3876
+    },
+    {
+      "epoch": 0.18370054489457474,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018387043625303068,
+      "loss": 0.0823,
+      "step": 3877
+    },
+    {
+      "epoch": 0.18374792703150913,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018386232501013786,
+      "loss": 0.7296,
+      "step": 3878
+    },
+    {
+      "epoch": 0.1837953091684435,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000183854211907268,
+      "loss": 0.1192,
+      "step": 3879
+    },
+    {
+      "epoch": 0.18384269130537786,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00018384609694460102,
+      "loss": 0.1747,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18389007344231226,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018383798012231695,
+      "loss": 1.5189,
+      "step": 3881
+    },
+    {
+      "epoch": 0.18393745557924662,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018382986144059577,
+      "loss": 0.0617,
+      "step": 3882
+    },
+    {
+      "epoch": 0.183984837716181,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018382174089961754,
+      "loss": 1.3788,
+      "step": 3883
+    },
+    {
+      "epoch": 0.18403221985311538,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001838136184995624,
+      "loss": 1.0907,
+      "step": 3884
+    },
+    {
+      "epoch": 0.18407960199004975,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00018380549424061045,
+      "loss": 0.2911,
+      "step": 3885
+    },
+    {
+      "epoch": 0.18412698412698414,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018379736812294194,
+      "loss": 1.1701,
+      "step": 3886
+    },
+    {
+      "epoch": 0.1841743662639185,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018378924014673703,
+      "loss": 0.0224,
+      "step": 3887
+    },
+    {
+      "epoch": 0.18422174840085287,
+      "grad_norm": 0.01446533203125,
+      "learning_rate": 0.00018378111031217605,
+      "loss": 0.0009,
+      "step": 3888
+    },
+    {
+      "epoch": 0.18426913053778726,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018377297861943927,
+      "loss": 0.7568,
+      "step": 3889
+    },
+    {
+      "epoch": 0.18431651267472163,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018376484506870707,
+      "loss": 0.6754,
+      "step": 3890
+    },
+    {
+      "epoch": 0.184363894811656,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001837567096601598,
+      "loss": 1.0827,
+      "step": 3891
+    },
+    {
+      "epoch": 0.18441127694859039,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018374857239397794,
+      "loss": 1.0899,
+      "step": 3892
+    },
+    {
+      "epoch": 0.18445865908552475,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018374043327034194,
+      "loss": 1.1447,
+      "step": 3893
+    },
+    {
+      "epoch": 0.18450604122245914,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001837322922894323,
+      "loss": 0.729,
+      "step": 3894
+    },
+    {
+      "epoch": 0.1845534233593935,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018372414945142963,
+      "loss": 0.6065,
+      "step": 3895
+    },
+    {
+      "epoch": 0.18460080549632787,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00018371600475651455,
+      "loss": 0.0833,
+      "step": 3896
+    },
+    {
+      "epoch": 0.18464818763326227,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001837078582048676,
+      "loss": 1.0352,
+      "step": 3897
+    },
+    {
+      "epoch": 0.18469556977019663,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018369970979666952,
+      "loss": 0.5847,
+      "step": 3898
+    },
+    {
+      "epoch": 0.184742951907131,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018369155953210103,
+      "loss": 0.871,
+      "step": 3899
+    },
+    {
+      "epoch": 0.1847903340440654,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018368340741134294,
+      "loss": 1.4962,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18483771618099976,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018367525343457596,
+      "loss": 0.5314,
+      "step": 3901
+    },
+    {
+      "epoch": 0.18488509831793415,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000183667097601981,
+      "loss": 1.46,
+      "step": 3902
+    },
+    {
+      "epoch": 0.18493248045486851,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00018365893991373892,
+      "loss": 0.071,
+      "step": 3903
+    },
+    {
+      "epoch": 0.18497986259180288,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018365078037003069,
+      "loss": 0.9815,
+      "step": 3904
+    },
+    {
+      "epoch": 0.18502724472873727,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00018364261897103724,
+      "loss": 1.2039,
+      "step": 3905
+    },
+    {
+      "epoch": 0.18507462686567164,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001836344557169396,
+      "loss": 1.0092,
+      "step": 3906
+    },
+    {
+      "epoch": 0.18512200900260603,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018362629060791877,
+      "loss": 1.2131,
+      "step": 3907
+    },
+    {
+      "epoch": 0.1851693911395404,
+      "grad_norm": 0.01556396484375,
+      "learning_rate": 0.00018361812364415595,
+      "loss": 0.001,
+      "step": 3908
+    },
+    {
+      "epoch": 0.18521677327647476,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001836099548258322,
+      "loss": 0.0326,
+      "step": 3909
+    },
+    {
+      "epoch": 0.18526415541340915,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018360178415312867,
+      "loss": 0.6078,
+      "step": 3910
+    },
+    {
+      "epoch": 0.18531153755034352,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018359361162622662,
+      "loss": 1.2607,
+      "step": 3911
+    },
+    {
+      "epoch": 0.18535891968727788,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018358543724530737,
+      "loss": 1.049,
+      "step": 3912
+    },
+    {
+      "epoch": 0.18540630182421228,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018357726101055207,
+      "loss": 0.0309,
+      "step": 3913
+    },
+    {
+      "epoch": 0.18545368396114664,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001835690829221422,
+      "loss": 1.0746,
+      "step": 3914
+    },
+    {
+      "epoch": 0.18550106609808104,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001835609029802591,
+      "loss": 0.0499,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1855484482350154,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018355272118508414,
+      "loss": 1.2679,
+      "step": 3916
+    },
+    {
+      "epoch": 0.18559583037194977,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018354453753679882,
+      "loss": 1.0898,
+      "step": 3917
+    },
+    {
+      "epoch": 0.18564321250888416,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018353635203558467,
+      "loss": 0.6901,
+      "step": 3918
+    },
+    {
+      "epoch": 0.18569059464581852,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018352816468162318,
+      "loss": 1.3228,
+      "step": 3919
+    },
+    {
+      "epoch": 0.1857379767827529,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018351997547509598,
+      "loss": 1.2299,
+      "step": 3920
+    },
+    {
+      "epoch": 0.18578535891968728,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018351178441618467,
+      "loss": 1.2638,
+      "step": 3921
+    },
+    {
+      "epoch": 0.18583274105662165,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018350359150507095,
+      "loss": 0.7652,
+      "step": 3922
+    },
+    {
+      "epoch": 0.18588012319355604,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018349539674193652,
+      "loss": 0.0936,
+      "step": 3923
+    },
+    {
+      "epoch": 0.1859275053304904,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018348720012696312,
+      "loss": 1.1508,
+      "step": 3924
+    },
+    {
+      "epoch": 0.18597488746742477,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00018347900166033253,
+      "loss": 0.3565,
+      "step": 3925
+    },
+    {
+      "epoch": 0.18602226960435916,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018347080134222663,
+      "loss": 0.5486,
+      "step": 3926
+    },
+    {
+      "epoch": 0.18606965174129353,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00018346259917282725,
+      "loss": 0.9104,
+      "step": 3927
+    },
+    {
+      "epoch": 0.1861170338782279,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018345439515231633,
+      "loss": 1.1383,
+      "step": 3928
+    },
+    {
+      "epoch": 0.1861644160151623,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00018344618928087584,
+      "loss": 0.5591,
+      "step": 3929
+    },
+    {
+      "epoch": 0.18621179815209665,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001834379815586877,
+      "loss": 0.1086,
+      "step": 3930
+    },
+    {
+      "epoch": 0.18625918028903105,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018342977198593404,
+      "loss": 0.0302,
+      "step": 3931
+    },
+    {
+      "epoch": 0.1863065624259654,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018342156056279686,
+      "loss": 0.8353,
+      "step": 3932
+    },
+    {
+      "epoch": 0.18635394456289978,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0001834133472894584,
+      "loss": 0.007,
+      "step": 3933
+    },
+    {
+      "epoch": 0.18640132669983417,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018340513216610066,
+      "loss": 1.2658,
+      "step": 3934
+    },
+    {
+      "epoch": 0.18644870883676853,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018339691519290597,
+      "loss": 0.1188,
+      "step": 3935
+    },
+    {
+      "epoch": 0.18649609097370293,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00018338869637005652,
+      "loss": 0.4885,
+      "step": 3936
+    },
+    {
+      "epoch": 0.1865434731106373,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018338047569773464,
+      "loss": 1.4999,
+      "step": 3937
+    },
+    {
+      "epoch": 0.18659085524757166,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018337225317612262,
+      "loss": 0.925,
+      "step": 3938
+    },
+    {
+      "epoch": 0.18663823738450605,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018336402880540277,
+      "loss": 0.9007,
+      "step": 3939
+    },
+    {
+      "epoch": 0.18668561952144042,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018335580258575763,
+      "loss": 1.3914,
+      "step": 3940
+    },
+    {
+      "epoch": 0.18673300165837478,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018334757451736951,
+      "loss": 1.3489,
+      "step": 3941
+    },
+    {
+      "epoch": 0.18678038379530917,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000183339344600421,
+      "loss": 0.9137,
+      "step": 3942
+    },
+    {
+      "epoch": 0.18682776593224354,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018333111283509458,
+      "loss": 0.9077,
+      "step": 3943
+    },
+    {
+      "epoch": 0.18687514806917793,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018332287922157288,
+      "loss": 0.8801,
+      "step": 3944
+    },
+    {
+      "epoch": 0.1869225302061123,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018331464376003843,
+      "loss": 0.5744,
+      "step": 3945
+    },
+    {
+      "epoch": 0.18696991234304666,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018330640645067392,
+      "loss": 0.6218,
+      "step": 3946
+    },
+    {
+      "epoch": 0.18701729447998106,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00018329816729366206,
+      "loss": 0.5785,
+      "step": 3947
+    },
+    {
+      "epoch": 0.18706467661691542,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018328992628918557,
+      "loss": 0.4056,
+      "step": 3948
+    },
+    {
+      "epoch": 0.18711205875384979,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001832816834374272,
+      "loss": 0.8687,
+      "step": 3949
+    },
+    {
+      "epoch": 0.18715944089078418,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018327343873856985,
+      "loss": 0.1306,
+      "step": 3950
+    },
+    {
+      "epoch": 0.18720682302771854,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018326519219279632,
+      "loss": 1.1476,
+      "step": 3951
+    },
+    {
+      "epoch": 0.18725420516465294,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018325694380028952,
+      "loss": 0.6137,
+      "step": 3952
+    },
+    {
+      "epoch": 0.1873015873015873,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018324869356123238,
+      "loss": 0.6959,
+      "step": 3953
+    },
+    {
+      "epoch": 0.18734896943852167,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001832404414758079,
+      "loss": 0.6488,
+      "step": 3954
+    },
+    {
+      "epoch": 0.18739635157545606,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00018323218754419905,
+      "loss": 0.7129,
+      "step": 3955
+    },
+    {
+      "epoch": 0.18744373371239043,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018322393176658898,
+      "loss": 0.9986,
+      "step": 3956
+    },
+    {
+      "epoch": 0.1874911158493248,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018321567414316073,
+      "loss": 0.7868,
+      "step": 3957
+    },
+    {
+      "epoch": 0.18753849798625918,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018320741467409748,
+      "loss": 0.3172,
+      "step": 3958
+    },
+    {
+      "epoch": 0.18758588012319355,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001831991533595824,
+      "loss": 1.3278,
+      "step": 3959
+    },
+    {
+      "epoch": 0.18763326226012794,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018319089019979868,
+      "loss": 0.6262,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1876806443970623,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018318262519492965,
+      "loss": 1.1593,
+      "step": 3961
+    },
+    {
+      "epoch": 0.18772802653399667,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00018317435834515862,
+      "loss": 0.0484,
+      "step": 3962
+    },
+    {
+      "epoch": 0.18777540867093107,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018316608965066887,
+      "loss": 0.9068,
+      "step": 3963
+    },
+    {
+      "epoch": 0.18782279080786543,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018315781911164386,
+      "loss": 0.1292,
+      "step": 3964
+    },
+    {
+      "epoch": 0.18787017294479982,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018314954672826703,
+      "loss": 0.2688,
+      "step": 3965
+    },
+    {
+      "epoch": 0.1879175550817342,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018314127250072178,
+      "loss": 1.1685,
+      "step": 3966
+    },
+    {
+      "epoch": 0.18796493721866855,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018313299642919167,
+      "loss": 1.2903,
+      "step": 3967
+    },
+    {
+      "epoch": 0.18801231935560295,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018312471851386026,
+      "loss": 0.9526,
+      "step": 3968
+    },
+    {
+      "epoch": 0.1880597014925373,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018311643875491113,
+      "loss": 1.6321,
+      "step": 3969
+    },
+    {
+      "epoch": 0.18810708362947168,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018310815715252788,
+      "loss": 1.1504,
+      "step": 3970
+    },
+    {
+      "epoch": 0.18815446576640607,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018309987370689428,
+      "loss": 0.6617,
+      "step": 3971
+    },
+    {
+      "epoch": 0.18820184790334044,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018309158841819397,
+      "loss": 0.7718,
+      "step": 3972
+    },
+    {
+      "epoch": 0.18824923004027483,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018308330128661074,
+      "loss": 0.8172,
+      "step": 3973
+    },
+    {
+      "epoch": 0.1882966121772092,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001830750123123284,
+      "loss": 0.7052,
+      "step": 3974
+    },
+    {
+      "epoch": 0.18834399431414356,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018306672149553076,
+      "loss": 1.2497,
+      "step": 3975
+    },
+    {
+      "epoch": 0.18839137645107795,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001830584288364017,
+      "loss": 0.7786,
+      "step": 3976
+    },
+    {
+      "epoch": 0.18843875858801232,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001830501343351252,
+      "loss": 0.1944,
+      "step": 3977
+    },
+    {
+      "epoch": 0.18848614072494668,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018304183799188514,
+      "loss": 0.043,
+      "step": 3978
+    },
+    {
+      "epoch": 0.18853352286188108,
+      "grad_norm": 0.0064697265625,
+      "learning_rate": 0.00018303353980686558,
+      "loss": 0.0004,
+      "step": 3979
+    },
+    {
+      "epoch": 0.18858090499881544,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018302523978025058,
+      "loss": 0.1154,
+      "step": 3980
+    },
+    {
+      "epoch": 0.18862828713574983,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018301693791222413,
+      "loss": 0.0054,
+      "step": 3981
+    },
+    {
+      "epoch": 0.1886756692726842,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001830086342029705,
+      "loss": 0.6819,
+      "step": 3982
+    },
+    {
+      "epoch": 0.18872305140961856,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001830003286526737,
+      "loss": 0.9113,
+      "step": 3983
+    },
+    {
+      "epoch": 0.18877043354655296,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018299202126151808,
+      "loss": 1.3575,
+      "step": 3984
+    },
+    {
+      "epoch": 0.18881781568348732,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018298371202968782,
+      "loss": 0.74,
+      "step": 3985
+    },
+    {
+      "epoch": 0.1888651978204217,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001829754009573672,
+      "loss": 1.3455,
+      "step": 3986
+    },
+    {
+      "epoch": 0.18891257995735608,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00018296708804474058,
+      "loss": 0.1818,
+      "step": 3987
+    },
+    {
+      "epoch": 0.18895996209429045,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0001829587732919923,
+      "loss": 0.0035,
+      "step": 3988
+    },
+    {
+      "epoch": 0.18900734423122484,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018295045669930682,
+      "loss": 0.9317,
+      "step": 3989
+    },
+    {
+      "epoch": 0.1890547263681592,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018294213826686853,
+      "loss": 1.188,
+      "step": 3990
+    },
+    {
+      "epoch": 0.18910210850509357,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000182933817994862,
+      "loss": 0.7484,
+      "step": 3991
+    },
+    {
+      "epoch": 0.18914949064202796,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018292549588347167,
+      "loss": 1.4723,
+      "step": 3992
+    },
+    {
+      "epoch": 0.18919687277896233,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018291717193288224,
+      "loss": 0.9236,
+      "step": 3993
+    },
+    {
+      "epoch": 0.18924425491589672,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018290884614327818,
+      "loss": 0.9172,
+      "step": 3994
+    },
+    {
+      "epoch": 0.18929163705283109,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001829005185148443,
+      "loss": 0.6576,
+      "step": 3995
+    },
+    {
+      "epoch": 0.18933901918976545,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018289218904776516,
+      "loss": 0.8123,
+      "step": 3996
+    },
+    {
+      "epoch": 0.18938640132669984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018288385774222558,
+      "loss": 1.4964,
+      "step": 3997
+    },
+    {
+      "epoch": 0.1894337834636342,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00018287552459841033,
+      "loss": 0.0254,
+      "step": 3998
+    },
+    {
+      "epoch": 0.18948116560056857,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018286718961650422,
+      "loss": 1.1917,
+      "step": 3999
+    },
+    {
+      "epoch": 0.18952854773750297,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018285885279669209,
+      "loss": 0.7383,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18957592987443733,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018285051413915885,
+      "loss": 1.2202,
+      "step": 4001
+    },
+    {
+      "epoch": 0.18962331201137173,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001828421736440895,
+      "loss": 0.1662,
+      "step": 4002
+    },
+    {
+      "epoch": 0.1896706941483061,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018283383131166895,
+      "loss": 0.8653,
+      "step": 4003
+    },
+    {
+      "epoch": 0.18971807628524046,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018282548714208227,
+      "loss": 1.4906,
+      "step": 4004
+    },
+    {
+      "epoch": 0.18976545842217485,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018281714113551448,
+      "loss": 0.043,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1898128405591092,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018280879329215076,
+      "loss": 0.642,
+      "step": 4006
+    },
+    {
+      "epoch": 0.18986022269604358,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018280044361217615,
+      "loss": 0.9417,
+      "step": 4007
+    },
+    {
+      "epoch": 0.18990760483297797,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018279209209577592,
+      "loss": 1.0475,
+      "step": 4008
+    },
+    {
+      "epoch": 0.18995498696991234,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001827837387431353,
+      "loss": 1.0232,
+      "step": 4009
+    },
+    {
+      "epoch": 0.19000236910684673,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001827753835544395,
+      "loss": 0.005,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1900497512437811,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018276702652987389,
+      "loss": 1.1849,
+      "step": 4011
+    },
+    {
+      "epoch": 0.19009713338071546,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00018275866766962376,
+      "loss": 0.1887,
+      "step": 4012
+    },
+    {
+      "epoch": 0.19014451551764985,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001827503069738745,
+      "loss": 1.0838,
+      "step": 4013
+    },
+    {
+      "epoch": 0.19019189765458422,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018274194444281165,
+      "loss": 1.0973,
+      "step": 4014
+    },
+    {
+      "epoch": 0.19023927979151858,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018273358007662055,
+      "loss": 0.9148,
+      "step": 4015
+    },
+    {
+      "epoch": 0.19028666192845298,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001827252138754868,
+      "loss": 0.805,
+      "step": 4016
+    },
+    {
+      "epoch": 0.19033404406538734,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001827168458395959,
+      "loss": 0.6369,
+      "step": 4017
+    },
+    {
+      "epoch": 0.19038142620232174,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001827084759691335,
+      "loss": 1.0181,
+      "step": 4018
+    },
+    {
+      "epoch": 0.1904288083392561,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018270010426428516,
+      "loss": 0.5452,
+      "step": 4019
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018269173072523663,
+      "loss": 1.1883,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19052357261312486,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018268335535217355,
+      "loss": 1.6229,
+      "step": 4021
+    },
+    {
+      "epoch": 0.19057095475005922,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018267497814528175,
+      "loss": 0.7858,
+      "step": 4022
+    },
+    {
+      "epoch": 0.19061833688699362,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018266659910474696,
+      "loss": 1.03,
+      "step": 4023
+    },
+    {
+      "epoch": 0.19066571902392798,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018265821823075512,
+      "loss": 0.8899,
+      "step": 4024
+    },
+    {
+      "epoch": 0.19071310116086235,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018264983552349198,
+      "loss": 0.2098,
+      "step": 4025
+    },
+    {
+      "epoch": 0.19076048329779674,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018264145098314356,
+      "loss": 0.0286,
+      "step": 4026
+    },
+    {
+      "epoch": 0.1908078654347311,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018263306460989575,
+      "loss": 0.5669,
+      "step": 4027
+    },
+    {
+      "epoch": 0.19085524757166547,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00018262467640393462,
+      "loss": 0.008,
+      "step": 4028
+    },
+    {
+      "epoch": 0.19090262970859986,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018261628636544614,
+      "loss": 0.5928,
+      "step": 4029
+    },
+    {
+      "epoch": 0.19095001184553423,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018260789449461645,
+      "loss": 0.8542,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19099739398246862,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00018259950079163164,
+      "loss": 0.9255,
+      "step": 4031
+    },
+    {
+      "epoch": 0.191044776119403,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001825911052566779,
+      "loss": 1.2408,
+      "step": 4032
+    },
+    {
+      "epoch": 0.19109215825633735,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018258270788994142,
+      "loss": 1.4613,
+      "step": 4033
+    },
+    {
+      "epoch": 0.19113954039327175,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018257430869160843,
+      "loss": 1.1682,
+      "step": 4034
+    },
+    {
+      "epoch": 0.1911869225302061,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00018256590766186522,
+      "loss": 1.0742,
+      "step": 4035
+    },
+    {
+      "epoch": 0.19123430466714048,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018255750480089812,
+      "loss": 0.4825,
+      "step": 4036
+    },
+    {
+      "epoch": 0.19128168680407487,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018254910010889354,
+      "loss": 0.9041,
+      "step": 4037
+    },
+    {
+      "epoch": 0.19132906894100923,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001825406935860378,
+      "loss": 1.0246,
+      "step": 4038
+    },
+    {
+      "epoch": 0.19137645107794363,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001825322852325174,
+      "loss": 1.1757,
+      "step": 4039
+    },
+    {
+      "epoch": 0.191423833214878,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018252387504851883,
+      "loss": 1.2713,
+      "step": 4040
+    },
+    {
+      "epoch": 0.19147121535181236,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018251546303422865,
+      "loss": 1.3427,
+      "step": 4041
+    },
+    {
+      "epoch": 0.19151859748874675,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018250704918983335,
+      "loss": 1.2905,
+      "step": 4042
+    },
+    {
+      "epoch": 0.19156597962568112,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018249863351551957,
+      "loss": 0.9729,
+      "step": 4043
+    },
+    {
+      "epoch": 0.19161336176261548,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000182490216011474,
+      "loss": 0.6328,
+      "step": 4044
+    },
+    {
+      "epoch": 0.19166074389954987,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001824817966778833,
+      "loss": 0.0854,
+      "step": 4045
+    },
+    {
+      "epoch": 0.19170812603648424,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0001824733755149342,
+      "loss": 0.0029,
+      "step": 4046
+    },
+    {
+      "epoch": 0.19175550817341863,
+      "grad_norm": 0.01312255859375,
+      "learning_rate": 0.0001824649525228135,
+      "loss": 0.0006,
+      "step": 4047
+    },
+    {
+      "epoch": 0.191802890310353,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018245652770170794,
+      "loss": 0.3941,
+      "step": 4048
+    },
+    {
+      "epoch": 0.19185027244728736,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018244810105180444,
+      "loss": 1.1245,
+      "step": 4049
+    },
+    {
+      "epoch": 0.19189765458422176,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001824396725732899,
+      "loss": 0.1825,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19194503672115612,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018243124226635123,
+      "loss": 0.8779,
+      "step": 4051
+    },
+    {
+      "epoch": 0.1919924188580905,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001824228101311754,
+      "loss": 0.842,
+      "step": 4052
+    },
+    {
+      "epoch": 0.19203980099502488,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018241437616794946,
+      "loss": 0.1626,
+      "step": 4053
+    },
+    {
+      "epoch": 0.19208718313195924,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001824059403768604,
+      "loss": 0.7362,
+      "step": 4054
+    },
+    {
+      "epoch": 0.19213456526889364,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018239750275809538,
+      "loss": 1.2624,
+      "step": 4055
+    },
+    {
+      "epoch": 0.192181947405828,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001823890633118415,
+      "loss": 0.8972,
+      "step": 4056
+    },
+    {
+      "epoch": 0.19222932954276237,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018238062203828598,
+      "loss": 0.5408,
+      "step": 4057
+    },
+    {
+      "epoch": 0.19227671167969676,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018237217893761598,
+      "loss": 1.1592,
+      "step": 4058
+    },
+    {
+      "epoch": 0.19232409381663113,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018236373401001878,
+      "loss": 1.0925,
+      "step": 4059
+    },
+    {
+      "epoch": 0.19237147595356552,
+      "grad_norm": 0.01043701171875,
+      "learning_rate": 0.00018235528725568174,
+      "loss": 0.0004,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19241885809049988,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001823468386747921,
+      "loss": 1.0052,
+      "step": 4061
+    },
+    {
+      "epoch": 0.19246624022743425,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018233838826753733,
+      "loss": 0.3306,
+      "step": 4062
+    },
+    {
+      "epoch": 0.19251362236436864,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018232993603410478,
+      "loss": 0.1146,
+      "step": 4063
+    },
+    {
+      "epoch": 0.192561004501303,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018232148197468194,
+      "loss": 1.229,
+      "step": 4064
+    },
+    {
+      "epoch": 0.19260838663823737,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00018231302608945636,
+      "loss": 0.1671,
+      "step": 4065
+    },
+    {
+      "epoch": 0.19265576877517177,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0001823045683786155,
+      "loss": 0.0051,
+      "step": 4066
+    },
+    {
+      "epoch": 0.19270315091210613,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018229610884234698,
+      "loss": 0.7844,
+      "step": 4067
+    },
+    {
+      "epoch": 0.19275053304904052,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018228764748083847,
+      "loss": 0.9091,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1927979151859749,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018227918429427752,
+      "loss": 0.7842,
+      "step": 4069
+    },
+    {
+      "epoch": 0.19284529732290925,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018227071928285192,
+      "loss": 1.0486,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19289267945984365,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018226225244674944,
+      "loss": 0.0681,
+      "step": 4071
+    },
+    {
+      "epoch": 0.192940061596778,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00018225378378615778,
+      "loss": 0.0134,
+      "step": 4072
+    },
+    {
+      "epoch": 0.19298744373371238,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018224531330126483,
+      "loss": 0.9821,
+      "step": 4073
+    },
+    {
+      "epoch": 0.19303482587064677,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00018223684099225843,
+      "loss": 0.246,
+      "step": 4074
+    },
+    {
+      "epoch": 0.19308220800758114,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018222836685932646,
+      "loss": 1.0779,
+      "step": 4075
+    },
+    {
+      "epoch": 0.19312959014451553,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018221989090265694,
+      "loss": 0.4981,
+      "step": 4076
+    },
+    {
+      "epoch": 0.1931769722814499,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001822114131224378,
+      "loss": 0.8945,
+      "step": 4077
+    },
+    {
+      "epoch": 0.19322435441838426,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001822029335188571,
+      "loss": 0.3565,
+      "step": 4078
+    },
+    {
+      "epoch": 0.19327173655531865,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018219445209210287,
+      "loss": 1.3734,
+      "step": 4079
+    },
+    {
+      "epoch": 0.19331911869225302,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018218596884236326,
+      "loss": 1.1543,
+      "step": 4080
+    },
+    {
+      "epoch": 0.1933665008291874,
+      "grad_norm": 0.0107421875,
+      "learning_rate": 0.00018217748376982636,
+      "loss": 0.0005,
+      "step": 4081
+    },
+    {
+      "epoch": 0.19341388296612178,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018216899687468048,
+      "loss": 0.8825,
+      "step": 4082
+    },
+    {
+      "epoch": 0.19346126510305614,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001821605081571137,
+      "loss": 1.0132,
+      "step": 4083
+    },
+    {
+      "epoch": 0.19350864723999053,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001821520176173144,
+      "loss": 1.0012,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1935560293769249,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018214352525547085,
+      "loss": 1.0989,
+      "step": 4085
+    },
+    {
+      "epoch": 0.19360341151385926,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001821350310717714,
+      "loss": 0.8541,
+      "step": 4086
+    },
+    {
+      "epoch": 0.19365079365079366,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018212653506640442,
+      "loss": 0.8196,
+      "step": 4087
+    },
+    {
+      "epoch": 0.19369817578772802,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001821180372395584,
+      "loss": 0.0037,
+      "step": 4088
+    },
+    {
+      "epoch": 0.19374555792466241,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018210953759142178,
+      "loss": 0.2064,
+      "step": 4089
+    },
+    {
+      "epoch": 0.19379294006159678,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00018210103612218307,
+      "loss": 0.0975,
+      "step": 4090
+    },
+    {
+      "epoch": 0.19384032219853115,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018209253283203082,
+      "loss": 0.7448,
+      "step": 4091
+    },
+    {
+      "epoch": 0.19388770433546554,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018208402772115366,
+      "loss": 0.0596,
+      "step": 4092
+    },
+    {
+      "epoch": 0.1939350864723999,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018207552078974018,
+      "loss": 0.0836,
+      "step": 4093
+    },
+    {
+      "epoch": 0.19398246860933427,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018206701203797907,
+      "loss": 0.9699,
+      "step": 4094
+    },
+    {
+      "epoch": 0.19402985074626866,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018205850146605905,
+      "loss": 0.9955,
+      "step": 4095
+    },
+    {
+      "epoch": 0.19407723288320303,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018204998907416888,
+      "loss": 1.0723,
+      "step": 4096
+    },
+    {
+      "epoch": 0.19412461502013742,
+      "grad_norm": 0.006317138671875,
+      "learning_rate": 0.00018204147486249733,
+      "loss": 0.0005,
+      "step": 4097
+    },
+    {
+      "epoch": 0.19417199715707179,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001820329588312333,
+      "loss": 1.2173,
+      "step": 4098
+    },
+    {
+      "epoch": 0.19421937929400615,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00018202444098056554,
+      "loss": 0.1614,
+      "step": 4099
+    },
+    {
+      "epoch": 0.19426676143094054,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018201592131068312,
+      "loss": 0.4065,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1943141435678749,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018200739982177487,
+      "loss": 0.0372,
+      "step": 4101
+    },
+    {
+      "epoch": 0.19436152570480927,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018199887651402987,
+      "loss": 0.4186,
+      "step": 4102
+    },
+    {
+      "epoch": 0.19440890784174367,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018199035138763712,
+      "loss": 0.9015,
+      "step": 4103
+    },
+    {
+      "epoch": 0.19445628997867803,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018198182444278572,
+      "loss": 1.0228,
+      "step": 4104
+    },
+    {
+      "epoch": 0.19450367211561242,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001819732956796648,
+      "loss": 1.0538,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1945510542525468,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018196476509846346,
+      "loss": 1.1581,
+      "step": 4106
+    },
+    {
+      "epoch": 0.19459843638948116,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018195623269937094,
+      "loss": 0.9356,
+      "step": 4107
+    },
+    {
+      "epoch": 0.19464581852641555,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018194769848257648,
+      "loss": 0.9523,
+      "step": 4108
+    },
+    {
+      "epoch": 0.1946932006633499,
+      "grad_norm": 0.0194091796875,
+      "learning_rate": 0.00018193916244826938,
+      "loss": 0.0006,
+      "step": 4109
+    },
+    {
+      "epoch": 0.1947405828002843,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001819306245966389,
+      "loss": 1.0516,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19478796493721867,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00018192208492787444,
+      "loss": 0.8879,
+      "step": 4111
+    },
+    {
+      "epoch": 0.19483534707415304,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001819135434421654,
+      "loss": 0.6391,
+      "step": 4112
+    },
+    {
+      "epoch": 0.19488272921108743,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001819050001397012,
+      "loss": 0.0754,
+      "step": 4113
+    },
+    {
+      "epoch": 0.1949301113480218,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001818964550206714,
+      "loss": 1.0995,
+      "step": 4114
+    },
+    {
+      "epoch": 0.19497749348495616,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001818879080852654,
+      "loss": 0.9117,
+      "step": 4115
+    },
+    {
+      "epoch": 0.19502487562189055,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018187935933367281,
+      "loss": 0.6531,
+      "step": 4116
+    },
+    {
+      "epoch": 0.19507225775882492,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00018187080876608328,
+      "loss": 0.0202,
+      "step": 4117
+    },
+    {
+      "epoch": 0.1951196398957593,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018186225638268643,
+      "loss": 0.0461,
+      "step": 4118
+    },
+    {
+      "epoch": 0.19516702203269368,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018185370218367188,
+      "loss": 0.8954,
+      "step": 4119
+    },
+    {
+      "epoch": 0.19521440416962804,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018184514616922945,
+      "loss": 0.0695,
+      "step": 4120
+    },
+    {
+      "epoch": 0.19526178630656243,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018183658833954884,
+      "loss": 0.4098,
+      "step": 4121
+    },
+    {
+      "epoch": 0.1953091684434968,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018182802869481986,
+      "loss": 1.5883,
+      "step": 4122
+    },
+    {
+      "epoch": 0.19535655058043117,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018181946723523236,
+      "loss": 1.1525,
+      "step": 4123
+    },
+    {
+      "epoch": 0.19540393271736556,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018181090396097625,
+      "loss": 1.054,
+      "step": 4124
+    },
+    {
+      "epoch": 0.19545131485429992,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018180233887224142,
+      "loss": 0.6363,
+      "step": 4125
+    },
+    {
+      "epoch": 0.19549869699123432,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001817937719692178,
+      "loss": 0.5821,
+      "step": 4126
+    },
+    {
+      "epoch": 0.19554607912816868,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001817852032520955,
+      "loss": 0.2442,
+      "step": 4127
+    },
+    {
+      "epoch": 0.19559346126510305,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018177663272106448,
+      "loss": 0.5642,
+      "step": 4128
+    },
+    {
+      "epoch": 0.19564084340203744,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018176806037631485,
+      "loss": 0.5534,
+      "step": 4129
+    },
+    {
+      "epoch": 0.1956882255389718,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018175948621803676,
+      "loss": 0.7327,
+      "step": 4130
+    },
+    {
+      "epoch": 0.19573560767590617,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00018175091024642034,
+      "loss": 0.142,
+      "step": 4131
+    },
+    {
+      "epoch": 0.19578298981284056,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001817423324616558,
+      "loss": 0.7017,
+      "step": 4132
+    },
+    {
+      "epoch": 0.19583037194977493,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001817337528639334,
+      "loss": 0.0416,
+      "step": 4133
+    },
+    {
+      "epoch": 0.19587775408670932,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018172517145344341,
+      "loss": 0.1415,
+      "step": 4134
+    },
+    {
+      "epoch": 0.1959251362236437,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018171658823037617,
+      "loss": 1.0319,
+      "step": 4135
+    },
+    {
+      "epoch": 0.19597251836057805,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018170800319492204,
+      "loss": 0.9732,
+      "step": 4136
+    },
+    {
+      "epoch": 0.19601990049751244,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018169941634727142,
+      "loss": 1.0862,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1960672826344468,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001816908276876148,
+      "loss": 0.0289,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1961146647713812,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018168223721614263,
+      "loss": 0.6891,
+      "step": 4139
+    },
+    {
+      "epoch": 0.19616204690831557,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018167364493304545,
+      "loss": 1.2633,
+      "step": 4140
+    },
+    {
+      "epoch": 0.19620942904524993,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001816650508385138,
+      "loss": 0.9559,
+      "step": 4141
+    },
+    {
+      "epoch": 0.19625681118218433,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001816564549327383,
+      "loss": 1.3139,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1963041933191187,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018164785721590961,
+      "loss": 1.0766,
+      "step": 4143
+    },
+    {
+      "epoch": 0.19635157545605306,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018163925768821843,
+      "loss": 0.7621,
+      "step": 4144
+    },
+    {
+      "epoch": 0.19639895759298745,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00018163065634985547,
+      "loss": 1.3806,
+      "step": 4145
+    },
+    {
+      "epoch": 0.19644633972992182,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001816220532010115,
+      "loss": 1.0591,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1964937218668562,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018161344824187733,
+      "loss": 1.3723,
+      "step": 4147
+    },
+    {
+      "epoch": 0.19654110400379057,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001816048414726438,
+      "loss": 0.2652,
+      "step": 4148
+    },
+    {
+      "epoch": 0.19658848614072494,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00018159623289350183,
+      "loss": 1.0016,
+      "step": 4149
+    },
+    {
+      "epoch": 0.19663586827765933,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001815876225046423,
+      "loss": 0.0529,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1966832504145937,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018157901030625625,
+      "loss": 0.7826,
+      "step": 4151
+    },
+    {
+      "epoch": 0.19673063255152806,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018157039629853458,
+      "loss": 0.7337,
+      "step": 4152
+    },
+    {
+      "epoch": 0.19677801468846245,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00018156178048166842,
+      "loss": 1.1036,
+      "step": 4153
+    },
+    {
+      "epoch": 0.19682539682539682,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018155316285584886,
+      "loss": 0.751,
+      "step": 4154
+    },
+    {
+      "epoch": 0.1968727789623312,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018154454342126703,
+      "loss": 1.113,
+      "step": 4155
+    },
+    {
+      "epoch": 0.19692016109926558,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00018153592217811407,
+      "loss": 1.2036,
+      "step": 4156
+    },
+    {
+      "epoch": 0.19696754323619994,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018152729912658118,
+      "loss": 0.1611,
+      "step": 4157
+    },
+    {
+      "epoch": 0.19701492537313434,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001815186742668597,
+      "loss": 0.1472,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1970623075100687,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001815100475991408,
+      "loss": 1.2335,
+      "step": 4159
+    },
+    {
+      "epoch": 0.19710968964700307,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018150141912361587,
+      "loss": 0.9885,
+      "step": 4160
+    },
+    {
+      "epoch": 0.19715707178393746,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001814927888404763,
+      "loss": 0.9222,
+      "step": 4161
+    },
+    {
+      "epoch": 0.19720445392087183,
+      "grad_norm": 0.0185546875,
+      "learning_rate": 0.00018148415674991344,
+      "loss": 0.0008,
+      "step": 4162
+    },
+    {
+      "epoch": 0.19725183605780622,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001814755228521188,
+      "loss": 0.2763,
+      "step": 4163
+    },
+    {
+      "epoch": 0.19729921819474058,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018146688714728386,
+      "loss": 0.9484,
+      "step": 4164
+    },
+    {
+      "epoch": 0.19734660033167495,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001814582496356001,
+      "loss": 1.0825,
+      "step": 4165
+    },
+    {
+      "epoch": 0.19739398246860934,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018144961031725916,
+      "loss": 1.1818,
+      "step": 4166
+    },
+    {
+      "epoch": 0.1974413646055437,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001814409691924526,
+      "loss": 1.0338,
+      "step": 4167
+    },
+    {
+      "epoch": 0.1974887467424781,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0001814323262613721,
+      "loss": 0.0489,
+      "step": 4168
+    },
+    {
+      "epoch": 0.19753612887941246,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018142368152420933,
+      "loss": 0.9375,
+      "step": 4169
+    },
+    {
+      "epoch": 0.19758351101634683,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018141503498115603,
+      "loss": 1.2822,
+      "step": 4170
+    },
+    {
+      "epoch": 0.19763089315328122,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018140638663240402,
+      "loss": 1.137,
+      "step": 4171
+    },
+    {
+      "epoch": 0.1976782752902156,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00018139773647814501,
+      "loss": 1.0599,
+      "step": 4172
+    },
+    {
+      "epoch": 0.19772565742714995,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018138908451857094,
+      "loss": 0.2821,
+      "step": 4173
+    },
+    {
+      "epoch": 0.19777303956408435,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00018138043075387365,
+      "loss": 1.3287,
+      "step": 4174
+    },
+    {
+      "epoch": 0.1978204217010187,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018137177518424508,
+      "loss": 0.9248,
+      "step": 4175
+    },
+    {
+      "epoch": 0.1978678038379531,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018136311780987723,
+      "loss": 0.8726,
+      "step": 4176
+    },
+    {
+      "epoch": 0.19791518597488747,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018135445863096206,
+      "loss": 0.7506,
+      "step": 4177
+    },
+    {
+      "epoch": 0.19796256811182183,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018134579764769168,
+      "loss": 0.2422,
+      "step": 4178
+    },
+    {
+      "epoch": 0.19800995024875623,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00018133713486025816,
+      "loss": 1.2308,
+      "step": 4179
+    },
+    {
+      "epoch": 0.1980573323856906,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001813284702688536,
+      "loss": 0.961,
+      "step": 4180
+    },
+    {
+      "epoch": 0.19810471452262496,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018131980387367023,
+      "loss": 1.2177,
+      "step": 4181
+    },
+    {
+      "epoch": 0.19815209665955935,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018131113567490021,
+      "loss": 0.0505,
+      "step": 4182
+    },
+    {
+      "epoch": 0.19819947879649372,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001813024656727358,
+      "loss": 0.7928,
+      "step": 4183
+    },
+    {
+      "epoch": 0.1982468609334281,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00018129379386736932,
+      "loss": 0.4456,
+      "step": 4184
+    },
+    {
+      "epoch": 0.19829424307036247,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018128512025899305,
+      "loss": 0.2851,
+      "step": 4185
+    },
+    {
+      "epoch": 0.19834162520729684,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00018127644484779943,
+      "loss": 0.0022,
+      "step": 4186
+    },
+    {
+      "epoch": 0.19838900734423123,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001812677676339808,
+      "loss": 0.8009,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1984363894811656,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018125908861772966,
+      "loss": 0.9012,
+      "step": 4188
+    },
+    {
+      "epoch": 0.19848377161809996,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001812504077992385,
+      "loss": 0.9618,
+      "step": 4189
+    },
+    {
+      "epoch": 0.19853115375503436,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018124172517869984,
+      "loss": 0.1837,
+      "step": 4190
+    },
+    {
+      "epoch": 0.19857853589196872,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018123304075630625,
+      "loss": 1.0119,
+      "step": 4191
+    },
+    {
+      "epoch": 0.19862591802890311,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018122435453225033,
+      "loss": 1.0928,
+      "step": 4192
+    },
+    {
+      "epoch": 0.19867330016583748,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018121566650672474,
+      "loss": 0.8389,
+      "step": 4193
+    },
+    {
+      "epoch": 0.19872068230277184,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018120697667992216,
+      "loss": 1.0344,
+      "step": 4194
+    },
+    {
+      "epoch": 0.19876806443970624,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018119828505203537,
+      "loss": 1.2736,
+      "step": 4195
+    },
+    {
+      "epoch": 0.1988154465766406,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018118959162325711,
+      "loss": 0.4803,
+      "step": 4196
+    },
+    {
+      "epoch": 0.19886282871357497,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018118089639378016,
+      "loss": 0.146,
+      "step": 4197
+    },
+    {
+      "epoch": 0.19891021085050936,
+      "grad_norm": 0.0037994384765625,
+      "learning_rate": 0.00018117219936379742,
+      "loss": 0.0002,
+      "step": 4198
+    },
+    {
+      "epoch": 0.19895759298744373,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018116350053350172,
+      "loss": 1.1075,
+      "step": 4199
+    },
+    {
+      "epoch": 0.19900497512437812,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001811547999030861,
+      "loss": 0.7409,
+      "step": 4200
+    },
+    {
+      "epoch": 0.19905235726131248,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001811460974727434,
+      "loss": 0.0631,
+      "step": 4201
+    },
+    {
+      "epoch": 0.19909973939824685,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018113739324266673,
+      "loss": 0.1505,
+      "step": 4202
+    },
+    {
+      "epoch": 0.19914712153518124,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018112868721304904,
+      "loss": 1.0933,
+      "step": 4203
+    },
+    {
+      "epoch": 0.1991945036721156,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018111997938408353,
+      "loss": 0.8051,
+      "step": 4204
+    },
+    {
+      "epoch": 0.19924188580905,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00018111126975596328,
+      "loss": 0.0196,
+      "step": 4205
+    },
+    {
+      "epoch": 0.19928926794598437,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00018110255832888146,
+      "loss": 0.1584,
+      "step": 4206
+    },
+    {
+      "epoch": 0.19933665008291873,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018109384510303127,
+      "loss": 0.1847,
+      "step": 4207
+    },
+    {
+      "epoch": 0.19938403221985312,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00018108513007860597,
+      "loss": 0.6955,
+      "step": 4208
+    },
+    {
+      "epoch": 0.1994314143567875,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018107641325579887,
+      "loss": 0.0291,
+      "step": 4209
+    },
+    {
+      "epoch": 0.19947879649372185,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001810676946348033,
+      "loss": 1.3563,
+      "step": 4210
+    },
+    {
+      "epoch": 0.19952617863065625,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018105897421581257,
+      "loss": 0.1529,
+      "step": 4211
+    },
+    {
+      "epoch": 0.1995735607675906,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018105025199902016,
+      "loss": 0.1252,
+      "step": 4212
+    },
+    {
+      "epoch": 0.199620942904525,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018104152798461946,
+      "loss": 0.1457,
+      "step": 4213
+    },
+    {
+      "epoch": 0.19966832504145937,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00018103280217280406,
+      "loss": 1.43,
+      "step": 4214
+    },
+    {
+      "epoch": 0.19971570717839374,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00018102407456376733,
+      "loss": 1.4493,
+      "step": 4215
+    },
+    {
+      "epoch": 0.19976308931532813,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018101534515770298,
+      "loss": 1.3808,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1998104714522625,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018100661395480455,
+      "loss": 0.8198,
+      "step": 4217
+    },
+    {
+      "epoch": 0.19985785358919686,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00018099788095526576,
+      "loss": 0.1655,
+      "step": 4218
+    },
+    {
+      "epoch": 0.19990523572613125,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018098914615928018,
+      "loss": 0.8965,
+      "step": 4219
+    },
+    {
+      "epoch": 0.19995261786306562,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00018098040956704162,
+      "loss": 0.0433,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018097167117874386,
+      "loss": 1.3795,
+      "step": 4221
+    },
+    {
+      "epoch": 0.20004738213693438,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018096293099458067,
+      "loss": 1.1495,
+      "step": 4222
+    },
+    {
+      "epoch": 0.20009476427386874,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018095418901474587,
+      "loss": 0.0622,
+      "step": 4223
+    },
+    {
+      "epoch": 0.20014214641080313,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001809454452394334,
+      "loss": 0.1261,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2001895285477375,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00018093669966883722,
+      "loss": 0.1317,
+      "step": 4225
+    },
+    {
+      "epoch": 0.20023691068467186,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001809279523031512,
+      "loss": 1.4214,
+      "step": 4226
+    },
+    {
+      "epoch": 0.20028429282160626,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018091920314256945,
+      "loss": 0.301,
+      "step": 4227
+    },
+    {
+      "epoch": 0.20033167495854062,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018091045218728593,
+      "loss": 0.1512,
+      "step": 4228
+    },
+    {
+      "epoch": 0.20037905709547502,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.1496,
+      "step": 4229
+    },
+    {
+      "epoch": 0.20042643923240938,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018089294489339008,
+      "loss": 0.88,
+      "step": 4230
+    },
+    {
+      "epoch": 0.20047382136934375,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018088418855516603,
+      "loss": 0.201,
+      "step": 4231
+    },
+    {
+      "epoch": 0.20052120350627814,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00018087543042301685,
+      "loss": 0.675,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2005685856432125,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018086667049713675,
+      "loss": 0.0207,
+      "step": 4233
+    },
+    {
+      "epoch": 0.2006159677801469,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018085790877772005,
+      "loss": 1.1899,
+      "step": 4234
+    },
+    {
+      "epoch": 0.20066334991708126,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018084914526496105,
+      "loss": 1.0218,
+      "step": 4235
+    },
+    {
+      "epoch": 0.20071073205401563,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001808403799590541,
+      "loss": 0.0254,
+      "step": 4236
+    },
+    {
+      "epoch": 0.20075811419095002,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018083161286019367,
+      "loss": 0.8982,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2008054963278844,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001808228439685742,
+      "loss": 1.0802,
+      "step": 4238
+    },
+    {
+      "epoch": 0.20085287846481875,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00018081407328439007,
+      "loss": 0.926,
+      "step": 4239
+    },
+    {
+      "epoch": 0.20090026060175314,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001808053008078359,
+      "loss": 0.8248,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2009476427386875,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001807965265391062,
+      "loss": 0.9299,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2009950248756219,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018078775047839564,
+      "loss": 0.2623,
+      "step": 4242
+    },
+    {
+      "epoch": 0.20104240701255627,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018077897262589883,
+      "loss": 0.1651,
+      "step": 4243
+    },
+    {
+      "epoch": 0.20108978914949063,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018077019298181043,
+      "loss": 0.0975,
+      "step": 4244
+    },
+    {
+      "epoch": 0.20113717128642503,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00018076141154632516,
+      "loss": 0.8288,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2011845534233594,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001807526283196378,
+      "loss": 0.7461,
+      "step": 4246
+    },
+    {
+      "epoch": 0.20123193556029376,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00018074384330194322,
+      "loss": 0.0066,
+      "step": 4247
+    },
+    {
+      "epoch": 0.20127931769722815,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018073505649343616,
+      "loss": 0.1853,
+      "step": 4248
+    },
+    {
+      "epoch": 0.20132669983416251,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018072626789431156,
+      "loss": 0.8787,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2013740819710969,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001807174775047643,
+      "loss": 0.4628,
+      "step": 4250
+    },
+    {
+      "epoch": 0.20142146410803127,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001807086853249894,
+      "loss": 0.3134,
+      "step": 4251
+    },
+    {
+      "epoch": 0.20146884624496564,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001806998913551818,
+      "loss": 0.8921,
+      "step": 4252
+    },
+    {
+      "epoch": 0.20151622838190003,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018069109559553655,
+      "loss": 0.7831,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2015636105188344,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018068229804624878,
+      "loss": 0.0292,
+      "step": 4254
+    },
+    {
+      "epoch": 0.20161099265576876,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018067349870751355,
+      "loss": 0.0452,
+      "step": 4255
+    },
+    {
+      "epoch": 0.20165837479270315,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018066469757952608,
+      "loss": 1.1553,
+      "step": 4256
+    },
+    {
+      "epoch": 0.20170575692963752,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018065589466248152,
+      "loss": 0.9697,
+      "step": 4257
+    },
+    {
+      "epoch": 0.2017531390665719,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018064708995657513,
+      "loss": 0.9858,
+      "step": 4258
+    },
+    {
+      "epoch": 0.20180052120350628,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001806382834620022,
+      "loss": 0.994,
+      "step": 4259
+    },
+    {
+      "epoch": 0.20184790334044064,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.000180629475178958,
+      "loss": 0.1986,
+      "step": 4260
+    },
+    {
+      "epoch": 0.20189528547737504,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00018062066510763795,
+      "loss": 0.2004,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2019426676143094,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018061185324823743,
+      "loss": 0.0201,
+      "step": 4262
+    },
+    {
+      "epoch": 0.2019900497512438,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018060303960095185,
+      "loss": 1.223,
+      "step": 4263
+    },
+    {
+      "epoch": 0.20203743188817816,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001805942241659767,
+      "loss": 0.8158,
+      "step": 4264
+    },
+    {
+      "epoch": 0.20208481402511252,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00018058540694350752,
+      "loss": 0.0056,
+      "step": 4265
+    },
+    {
+      "epoch": 0.20213219616204692,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018057658793373983,
+      "loss": 1.104,
+      "step": 4266
+    },
+    {
+      "epoch": 0.20217957829898128,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00018056776713686927,
+      "loss": 1.2196,
+      "step": 4267
+    },
+    {
+      "epoch": 0.20222696043591565,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018055894455309144,
+      "loss": 1.0654,
+      "step": 4268
+    },
+    {
+      "epoch": 0.20227434257285004,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018055012018260204,
+      "loss": 1.1841,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2023217247097844,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018054129402559675,
+      "loss": 1.1881,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2023691068467188,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018053246608227137,
+      "loss": 0.5166,
+      "step": 4271
+    },
+    {
+      "epoch": 0.20241648898365316,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018052363635282163,
+      "loss": 0.0744,
+      "step": 4272
+    },
+    {
+      "epoch": 0.20246387112058753,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018051480483744346,
+      "loss": 0.9402,
+      "step": 4273
+    },
+    {
+      "epoch": 0.20251125325752192,
+      "grad_norm": 0.02197265625,
+      "learning_rate": 0.00018050597153633263,
+      "loss": 0.0009,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2025586353944563,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00018049713644968516,
+      "loss": 0.1977,
+      "step": 4275
+    },
+    {
+      "epoch": 0.20260601753139065,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001804882995776969,
+      "loss": 1.4933,
+      "step": 4276
+    },
+    {
+      "epoch": 0.20265339966832505,
+      "grad_norm": 0.01129150390625,
+      "learning_rate": 0.00018047946092056391,
+      "loss": 0.0006,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2027007818052594,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001804706204784822,
+      "loss": 1.4036,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2027481639421938,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018046177825164784,
+      "loss": 0.015,
+      "step": 4279
+    },
+    {
+      "epoch": 0.20279554607912817,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001804529342402569,
+      "loss": 0.6647,
+      "step": 4280
+    },
+    {
+      "epoch": 0.20284292821606253,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018044408844450563,
+      "loss": 0.3855,
+      "step": 4281
+    },
+    {
+      "epoch": 0.20289031035299693,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018043524086459013,
+      "loss": 0.9161,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2029376924899313,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001804263915007067,
+      "loss": 1.0379,
+      "step": 4283
+    },
+    {
+      "epoch": 0.20298507462686566,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018041754035305152,
+      "loss": 0.0514,
+      "step": 4284
+    },
+    {
+      "epoch": 0.20303245676380005,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00018040868742182098,
+      "loss": 0.0456,
+      "step": 4285
+    },
+    {
+      "epoch": 0.20307983890073442,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001803998327072114,
+      "loss": 1.2112,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2031272210376688,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018039097620941915,
+      "loss": 0.0263,
+      "step": 4287
+    },
+    {
+      "epoch": 0.20317460317460317,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001803821179286407,
+      "loss": 0.0274,
+      "step": 4288
+    },
+    {
+      "epoch": 0.20322198531153754,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018037325786507248,
+      "loss": 0.0196,
+      "step": 4289
+    },
+    {
+      "epoch": 0.20326936744847193,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018036439601891102,
+      "loss": 1.1825,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2033167495854063,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018035553239035285,
+      "loss": 0.8397,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2033641317223407,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018034666697959456,
+      "loss": 0.9871,
+      "step": 4292
+    },
+    {
+      "epoch": 0.20341151385927506,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001803377997868328,
+      "loss": 0.7991,
+      "step": 4293
+    },
+    {
+      "epoch": 0.20345889599620942,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001803289308122642,
+      "loss": 0.9436,
+      "step": 4294
+    },
+    {
+      "epoch": 0.20350627813314381,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001803200600560855,
+      "loss": 1.4093,
+      "step": 4295
+    },
+    {
+      "epoch": 0.20355366027007818,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018031118751849338,
+      "loss": 0.0173,
+      "step": 4296
+    },
+    {
+      "epoch": 0.20360104240701254,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00018030231319968466,
+      "loss": 0.1785,
+      "step": 4297
+    },
+    {
+      "epoch": 0.20364842454394694,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001802934370998562,
+      "loss": 0.6897,
+      "step": 4298
+    },
+    {
+      "epoch": 0.2036958066808813,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001802845592192048,
+      "loss": 0.0282,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2037431888178157,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001802756795579274,
+      "loss": 1.045,
+      "step": 4300
+    },
+    {
+      "epoch": 0.20379057095475006,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018026679811622098,
+      "loss": 1.0125,
+      "step": 4301
+    },
+    {
+      "epoch": 0.20383795309168443,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001802579148942824,
+      "loss": 1.3856,
+      "step": 4302
+    },
+    {
+      "epoch": 0.20388533522861882,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018024902989230882,
+      "loss": 0.8329,
+      "step": 4303
+    },
+    {
+      "epoch": 0.20393271736555318,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001802401431104972,
+      "loss": 1.3463,
+      "step": 4304
+    },
+    {
+      "epoch": 0.20398009950248755,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00018023125454904467,
+      "loss": 0.2841,
+      "step": 4305
+    },
+    {
+      "epoch": 0.20402748163942194,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018022236420814838,
+      "loss": 0.7527,
+      "step": 4306
+    },
+    {
+      "epoch": 0.2040748637763563,
+      "grad_norm": 0.00122833251953125,
+      "learning_rate": 0.0001802134720880055,
+      "loss": 0.0001,
+      "step": 4307
+    },
+    {
+      "epoch": 0.2041222459132907,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018020457818881326,
+      "loss": 0.7237,
+      "step": 4308
+    },
+    {
+      "epoch": 0.20416962805022507,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001801956825107689,
+      "loss": 1.564,
+      "step": 4309
+    },
+    {
+      "epoch": 0.20421701018715943,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00018018678505406972,
+      "loss": 1.4349,
+      "step": 4310
+    },
+    {
+      "epoch": 0.20426439232409382,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018017788581891307,
+      "loss": 1.4531,
+      "step": 4311
+    },
+    {
+      "epoch": 0.2043117744610282,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001801689848054963,
+      "loss": 0.8812,
+      "step": 4312
+    },
+    {
+      "epoch": 0.20435915659796255,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018016008201401684,
+      "loss": 0.8732,
+      "step": 4313
+    },
+    {
+      "epoch": 0.20440653873489695,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00018015117744467213,
+      "loss": 0.1634,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2044539208718313,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0001801422710976597,
+      "loss": 0.0031,
+      "step": 4315
+    },
+    {
+      "epoch": 0.2045013030087657,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00018013336297317703,
+      "loss": 0.8116,
+      "step": 4316
+    },
+    {
+      "epoch": 0.20454868514570007,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018012445307142175,
+      "loss": 1.1377,
+      "step": 4317
+    },
+    {
+      "epoch": 0.20459606728263444,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018011554139259146,
+      "loss": 0.8103,
+      "step": 4318
+    },
+    {
+      "epoch": 0.20464344941956883,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018010662793688378,
+      "loss": 0.8936,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2046908315565032,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001800977127044964,
+      "loss": 0.1203,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2047382136934376,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001800887956956271,
+      "loss": 0.9999,
+      "step": 4321
+    },
+    {
+      "epoch": 0.20478559583037195,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00018007987691047358,
+      "loss": 1.2671,
+      "step": 4322
+    },
+    {
+      "epoch": 0.20483297796730632,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018007095634923372,
+      "loss": 1.2057,
+      "step": 4323
+    },
+    {
+      "epoch": 0.2048803601042407,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00018006203401210535,
+      "loss": 1.0501,
+      "step": 4324
+    },
+    {
+      "epoch": 0.20492774224117508,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018005310989928632,
+      "loss": 1.1723,
+      "step": 4325
+    },
+    {
+      "epoch": 0.20497512437810944,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018004418401097456,
+      "loss": 1.1301,
+      "step": 4326
+    },
+    {
+      "epoch": 0.20502250651504383,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018003525634736808,
+      "loss": 0.8912,
+      "step": 4327
+    },
+    {
+      "epoch": 0.2050698886519782,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018002632690866487,
+      "loss": 1.0652,
+      "step": 4328
+    },
+    {
+      "epoch": 0.2051172707889126,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.000180017395695063,
+      "loss": 1.0792,
+      "step": 4329
+    },
+    {
+      "epoch": 0.20516465292584696,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00018000846270676047,
+      "loss": 1.1601,
+      "step": 4330
+    },
+    {
+      "epoch": 0.20521203506278132,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017999952794395548,
+      "loss": 0.0512,
+      "step": 4331
+    },
+    {
+      "epoch": 0.20525941719971572,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017999059140684615,
+      "loss": 0.7923,
+      "step": 4332
+    },
+    {
+      "epoch": 0.20530679933665008,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00017998165309563073,
+      "loss": 0.7649,
+      "step": 4333
+    },
+    {
+      "epoch": 0.20535418147358445,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001799727130105074,
+      "loss": 0.9117,
+      "step": 4334
+    },
+    {
+      "epoch": 0.20540156361051884,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017996377115167452,
+      "loss": 0.3355,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2054489457474532,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017995482751933037,
+      "loss": 1.2005,
+      "step": 4336
+    },
+    {
+      "epoch": 0.2054963278843876,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001799458821136733,
+      "loss": 0.8636,
+      "step": 4337
+    },
+    {
+      "epoch": 0.20554371002132196,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00017993693493490175,
+      "loss": 0.5847,
+      "step": 4338
+    },
+    {
+      "epoch": 0.20559109215825633,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017992798598321405,
+      "loss": 1.2174,
+      "step": 4339
+    },
+    {
+      "epoch": 0.20563847429519072,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017991903525880882,
+      "loss": 1.2868,
+      "step": 4340
+    },
+    {
+      "epoch": 0.20568585643212509,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017991008276188448,
+      "loss": 0.0422,
+      "step": 4341
+    },
+    {
+      "epoch": 0.20573323856905945,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017990112849263965,
+      "loss": 0.8283,
+      "step": 4342
+    },
+    {
+      "epoch": 0.20578062070599384,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017989217245127287,
+      "loss": 1.3559,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2058280028429282,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017988321463798283,
+      "loss": 0.6048,
+      "step": 4344
+    },
+    {
+      "epoch": 0.2058753849798626,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017987425505296815,
+      "loss": 0.9089,
+      "step": 4345
+    },
+    {
+      "epoch": 0.20592276711679697,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017986529369642758,
+      "loss": 0.8273,
+      "step": 4346
+    },
+    {
+      "epoch": 0.20597014925373133,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017985633056855988,
+      "loss": 0.1078,
+      "step": 4347
+    },
+    {
+      "epoch": 0.20601753139066573,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017984736566956382,
+      "loss": 0.8183,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2060649135276001,
+      "grad_norm": 0.00225830078125,
+      "learning_rate": 0.00017983839899963822,
+      "loss": 0.0002,
+      "step": 4349
+    },
+    {
+      "epoch": 0.20611229566453448,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000179829430558982,
+      "loss": 0.1592,
+      "step": 4350
+    },
+    {
+      "epoch": 0.20615967780146885,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000179820460347794,
+      "loss": 0.9371,
+      "step": 4351
+    },
+    {
+      "epoch": 0.20620705993840321,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017981148836627325,
+      "loss": 1.0387,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2062544420753376,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00017980251461461867,
+      "loss": 0.4302,
+      "step": 4353
+    },
+    {
+      "epoch": 0.20630182421227197,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017979353909302934,
+      "loss": 0.5641,
+      "step": 4354
+    },
+    {
+      "epoch": 0.20634920634920634,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001797845618017043,
+      "loss": 0.986,
+      "step": 4355
+    },
+    {
+      "epoch": 0.20639658848614073,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017977558274084266,
+      "loss": 1.3083,
+      "step": 4356
+    },
+    {
+      "epoch": 0.2064439706230751,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017976660191064357,
+      "loss": 0.7942,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2064913527600095,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001797576193113062,
+      "loss": 0.26,
+      "step": 4358
+    },
+    {
+      "epoch": 0.20653873489694385,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017974863494302982,
+      "loss": 0.9274,
+      "step": 4359
+    },
+    {
+      "epoch": 0.20658611703387822,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017973964880601363,
+      "loss": 0.7014,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2066334991708126,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000179730660900457,
+      "loss": 0.059,
+      "step": 4361
+    },
+    {
+      "epoch": 0.20668088130774698,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017972167122655918,
+      "loss": 0.1828,
+      "step": 4362
+    },
+    {
+      "epoch": 0.20672826344468134,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00017971267978451967,
+      "loss": 0.6896,
+      "step": 4363
+    },
+    {
+      "epoch": 0.20677564558161574,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017970368657453778,
+      "loss": 0.767,
+      "step": 4364
+    },
+    {
+      "epoch": 0.2068230277185501,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000179694691596813,
+      "loss": 0.1646,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2068704098554845,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017968569485154489,
+      "loss": 1.4281,
+      "step": 4366
+    },
+    {
+      "epoch": 0.20691779199241886,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017967669633893297,
+      "loss": 0.9969,
+      "step": 4367
+    },
+    {
+      "epoch": 0.20696517412935322,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017966769605917674,
+      "loss": 0.8421,
+      "step": 4368
+    },
+    {
+      "epoch": 0.20701255626628762,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001796586940124759,
+      "loss": 0.9112,
+      "step": 4369
+    },
+    {
+      "epoch": 0.20705993840322198,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017964969019903005,
+      "loss": 0.5215,
+      "step": 4370
+    },
+    {
+      "epoch": 0.20710732054015635,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017964068461903896,
+      "loss": 0.8646,
+      "step": 4371
+    },
+    {
+      "epoch": 0.20715470267709074,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00017963167727270225,
+      "loss": 0.1278,
+      "step": 4372
+    },
+    {
+      "epoch": 0.2072020848140251,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001796226681602198,
+      "loss": 1.0172,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2072494669509595,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017961365728179138,
+      "loss": 0.7388,
+      "step": 4374
+    },
+    {
+      "epoch": 0.20729684908789386,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017960464463761682,
+      "loss": 0.9097,
+      "step": 4375
+    },
+    {
+      "epoch": 0.20734423122482823,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00017959563022789605,
+      "loss": 0.4247,
+      "step": 4376
+    },
+    {
+      "epoch": 0.20739161336176262,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017958661405282898,
+      "loss": 0.4378,
+      "step": 4377
+    },
+    {
+      "epoch": 0.207438995498697,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001795775961126156,
+      "loss": 0.8121,
+      "step": 4378
+    },
+    {
+      "epoch": 0.20748637763563138,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017956857640745585,
+      "loss": 0.1325,
+      "step": 4379
+    },
+    {
+      "epoch": 0.20753375977256575,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017955955493754986,
+      "loss": 0.1541,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2075811419095001,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001795505317030977,
+      "loss": 1.0451,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2076285240464345,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017954150670429944,
+      "loss": 0.6229,
+      "step": 4382
+    },
+    {
+      "epoch": 0.20767590618336887,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001795324799413553,
+      "loss": 1.2359,
+      "step": 4383
+    },
+    {
+      "epoch": 0.20772328832030323,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017952345141446552,
+      "loss": 0.1026,
+      "step": 4384
+    },
+    {
+      "epoch": 0.20777067045723763,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001795144211238302,
+      "loss": 0.7769,
+      "step": 4385
+    },
+    {
+      "epoch": 0.207818052594172,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017950538906964979,
+      "loss": 0.53,
+      "step": 4386
+    },
+    {
+      "epoch": 0.20786543473110639,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001794963552521245,
+      "loss": 1.4608,
+      "step": 4387
+    },
+    {
+      "epoch": 0.20791281686804075,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017948731967145474,
+      "loss": 1.4962,
+      "step": 4388
+    },
+    {
+      "epoch": 0.20796019900497512,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001794782823278409,
+      "loss": 1.2669,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2080075811419095,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017946924322148343,
+      "loss": 1.1537,
+      "step": 4390
+    },
+    {
+      "epoch": 0.20805496327884387,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017946020235258274,
+      "loss": 0.9543,
+      "step": 4391
+    },
+    {
+      "epoch": 0.20810234541577824,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017945115972133944,
+      "loss": 0.7522,
+      "step": 4392
+    },
+    {
+      "epoch": 0.20814972755271263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000179442115327954,
+      "loss": 0.1398,
+      "step": 4393
+    },
+    {
+      "epoch": 0.208197109689647,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001794330691726271,
+      "loss": 0.2923,
+      "step": 4394
+    },
+    {
+      "epoch": 0.2082444918265814,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017942402125555935,
+      "loss": 0.1591,
+      "step": 4395
+    },
+    {
+      "epoch": 0.20829187396351576,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00017941497157695138,
+      "loss": 0.0288,
+      "step": 4396
+    },
+    {
+      "epoch": 0.20833925610045012,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017940592013700394,
+      "loss": 1.0994,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2083866382373845,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017939686693591776,
+      "loss": 1.5042,
+      "step": 4398
+    },
+    {
+      "epoch": 0.20843402037431888,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017938781197389365,
+      "loss": 0.4372,
+      "step": 4399
+    },
+    {
+      "epoch": 0.20848140251125324,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017937875525113242,
+      "loss": 1.1026,
+      "step": 4400
+    },
+    {
+      "epoch": 0.20852878464818764,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017936969676783494,
+      "loss": 0.1716,
+      "step": 4401
+    },
+    {
+      "epoch": 0.208576166785122,
+      "grad_norm": 0.0037689208984375,
+      "learning_rate": 0.00017936063652420215,
+      "loss": 0.0003,
+      "step": 4402
+    },
+    {
+      "epoch": 0.2086235489220564,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00017935157452043495,
+      "loss": 0.0158,
+      "step": 4403
+    },
+    {
+      "epoch": 0.20867093105899076,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00017934251075673436,
+      "loss": 0.0206,
+      "step": 4404
+    },
+    {
+      "epoch": 0.20871831319592513,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001793334452333014,
+      "loss": 0.8304,
+      "step": 4405
+    },
+    {
+      "epoch": 0.20876569533285952,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017932437795033712,
+      "loss": 0.0437,
+      "step": 4406
+    },
+    {
+      "epoch": 0.20881307746979388,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017931530890804263,
+      "loss": 1.0193,
+      "step": 4407
+    },
+    {
+      "epoch": 0.20886045960672828,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001793062381066191,
+      "loss": 0.4923,
+      "step": 4408
+    },
+    {
+      "epoch": 0.20890784174366264,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00017929716554626764,
+      "loss": 0.4406,
+      "step": 4409
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017928809122718953,
+      "loss": 0.8709,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2090026060175314,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000179279015149586,
+      "loss": 0.6877,
+      "step": 4411
+    },
+    {
+      "epoch": 0.20904998815446577,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00017926993731365838,
+      "loss": 0.0178,
+      "step": 4412
+    },
+    {
+      "epoch": 0.20909737029140013,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017926085771960797,
+      "loss": 0.9612,
+      "step": 4413
+    },
+    {
+      "epoch": 0.20914475242833452,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017925177636763615,
+      "loss": 0.3026,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2091921345652689,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017924269325794435,
+      "loss": 0.6007,
+      "step": 4415
+    },
+    {
+      "epoch": 0.20923951670220328,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017923360839073406,
+      "loss": 0.7906,
+      "step": 4416
+    },
+    {
+      "epoch": 0.20928689883913765,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017922452176620667,
+      "loss": 0.7725,
+      "step": 4417
+    },
+    {
+      "epoch": 0.209334280976072,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001792154333845638,
+      "loss": 1.5173,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2093816631130064,
+      "grad_norm": 0.013671875,
+      "learning_rate": 0.000179206343246007,
+      "loss": 0.0007,
+      "step": 4419
+    },
+    {
+      "epoch": 0.20942904524994077,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017919725135073788,
+      "loss": 0.9344,
+      "step": 4420
+    },
+    {
+      "epoch": 0.20947642738687514,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017918815769895808,
+      "loss": 0.0335,
+      "step": 4421
+    },
+    {
+      "epoch": 0.20952380952380953,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017917906229086925,
+      "loss": 0.0333,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2095711916607439,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001791699651266732,
+      "loss": 1.1535,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2096185737976783,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017916086620657165,
+      "loss": 1.2566,
+      "step": 4424
+    },
+    {
+      "epoch": 0.20966595593461265,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001791517655307664,
+      "loss": 0.9216,
+      "step": 4425
+    },
+    {
+      "epoch": 0.20971333807154702,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00017914266309945927,
+      "loss": 0.033,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2097607202084814,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001791335589128522,
+      "loss": 0.6292,
+      "step": 4427
+    },
+    {
+      "epoch": 0.20980810234541578,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017912445297114707,
+      "loss": 0.8958,
+      "step": 4428
+    },
+    {
+      "epoch": 0.20985548448235014,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017911534527454586,
+      "loss": 0.7853,
+      "step": 4429
+    },
+    {
+      "epoch": 0.20990286661928453,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017910623582325058,
+      "loss": 1.0217,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2099502487562189,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001790971246174632,
+      "loss": 0.9569,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2099976308931533,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001790880116573859,
+      "loss": 1.152,
+      "step": 4432
+    },
+    {
+      "epoch": 0.21004501303008766,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001790788969432207,
+      "loss": 0.9903,
+      "step": 4433
+    },
+    {
+      "epoch": 0.21009239516702202,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017906978047516983,
+      "loss": 0.6095,
+      "step": 4434
+    },
+    {
+      "epoch": 0.21013977730395642,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017906066225343542,
+      "loss": 0.7025,
+      "step": 4435
+    },
+    {
+      "epoch": 0.21018715944089078,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00017905154227821978,
+      "loss": 0.0122,
+      "step": 4436
+    },
+    {
+      "epoch": 0.21023454157782517,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001790424205497251,
+      "loss": 1.0213,
+      "step": 4437
+    },
+    {
+      "epoch": 0.21028192371475954,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017903329706815373,
+      "loss": 1.4509,
+      "step": 4438
+    },
+    {
+      "epoch": 0.2103293058516939,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000179024171833708,
+      "loss": 0.7588,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2103766879886283,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017901504484659034,
+      "loss": 1.0901,
+      "step": 4440
+    },
+    {
+      "epoch": 0.21042407012556266,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017900591610700312,
+      "loss": 1.1255,
+      "step": 4441
+    },
+    {
+      "epoch": 0.21047145226249703,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017899678561514884,
+      "loss": 0.5801,
+      "step": 4442
+    },
+    {
+      "epoch": 0.21051883439943142,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017898765337123003,
+      "loss": 0.9075,
+      "step": 4443
+    },
+    {
+      "epoch": 0.21056621653636579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017897851937544916,
+      "loss": 0.7969,
+      "step": 4444
+    },
+    {
+      "epoch": 0.21061359867330018,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017896938362800885,
+      "loss": 0.7594,
+      "step": 4445
+    },
+    {
+      "epoch": 0.21066098081023454,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017896024612911178,
+      "loss": 0.959,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2107083629471689,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017895110687896052,
+      "loss": 1.2508,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2107557450841033,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001789419658777578,
+      "loss": 0.6613,
+      "step": 4448
+    },
+    {
+      "epoch": 0.21080312722103767,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017893282312570635,
+      "loss": 1.3908,
+      "step": 4449
+    },
+    {
+      "epoch": 0.21085050935797203,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.000178923678623009,
+      "loss": 1.1284,
+      "step": 4450
+    },
+    {
+      "epoch": 0.21089789149490643,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017891453236986847,
+      "loss": 0.7242,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2109452736318408,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001789053843664877,
+      "loss": 1.0444,
+      "step": 4452
+    },
+    {
+      "epoch": 0.21099265576877518,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017889623461306956,
+      "loss": 0.8223,
+      "step": 4453
+    },
+    {
+      "epoch": 0.21104003790570955,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00017888708310981696,
+      "loss": 0.1471,
+      "step": 4454
+    },
+    {
+      "epoch": 0.21108742004264391,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001788779298569329,
+      "loss": 1.1383,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2111348021795783,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017886877485462038,
+      "loss": 0.0846,
+      "step": 4456
+    },
+    {
+      "epoch": 0.21118218431651267,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017885961810308238,
+      "loss": 0.1903,
+      "step": 4457
+    },
+    {
+      "epoch": 0.21122956645344704,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001788504596025221,
+      "loss": 0.0582,
+      "step": 4458
+    },
+    {
+      "epoch": 0.21127694859038143,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017884129935314262,
+      "loss": 1.2312,
+      "step": 4459
+    },
+    {
+      "epoch": 0.2113243307273158,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017883213735514706,
+      "loss": 0.8433,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2113717128642502,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001788229736087387,
+      "loss": 0.9221,
+      "step": 4461
+    },
+    {
+      "epoch": 0.21141909500118455,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00017881380811412073,
+      "loss": 0.8076,
+      "step": 4462
+    },
+    {
+      "epoch": 0.21146647713811892,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017880464087149643,
+      "loss": 1.1539,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2115138592750533,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017879547188106914,
+      "loss": 0.7532,
+      "step": 4464
+    },
+    {
+      "epoch": 0.21156124141198768,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017878630114304223,
+      "loss": 0.9691,
+      "step": 4465
+    },
+    {
+      "epoch": 0.21160862354892207,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017877712865761906,
+      "loss": 0.5623,
+      "step": 4466
+    },
+    {
+      "epoch": 0.21165600568585644,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001787679544250031,
+      "loss": 1.3412,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2117033878227908,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0001787587784453978,
+      "loss": 0.0055,
+      "step": 4468
+    },
+    {
+      "epoch": 0.2117507699597252,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017874960071900668,
+      "loss": 1.0678,
+      "step": 4469
+    },
+    {
+      "epoch": 0.21179815209665956,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001787404212460333,
+      "loss": 1.2227,
+      "step": 4470
+    },
+    {
+      "epoch": 0.21184553423359392,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00017873124002668126,
+      "loss": 0.1423,
+      "step": 4471
+    },
+    {
+      "epoch": 0.21189291637052832,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017872205706115417,
+      "loss": 0.7837,
+      "step": 4472
+    },
+    {
+      "epoch": 0.21194029850746268,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001787128723496557,
+      "loss": 0.4995,
+      "step": 4473
+    },
+    {
+      "epoch": 0.21198768064439708,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017870368589238957,
+      "loss": 1.6933,
+      "step": 4474
+    },
+    {
+      "epoch": 0.21203506278133144,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017869449768955953,
+      "loss": 0.0053,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2120824449182658,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017868530774136932,
+      "loss": 1.1833,
+      "step": 4476
+    },
+    {
+      "epoch": 0.2121298270552002,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017867611604802282,
+      "loss": 0.7067,
+      "step": 4477
+    },
+    {
+      "epoch": 0.21217720919213456,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017866692260972388,
+      "loss": 0.8348,
+      "step": 4478
+    },
+    {
+      "epoch": 0.21222459132906893,
+      "grad_norm": 0.0303955078125,
+      "learning_rate": 0.00017865772742667636,
+      "loss": 0.0019,
+      "step": 4479
+    },
+    {
+      "epoch": 0.21227197346600332,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017864853049908422,
+      "loss": 0.9081,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2123193556029377,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017863933182715147,
+      "loss": 0.5684,
+      "step": 4481
+    },
+    {
+      "epoch": 0.21236673773987208,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001786301314110821,
+      "loss": 0.9752,
+      "step": 4482
+    },
+    {
+      "epoch": 0.21241411987680645,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017862092925108016,
+      "loss": 0.0964,
+      "step": 4483
+    },
+    {
+      "epoch": 0.2124615020137408,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017861172534734977,
+      "loss": 0.0359,
+      "step": 4484
+    },
+    {
+      "epoch": 0.2125088841506752,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017860251970009503,
+      "loss": 1.3729,
+      "step": 4485
+    },
+    {
+      "epoch": 0.21255626628760957,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017859331230952013,
+      "loss": 0.6791,
+      "step": 4486
+    },
+    {
+      "epoch": 0.21260364842454393,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017858410317582929,
+      "loss": 0.2489,
+      "step": 4487
+    },
+    {
+      "epoch": 0.21265103056147833,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001785748922992267,
+      "loss": 0.7125,
+      "step": 4488
+    },
+    {
+      "epoch": 0.2126984126984127,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017856567967991675,
+      "loss": 0.8741,
+      "step": 4489
+    },
+    {
+      "epoch": 0.21274579483534709,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017855646531810367,
+      "loss": 0.2095,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21279317697228145,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001785472492139919,
+      "loss": 0.9846,
+      "step": 4491
+    },
+    {
+      "epoch": 0.21284055910921582,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001785380313677858,
+      "loss": 0.8484,
+      "step": 4492
+    },
+    {
+      "epoch": 0.2128879412461502,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017852881177968978,
+      "loss": 0.8543,
+      "step": 4493
+    },
+    {
+      "epoch": 0.21293532338308457,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001785195904499084,
+      "loss": 1.2075,
+      "step": 4494
+    },
+    {
+      "epoch": 0.21298270552001897,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017851036737864614,
+      "loss": 0.9833,
+      "step": 4495
+    },
+    {
+      "epoch": 0.21303008765695333,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017850114256610754,
+      "loss": 1.2396,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2130774697938877,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001784919160124972,
+      "loss": 0.0141,
+      "step": 4497
+    },
+    {
+      "epoch": 0.2131248519308221,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017848268771801978,
+      "loss": 0.5485,
+      "step": 4498
+    },
+    {
+      "epoch": 0.21317223406775646,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017847345768287994,
+      "loss": 1.1611,
+      "step": 4499
+    },
+    {
+      "epoch": 0.21321961620469082,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017846422590728241,
+      "loss": 1.4193,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2132669983416252,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001784549923914319,
+      "loss": 1.0991,
+      "step": 4501
+    },
+    {
+      "epoch": 0.21331438047855958,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017844575713553325,
+      "loss": 1.0719,
+      "step": 4502
+    },
+    {
+      "epoch": 0.21336176261549397,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00017843652013979123,
+      "loss": 1.1205,
+      "step": 4503
+    },
+    {
+      "epoch": 0.21340914475242834,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017842728140441078,
+      "loss": 0.7693,
+      "step": 4504
+    },
+    {
+      "epoch": 0.2134565268893627,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017841804092959675,
+      "loss": 1.0125,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2135039090262971,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017840879871555408,
+      "loss": 1.2968,
+      "step": 4506
+    },
+    {
+      "epoch": 0.21355129116323146,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017839955476248783,
+      "loss": 1.0322,
+      "step": 4507
+    },
+    {
+      "epoch": 0.21359867330016583,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001783903090706029,
+      "loss": 0.039,
+      "step": 4508
+    },
+    {
+      "epoch": 0.21364605543710022,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017838106164010445,
+      "loss": 1.449,
+      "step": 4509
+    },
+    {
+      "epoch": 0.21369343757403458,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00017837181247119753,
+      "loss": 0.2595,
+      "step": 4510
+    },
+    {
+      "epoch": 0.21374081971096898,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017836256156408728,
+      "loss": 0.5309,
+      "step": 4511
+    },
+    {
+      "epoch": 0.21378820184790334,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001783533089189789,
+      "loss": 1.3286,
+      "step": 4512
+    },
+    {
+      "epoch": 0.2138355839848377,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001783440545360776,
+      "loss": 0.9724,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2138829661217721,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017833479841558857,
+      "loss": 0.7573,
+      "step": 4514
+    },
+    {
+      "epoch": 0.21393034825870647,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017832554055771714,
+      "loss": 0.1716,
+      "step": 4515
+    },
+    {
+      "epoch": 0.21397773039564083,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001783162809626687,
+      "loss": 0.066,
+      "step": 4516
+    },
+    {
+      "epoch": 0.21402511253257522,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017830701963064856,
+      "loss": 0.9212,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2140724946695096,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017829775656186213,
+      "loss": 1.08,
+      "step": 4518
+    },
+    {
+      "epoch": 0.21411987680644398,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017828849175651482,
+      "loss": 1.1641,
+      "step": 4519
+    },
+    {
+      "epoch": 0.21416725894337835,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001782792252148122,
+      "loss": 0.5539,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2142146410803127,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00017826995693695974,
+      "loss": 0.356,
+      "step": 4521
+    },
+    {
+      "epoch": 0.2142620232172471,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017826068692316298,
+      "loss": 1.1708,
+      "step": 4522
+    },
+    {
+      "epoch": 0.21430940535418147,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017825141517362756,
+      "loss": 0.0662,
+      "step": 4523
+    },
+    {
+      "epoch": 0.21435678749111584,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017824214168855908,
+      "loss": 0.6175,
+      "step": 4524
+    },
+    {
+      "epoch": 0.21440416962805023,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00017823286646816326,
+      "loss": 0.9617,
+      "step": 4525
+    },
+    {
+      "epoch": 0.2144515517649846,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001782235895126458,
+      "loss": 1.2364,
+      "step": 4526
+    },
+    {
+      "epoch": 0.214498933901919,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001782143108222124,
+      "loss": 1.1671,
+      "step": 4527
+    },
+    {
+      "epoch": 0.21454631603885335,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017820503039706894,
+      "loss": 0.1485,
+      "step": 4528
+    },
+    {
+      "epoch": 0.21459369817578772,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001781957482374212,
+      "loss": 1.2461,
+      "step": 4529
+    },
+    {
+      "epoch": 0.2146410803127221,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017818646434347504,
+      "loss": 0.8072,
+      "step": 4530
+    },
+    {
+      "epoch": 0.21468846244965648,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001781771787154364,
+      "loss": 1.2577,
+      "step": 4531
+    },
+    {
+      "epoch": 0.21473584458659087,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017816789135351117,
+      "loss": 1.0809,
+      "step": 4532
+    },
+    {
+      "epoch": 0.21478322672352523,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001781586022579054,
+      "loss": 0.0052,
+      "step": 4533
+    },
+    {
+      "epoch": 0.2148306088604596,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017814931142882506,
+      "loss": 0.7403,
+      "step": 4534
+    },
+    {
+      "epoch": 0.214877990997394,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017814001886647628,
+      "loss": 0.9979,
+      "step": 4535
+    },
+    {
+      "epoch": 0.21492537313432836,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017813072457106504,
+      "loss": 0.1139,
+      "step": 4536
+    },
+    {
+      "epoch": 0.21497275527126272,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017812142854279763,
+      "loss": 1.0816,
+      "step": 4537
+    },
+    {
+      "epoch": 0.21502013740819712,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017811213078188008,
+      "loss": 0.0649,
+      "step": 4538
+    },
+    {
+      "epoch": 0.21506751954513148,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017810283128851873,
+      "loss": 0.0291,
+      "step": 4539
+    },
+    {
+      "epoch": 0.21511490168206587,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00017809353006291974,
+      "loss": 0.0044,
+      "step": 4540
+    },
+    {
+      "epoch": 0.21516228381900024,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017808422710528942,
+      "loss": 1.1405,
+      "step": 4541
+    },
+    {
+      "epoch": 0.2152096659559346,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017807492241583413,
+      "loss": 0.5905,
+      "step": 4542
+    },
+    {
+      "epoch": 0.215257048092869,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017806561599476023,
+      "loss": 0.9847,
+      "step": 4543
+    },
+    {
+      "epoch": 0.21530443022980336,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001780563078422741,
+      "loss": 0.9864,
+      "step": 4544
+    },
+    {
+      "epoch": 0.21535181236673773,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017804699795858224,
+      "loss": 0.1668,
+      "step": 4545
+    },
+    {
+      "epoch": 0.21539919450367212,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00017803768634389107,
+      "loss": 0.5681,
+      "step": 4546
+    },
+    {
+      "epoch": 0.21544657664060649,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017802837299840714,
+      "loss": 0.9761,
+      "step": 4547
+    },
+    {
+      "epoch": 0.21549395877754088,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017801905792233703,
+      "loss": 1.0556,
+      "step": 4548
+    },
+    {
+      "epoch": 0.21554134091447524,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001780097411158873,
+      "loss": 1.2234,
+      "step": 4549
+    },
+    {
+      "epoch": 0.2155887230514096,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001780004225792646,
+      "loss": 1.0361,
+      "step": 4550
+    },
+    {
+      "epoch": 0.215636105188344,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001779911023126756,
+      "loss": 0.9324,
+      "step": 4551
+    },
+    {
+      "epoch": 0.21568348732527837,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017798178031632703,
+      "loss": 1.1086,
+      "step": 4552
+    },
+    {
+      "epoch": 0.21573086946221273,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00017797245659042566,
+      "loss": 0.0465,
+      "step": 4553
+    },
+    {
+      "epoch": 0.21577825159914713,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017796313113517824,
+      "loss": 0.846,
+      "step": 4554
+    },
+    {
+      "epoch": 0.2158256337360815,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001779538039507916,
+      "loss": 0.7183,
+      "step": 4555
+    },
+    {
+      "epoch": 0.21587301587301588,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017794447503747263,
+      "loss": 0.9674,
+      "step": 4556
+    },
+    {
+      "epoch": 0.21592039800995025,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017793514439542825,
+      "loss": 0.6377,
+      "step": 4557
+    },
+    {
+      "epoch": 0.2159677801468846,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017792581202486535,
+      "loss": 0.7499,
+      "step": 4558
+    },
+    {
+      "epoch": 0.216015162283819,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017791647792599093,
+      "loss": 0.9097,
+      "step": 4559
+    },
+    {
+      "epoch": 0.21606254442075337,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017790714209901207,
+      "loss": 1.3563,
+      "step": 4560
+    },
+    {
+      "epoch": 0.21610992655768776,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017789780454413573,
+      "loss": 1.0608,
+      "step": 4561
+    },
+    {
+      "epoch": 0.21615730869462213,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017788846526156911,
+      "loss": 0.089,
+      "step": 4562
+    },
+    {
+      "epoch": 0.2162046908315565,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017787912425151924,
+      "loss": 0.5966,
+      "step": 4563
+    },
+    {
+      "epoch": 0.2162520729684909,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017786978151419338,
+      "loss": 0.7353,
+      "step": 4564
+    },
+    {
+      "epoch": 0.21629945510542525,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001778604370497987,
+      "loss": 0.1329,
+      "step": 4565
+    },
+    {
+      "epoch": 0.21634683724235962,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017785109085854245,
+      "loss": 0.356,
+      "step": 4566
+    },
+    {
+      "epoch": 0.216394219379294,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00017784174294063195,
+      "loss": 0.0174,
+      "step": 4567
+    },
+    {
+      "epoch": 0.21644160151622838,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017783239329627454,
+      "loss": 1.3163,
+      "step": 4568
+    },
+    {
+      "epoch": 0.21648898365316277,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017782304192567747,
+      "loss": 0.1526,
+      "step": 4569
+    },
+    {
+      "epoch": 0.21653636579009714,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017781368882904824,
+      "loss": 0.0569,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2165837479270315,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017780433400659432,
+      "loss": 1.3177,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2166311300639659,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001777949774585231,
+      "loss": 0.0829,
+      "step": 4572
+    },
+    {
+      "epoch": 0.21667851220090026,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017778561918504218,
+      "loss": 0.0749,
+      "step": 4573
+    },
+    {
+      "epoch": 0.21672589433783462,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017777625918635907,
+      "loss": 1.1553,
+      "step": 4574
+    },
+    {
+      "epoch": 0.21677327647476902,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017776689746268137,
+      "loss": 1.1068,
+      "step": 4575
+    },
+    {
+      "epoch": 0.21682065861170338,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001777575340142167,
+      "loss": 0.6795,
+      "step": 4576
+    },
+    {
+      "epoch": 0.21686804074863777,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017774816884117277,
+      "loss": 1.1908,
+      "step": 4577
+    },
+    {
+      "epoch": 0.21691542288557214,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017773880194375726,
+      "loss": 1.3618,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2169628050225065,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017772943332217792,
+      "loss": 0.0409,
+      "step": 4579
+    },
+    {
+      "epoch": 0.2170101871594409,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00017772006297664256,
+      "loss": 0.6163,
+      "step": 4580
+    },
+    {
+      "epoch": 0.21705756929637526,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017771069090735897,
+      "loss": 1.048,
+      "step": 4581
+    },
+    {
+      "epoch": 0.21710495143330963,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017770131711453507,
+      "loss": 1.1447,
+      "step": 4582
+    },
+    {
+      "epoch": 0.21715233357024402,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017769194159837867,
+      "loss": 1.2723,
+      "step": 4583
+    },
+    {
+      "epoch": 0.2171997157071784,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017768256435909778,
+      "loss": 1.3821,
+      "step": 4584
+    },
+    {
+      "epoch": 0.21724709784411278,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00017767318539690035,
+      "loss": 0.0339,
+      "step": 4585
+    },
+    {
+      "epoch": 0.21729447998104715,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001776638047119944,
+      "loss": 0.6054,
+      "step": 4586
+    },
+    {
+      "epoch": 0.2173418621179815,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000177654422304588,
+      "loss": 1.1137,
+      "step": 4587
+    },
+    {
+      "epoch": 0.2173892442549159,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017764503817488923,
+      "loss": 1.4631,
+      "step": 4588
+    },
+    {
+      "epoch": 0.21743662639185027,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001776356523231062,
+      "loss": 0.9437,
+      "step": 4589
+    },
+    {
+      "epoch": 0.21748400852878466,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001776262647494471,
+      "loss": 0.6411,
+      "step": 4590
+    },
+    {
+      "epoch": 0.21753139066571903,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017761687545412013,
+      "loss": 0.7975,
+      "step": 4591
+    },
+    {
+      "epoch": 0.2175787728026534,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017760748443733352,
+      "loss": 0.1838,
+      "step": 4592
+    },
+    {
+      "epoch": 0.21762615493958778,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001775980916992956,
+      "loss": 1.1014,
+      "step": 4593
+    },
+    {
+      "epoch": 0.21767353707652215,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017758869724021463,
+      "loss": 0.7344,
+      "step": 4594
+    },
+    {
+      "epoch": 0.21772091921345652,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017757930106029903,
+      "loss": 0.9556,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2177683013503909,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017756990315975714,
+      "loss": 1.1179,
+      "step": 4596
+    },
+    {
+      "epoch": 0.21781568348732527,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00017756050353879746,
+      "loss": 0.6082,
+      "step": 4597
+    },
+    {
+      "epoch": 0.21786306562425967,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001775511021976284,
+      "loss": 0.6329,
+      "step": 4598
+    },
+    {
+      "epoch": 0.21791044776119403,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001775416991364585,
+      "loss": 1.2488,
+      "step": 4599
+    },
+    {
+      "epoch": 0.2179578298981284,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00017753229435549627,
+      "loss": 1.1134,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2180052120350628,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017752288785495037,
+      "loss": 0.9355,
+      "step": 4601
+    },
+    {
+      "epoch": 0.21805259417199715,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001775134796350294,
+      "loss": 0.064,
+      "step": 4602
+    },
+    {
+      "epoch": 0.21809997630893152,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.000177504069695942,
+      "loss": 0.3083,
+      "step": 4603
+    },
+    {
+      "epoch": 0.2181473584458659,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001774946580378969,
+      "loss": 0.1614,
+      "step": 4604
+    },
+    {
+      "epoch": 0.21819474058280028,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001774852446611028,
+      "loss": 0.2464,
+      "step": 4605
+    },
+    {
+      "epoch": 0.21824212271973467,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017747582956576853,
+      "loss": 0.5314,
+      "step": 4606
+    },
+    {
+      "epoch": 0.21828950485666904,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001774664127521029,
+      "loss": 0.0323,
+      "step": 4607
+    },
+    {
+      "epoch": 0.2183368869936034,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001774569942203147,
+      "loss": 0.5314,
+      "step": 4608
+    },
+    {
+      "epoch": 0.2183842691305378,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001774475739706129,
+      "loss": 0.8986,
+      "step": 4609
+    },
+    {
+      "epoch": 0.21843165126747216,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001774381520032064,
+      "loss": 1.1637,
+      "step": 4610
+    },
+    {
+      "epoch": 0.21847903340440653,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017742872831830422,
+      "loss": 0.8177,
+      "step": 4611
+    },
+    {
+      "epoch": 0.21852641554134092,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017741930291611526,
+      "loss": 1.1722,
+      "step": 4612
+    },
+    {
+      "epoch": 0.21857379767827528,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017740987579684863,
+      "loss": 0.8712,
+      "step": 4613
+    },
+    {
+      "epoch": 0.21862117981520968,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017740044696071342,
+      "loss": 0.062,
+      "step": 4614
+    },
+    {
+      "epoch": 0.21866856195214404,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017739101640791875,
+      "loss": 1.4296,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2187159440890784,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.00017738158413867377,
+      "loss": 0.0048,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2187633262260128,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017737215015318765,
+      "loss": 0.1394,
+      "step": 4617
+    },
+    {
+      "epoch": 0.21881070836294716,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017736271445166965,
+      "loss": 1.2406,
+      "step": 4618
+    },
+    {
+      "epoch": 0.21885809049988156,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017735327703432906,
+      "loss": 1.0548,
+      "step": 4619
+    },
+    {
+      "epoch": 0.21890547263681592,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001773438379013752,
+      "loss": 1.4781,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2189528547737503,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017733439705301738,
+      "loss": 1.6623,
+      "step": 4621
+    },
+    {
+      "epoch": 0.21900023691068468,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000177324954489465,
+      "loss": 1.0556,
+      "step": 4622
+    },
+    {
+      "epoch": 0.21904761904761905,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017731551021092748,
+      "loss": 0.0364,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2190950011845534,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00017730606421761434,
+      "loss": 0.5981,
+      "step": 4624
+    },
+    {
+      "epoch": 0.2191423833214878,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017729661650973502,
+      "loss": 1.2069,
+      "step": 4625
+    },
+    {
+      "epoch": 0.21918976545842217,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017728716708749907,
+      "loss": 0.628,
+      "step": 4626
+    },
+    {
+      "epoch": 0.21923714759535656,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001772777159511161,
+      "loss": 0.9068,
+      "step": 4627
+    },
+    {
+      "epoch": 0.21928452973229093,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017726826310079566,
+      "loss": 0.7564,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2193319118692253,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017725880853674744,
+      "loss": 0.256,
+      "step": 4629
+    },
+    {
+      "epoch": 0.2193792940061597,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001772493522591812,
+      "loss": 0.9717,
+      "step": 4630
+    },
+    {
+      "epoch": 0.21942667614309405,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017723989426830656,
+      "loss": 0.1078,
+      "step": 4631
+    },
+    {
+      "epoch": 0.21947405828002842,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017723043456433334,
+      "loss": 1.1237,
+      "step": 4632
+    },
+    {
+      "epoch": 0.2195214404169628,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017722097314747137,
+      "loss": 1.2465,
+      "step": 4633
+    },
+    {
+      "epoch": 0.21956882255389717,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017721151001793046,
+      "loss": 1.145,
+      "step": 4634
+    },
+    {
+      "epoch": 0.21961620469083157,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001772020451759205,
+      "loss": 0.5836,
+      "step": 4635
+    },
+    {
+      "epoch": 0.21966358682776593,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017719257862165139,
+      "loss": 0.0576,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2197109689647003,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017718311035533314,
+      "loss": 1.0329,
+      "step": 4637
+    },
+    {
+      "epoch": 0.2197583511016347,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017717364037717566,
+      "loss": 1.1943,
+      "step": 4638
+    },
+    {
+      "epoch": 0.21980573323856906,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001771641686873891,
+      "loss": 0.9518,
+      "step": 4639
+    },
+    {
+      "epoch": 0.21985311537550342,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017715469528618342,
+      "loss": 0.5623,
+      "step": 4640
+    },
+    {
+      "epoch": 0.21990049751243781,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001771452201737688,
+      "loss": 0.1291,
+      "step": 4641
+    },
+    {
+      "epoch": 0.21994787964937218,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00017713574335035542,
+      "loss": 0.7221,
+      "step": 4642
+    },
+    {
+      "epoch": 0.21999526178630657,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017712626481615335,
+      "loss": 0.8539,
+      "step": 4643
+    },
+    {
+      "epoch": 0.22004264392324094,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017711678457137288,
+      "loss": 1.3783,
+      "step": 4644
+    },
+    {
+      "epoch": 0.2200900260601753,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001771073026162243,
+      "loss": 0.0489,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2201374081971097,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017709781895091785,
+      "loss": 1.1835,
+      "step": 4646
+    },
+    {
+      "epoch": 0.22018479033404406,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001770883335756639,
+      "loss": 0.906,
+      "step": 4647
+    },
+    {
+      "epoch": 0.22023217247097845,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017707884649067283,
+      "loss": 0.7656,
+      "step": 4648
+    },
+    {
+      "epoch": 0.22027955460791282,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00017706935769615508,
+      "loss": 0.1213,
+      "step": 4649
+    },
+    {
+      "epoch": 0.22032693674484718,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017705986719232102,
+      "loss": 0.9539,
+      "step": 4650
+    },
+    {
+      "epoch": 0.22037431888178158,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001770503749793812,
+      "loss": 0.8432,
+      "step": 4651
+    },
+    {
+      "epoch": 0.22042170101871594,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017704088105754612,
+      "loss": 0.7951,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2204690831556503,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001770313854270264,
+      "loss": 0.9206,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2205164652925847,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017702188808803255,
+      "loss": 0.8624,
+      "step": 4654
+    },
+    {
+      "epoch": 0.22056384742951907,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017701238904077528,
+      "loss": 0.0222,
+      "step": 4655
+    },
+    {
+      "epoch": 0.22061122956645346,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017700288828546525,
+      "loss": 0.0418,
+      "step": 4656
+    },
+    {
+      "epoch": 0.22065861170338782,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00017699338582231316,
+      "loss": 0.0485,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2207059938403222,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017698388165152983,
+      "loss": 1.0035,
+      "step": 4658
+    },
+    {
+      "epoch": 0.22075337597725658,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000176974375773326,
+      "loss": 1.1605,
+      "step": 4659
+    },
+    {
+      "epoch": 0.22080075811419095,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017696486818791248,
+      "loss": 1.1143,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2208481402511253,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00017695535889550014,
+      "loss": 0.046,
+      "step": 4661
+    },
+    {
+      "epoch": 0.2208955223880597,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017694584789629996,
+      "loss": 0.9403,
+      "step": 4662
+    },
+    {
+      "epoch": 0.22094290452499407,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001769363351905228,
+      "loss": 1.1337,
+      "step": 4663
+    },
+    {
+      "epoch": 0.22099028666192846,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017692682077837966,
+      "loss": 1.2104,
+      "step": 4664
+    },
+    {
+      "epoch": 0.22103766879886283,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017691730466008163,
+      "loss": 0.6819,
+      "step": 4665
+    },
+    {
+      "epoch": 0.2210850509357972,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017690778683583967,
+      "loss": 0.8609,
+      "step": 4666
+    },
+    {
+      "epoch": 0.2211324330727316,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017689826730586493,
+      "loss": 1.376,
+      "step": 4667
+    },
+    {
+      "epoch": 0.22117981520966595,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017688874607036853,
+      "loss": 1.4283,
+      "step": 4668
+    },
+    {
+      "epoch": 0.22122719734660032,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00017687922312956163,
+      "loss": 0.2257,
+      "step": 4669
+    },
+    {
+      "epoch": 0.2212745794835347,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017686969848365545,
+      "loss": 0.0948,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22132196162046908,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017686017213286123,
+      "loss": 1.052,
+      "step": 4671
+    },
+    {
+      "epoch": 0.22136934375740347,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017685064407739026,
+      "loss": 1.2214,
+      "step": 4672
+    },
+    {
+      "epoch": 0.22141672589433783,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001768411143174539,
+      "loss": 1.276,
+      "step": 4673
+    },
+    {
+      "epoch": 0.2214641080312722,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017683158285326344,
+      "loss": 0.1367,
+      "step": 4674
+    },
+    {
+      "epoch": 0.2215114901682066,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017682204968503032,
+      "loss": 1.3712,
+      "step": 4675
+    },
+    {
+      "epoch": 0.22155887230514096,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017681251481296595,
+      "loss": 1.4865,
+      "step": 4676
+    },
+    {
+      "epoch": 0.22160625444207535,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00017680297823728183,
+      "loss": 0.5422,
+      "step": 4677
+    },
+    {
+      "epoch": 0.22165363657900972,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017679343995818947,
+      "loss": 0.7689,
+      "step": 4678
+    },
+    {
+      "epoch": 0.22170101871594408,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017678389997590038,
+      "loss": 0.1221,
+      "step": 4679
+    },
+    {
+      "epoch": 0.22174840085287847,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017677435829062618,
+      "loss": 0.6552,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22179578298981284,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001767648149025785,
+      "loss": 0.1533,
+      "step": 4681
+    },
+    {
+      "epoch": 0.2218431651267472,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.000176755269811969,
+      "loss": 1.3087,
+      "step": 4682
+    },
+    {
+      "epoch": 0.2218905472636816,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00017674572301900934,
+      "loss": 0.4479,
+      "step": 4683
+    },
+    {
+      "epoch": 0.22193792940061596,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017673617452391134,
+      "loss": 0.0224,
+      "step": 4684
+    },
+    {
+      "epoch": 0.22198531153755036,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001767266243268867,
+      "loss": 0.4738,
+      "step": 4685
+    },
+    {
+      "epoch": 0.22203269367448472,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00017671707242814723,
+      "loss": 0.8659,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2220800758114191,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017670751882790486,
+      "loss": 1.4518,
+      "step": 4687
+    },
+    {
+      "epoch": 0.22212745794835348,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017669796352637137,
+      "loss": 1.1001,
+      "step": 4688
+    },
+    {
+      "epoch": 0.22217484008528784,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001766884065237588,
+      "loss": 0.6323,
+      "step": 4689
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017667884782027903,
+      "loss": 1.1724,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2222696043591566,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001766692874161441,
+      "loss": 1.0944,
+      "step": 4691
+    },
+    {
+      "epoch": 0.22231698649609097,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017665972531156603,
+      "loss": 0.0107,
+      "step": 4692
+    },
+    {
+      "epoch": 0.22236436863302536,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001766501615067569,
+      "loss": 0.9774,
+      "step": 4693
+    },
+    {
+      "epoch": 0.22241175076995973,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017664059600192884,
+      "loss": 1.0095,
+      "step": 4694
+    },
+    {
+      "epoch": 0.2224591329068941,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017663102879729401,
+      "loss": 1.158,
+      "step": 4695
+    },
+    {
+      "epoch": 0.22250651504382848,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017662145989306456,
+      "loss": 0.918,
+      "step": 4696
+    },
+    {
+      "epoch": 0.22255389718076285,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017661188928945275,
+      "loss": 0.0291,
+      "step": 4697
+    },
+    {
+      "epoch": 0.22260127931769721,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017660231698667084,
+      "loss": 0.0249,
+      "step": 4698
+    },
+    {
+      "epoch": 0.2226486614546316,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00017659274298493114,
+      "loss": 0.0116,
+      "step": 4699
+    },
+    {
+      "epoch": 0.22269604359156597,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017658316728444597,
+      "loss": 0.6287,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22274342572850037,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017657358988542774,
+      "loss": 1.0842,
+      "step": 4701
+    },
+    {
+      "epoch": 0.22279080786543473,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017656401078808883,
+      "loss": 0.6734,
+      "step": 4702
+    },
+    {
+      "epoch": 0.2228381900023691,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017655442999264174,
+      "loss": 0.7644,
+      "step": 4703
+    },
+    {
+      "epoch": 0.2228855721393035,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017654484749929893,
+      "loss": 0.8459,
+      "step": 4704
+    },
+    {
+      "epoch": 0.22293295427623785,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001765352633082729,
+      "loss": 0.315,
+      "step": 4705
+    },
+    {
+      "epoch": 0.22298033641317225,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017652567741977627,
+      "loss": 0.5902,
+      "step": 4706
+    },
+    {
+      "epoch": 0.2230277185501066,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017651608983402164,
+      "loss": 1.2075,
+      "step": 4707
+    },
+    {
+      "epoch": 0.22307510068704098,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001765065005512216,
+      "loss": 0.7806,
+      "step": 4708
+    },
+    {
+      "epoch": 0.22312248282397537,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00017649690957158892,
+      "loss": 0.3137,
+      "step": 4709
+    },
+    {
+      "epoch": 0.22316986496090974,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017648731689533627,
+      "loss": 0.5101,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2232172470978441,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017647772252267637,
+      "loss": 0.6607,
+      "step": 4711
+    },
+    {
+      "epoch": 0.2232646292347785,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017646812645382201,
+      "loss": 1.1553,
+      "step": 4712
+    },
+    {
+      "epoch": 0.22331201137171286,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017645852868898611,
+      "loss": 1.0347,
+      "step": 4713
+    },
+    {
+      "epoch": 0.22335939350864725,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017644892922838147,
+      "loss": 0.8938,
+      "step": 4714
+    },
+    {
+      "epoch": 0.22340677564558162,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017643932807222102,
+      "loss": 1.466,
+      "step": 4715
+    },
+    {
+      "epoch": 0.22345415778251598,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017642972522071768,
+      "loss": 0.7273,
+      "step": 4716
+    },
+    {
+      "epoch": 0.22350153991945038,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017642012067408444,
+      "loss": 1.5163,
+      "step": 4717
+    },
+    {
+      "epoch": 0.22354892205638474,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017641051443253433,
+      "loss": 0.925,
+      "step": 4718
+    },
+    {
+      "epoch": 0.2235963041933191,
+      "grad_norm": 0.00823974609375,
+      "learning_rate": 0.00017640090649628042,
+      "loss": 0.0005,
+      "step": 4719
+    },
+    {
+      "epoch": 0.2236436863302535,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017639129686553573,
+      "loss": 0.9881,
+      "step": 4720
+    },
+    {
+      "epoch": 0.22369106846718786,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017638168554051352,
+      "loss": 0.5051,
+      "step": 4721
+    },
+    {
+      "epoch": 0.22373845060412226,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001763720725214268,
+      "loss": 0.9316,
+      "step": 4722
+    },
+    {
+      "epoch": 0.22378583274105662,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001763624578084889,
+      "loss": 1.0418,
+      "step": 4723
+    },
+    {
+      "epoch": 0.223833214877991,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017635284140191302,
+      "loss": 1.1204,
+      "step": 4724
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017634322330191244,
+      "loss": 0.8188,
+      "step": 4725
+    },
+    {
+      "epoch": 0.22392797915185975,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001763336035087005,
+      "loss": 1.2027,
+      "step": 4726
+    },
+    {
+      "epoch": 0.2239753612887941,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017632398202249055,
+      "loss": 1.5133,
+      "step": 4727
+    },
+    {
+      "epoch": 0.2240227434257285,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00017631435884349594,
+      "loss": 0.3672,
+      "step": 4728
+    },
+    {
+      "epoch": 0.22407012556266287,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017630473397193015,
+      "loss": 0.3307,
+      "step": 4729
+    },
+    {
+      "epoch": 0.22411750769959726,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017629510740800662,
+      "loss": 0.0769,
+      "step": 4730
+    },
+    {
+      "epoch": 0.22416488983653163,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017628547915193892,
+      "loss": 1.293,
+      "step": 4731
+    },
+    {
+      "epoch": 0.224212271973466,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001762758492039405,
+      "loss": 0.8374,
+      "step": 4732
+    },
+    {
+      "epoch": 0.22425965411040039,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017626621756422503,
+      "loss": 0.2328,
+      "step": 4733
+    },
+    {
+      "epoch": 0.22430703624733475,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017625658423300606,
+      "loss": 1.0732,
+      "step": 4734
+    },
+    {
+      "epoch": 0.22435441838426914,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017624694921049728,
+      "loss": 0.0418,
+      "step": 4735
+    },
+    {
+      "epoch": 0.2244018005212035,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001762373124969124,
+      "loss": 0.1589,
+      "step": 4736
+    },
+    {
+      "epoch": 0.22444918265813787,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001762276740924651,
+      "loss": 1.3373,
+      "step": 4737
+    },
+    {
+      "epoch": 0.22449656479507227,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017621803399736922,
+      "loss": 0.5409,
+      "step": 4738
+    },
+    {
+      "epoch": 0.22454394693200663,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017620839221183852,
+      "loss": 0.2755,
+      "step": 4739
+    },
+    {
+      "epoch": 0.224591329068941,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017619874873608685,
+      "loss": 0.9164,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2246387112058754,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017618910357032808,
+      "loss": 0.3311,
+      "step": 4741
+    },
+    {
+      "epoch": 0.22468609334280976,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00017617945671477618,
+      "loss": 1.7412,
+      "step": 4742
+    },
+    {
+      "epoch": 0.22473347547974415,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017616980816964503,
+      "loss": 0.0396,
+      "step": 4743
+    },
+    {
+      "epoch": 0.22478085761667851,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001761601579351487,
+      "loss": 0.7049,
+      "step": 4744
+    },
+    {
+      "epoch": 0.22482823975361288,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017615050601150117,
+      "loss": 0.0126,
+      "step": 4745
+    },
+    {
+      "epoch": 0.22487562189054727,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017614085239891654,
+      "loss": 1.1157,
+      "step": 4746
+    },
+    {
+      "epoch": 0.22492300402748164,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001761311970976089,
+      "loss": 0.8671,
+      "step": 4747
+    },
+    {
+      "epoch": 0.224970386164416,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017612154010779242,
+      "loss": 0.0453,
+      "step": 4748
+    },
+    {
+      "epoch": 0.2250177683013504,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017611188142968126,
+      "loss": 0.8946,
+      "step": 4749
+    },
+    {
+      "epoch": 0.22506515043828476,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001761022210634896,
+      "loss": 0.024,
+      "step": 4750
+    },
+    {
+      "epoch": 0.22511253257521915,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017609255900943177,
+      "loss": 1.0729,
+      "step": 4751
+    },
+    {
+      "epoch": 0.22515991471215352,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000176082895267722,
+      "loss": 0.9089,
+      "step": 4752
+    },
+    {
+      "epoch": 0.22520729684908788,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001760732298385747,
+      "loss": 1.1987,
+      "step": 4753
+    },
+    {
+      "epoch": 0.22525467898602228,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017606356272220416,
+      "loss": 0.9727,
+      "step": 4754
+    },
+    {
+      "epoch": 0.22530206112295664,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00017605389391882483,
+      "loss": 0.0066,
+      "step": 4755
+    },
+    {
+      "epoch": 0.225349443259891,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017604422342865113,
+      "loss": 0.1331,
+      "step": 4756
+    },
+    {
+      "epoch": 0.2253968253968254,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017603455125189758,
+      "loss": 0.9788,
+      "step": 4757
+    },
+    {
+      "epoch": 0.22544420753375977,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017602487738877862,
+      "loss": 1.3464,
+      "step": 4758
+    },
+    {
+      "epoch": 0.22549158967069416,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001760152018395089,
+      "loss": 1.2365,
+      "step": 4759
+    },
+    {
+      "epoch": 0.22553897180762852,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017600552460430297,
+      "loss": 0.9028,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2255863539445629,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017599584568337546,
+      "loss": 1.0353,
+      "step": 4761
+    },
+    {
+      "epoch": 0.22563373608149728,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017598616507694104,
+      "loss": 1.1336,
+      "step": 4762
+    },
+    {
+      "epoch": 0.22568111821843165,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001759764827852144,
+      "loss": 0.012,
+      "step": 4763
+    },
+    {
+      "epoch": 0.22572850035536604,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001759667988084103,
+      "loss": 0.8005,
+      "step": 4764
+    },
+    {
+      "epoch": 0.2257758824923004,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017595711314674352,
+      "loss": 0.0862,
+      "step": 4765
+    },
+    {
+      "epoch": 0.22582326462923477,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017594742580042888,
+      "loss": 1.197,
+      "step": 4766
+    },
+    {
+      "epoch": 0.22587064676616916,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017593773676968124,
+      "loss": 0.0649,
+      "step": 4767
+    },
+    {
+      "epoch": 0.22591802890310353,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00017592804605471546,
+      "loss": 0.6548,
+      "step": 4768
+    },
+    {
+      "epoch": 0.2259654110400379,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017591835365574654,
+      "loss": 0.013,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2260127931769723,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017590865957298934,
+      "loss": 1.1218,
+      "step": 4770
+    },
+    {
+      "epoch": 0.22606017531390665,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017589896380665896,
+      "loss": 0.7666,
+      "step": 4771
+    },
+    {
+      "epoch": 0.22610755745084105,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001758892663569704,
+      "loss": 0.8846,
+      "step": 4772
+    },
+    {
+      "epoch": 0.2261549395877754,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017587956722413877,
+      "loss": 0.6879,
+      "step": 4773
+    },
+    {
+      "epoch": 0.22620232172470978,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00017586986640837914,
+      "loss": 0.0823,
+      "step": 4774
+    },
+    {
+      "epoch": 0.22624970386164417,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017586016390990668,
+      "loss": 1.276,
+      "step": 4775
+    },
+    {
+      "epoch": 0.22629708599857853,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017585045972893658,
+      "loss": 0.008,
+      "step": 4776
+    },
+    {
+      "epoch": 0.2263444681355129,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001758407538656841,
+      "loss": 0.9002,
+      "step": 4777
+    },
+    {
+      "epoch": 0.2263918502724473,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017583104632036447,
+      "loss": 0.7278,
+      "step": 4778
+    },
+    {
+      "epoch": 0.22643923240938166,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017582133709319298,
+      "loss": 0.2357,
+      "step": 4779
+    },
+    {
+      "epoch": 0.22648661454631605,
+      "grad_norm": 0.033203125,
+      "learning_rate": 0.00017581162618438505,
+      "loss": 0.0013,
+      "step": 4780
+    },
+    {
+      "epoch": 0.22653399668325042,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00017580191359415595,
+      "loss": 0.0245,
+      "step": 4781
+    },
+    {
+      "epoch": 0.22658137882018478,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017579219932272117,
+      "loss": 1.1036,
+      "step": 4782
+    },
+    {
+      "epoch": 0.22662876095711917,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017578248337029613,
+      "loss": 0.8106,
+      "step": 4783
+    },
+    {
+      "epoch": 0.22667614309405354,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017577276573709635,
+      "loss": 0.772,
+      "step": 4784
+    },
+    {
+      "epoch": 0.2267235252309879,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017576304642333732,
+      "loss": 1.1472,
+      "step": 4785
+    },
+    {
+      "epoch": 0.2267709073679223,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017575332542923465,
+      "loss": 0.1988,
+      "step": 4786
+    },
+    {
+      "epoch": 0.22681828950485666,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017574360275500386,
+      "loss": 1.0495,
+      "step": 4787
+    },
+    {
+      "epoch": 0.22686567164179106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017573387840086067,
+      "loss": 0.1959,
+      "step": 4788
+    },
+    {
+      "epoch": 0.22691305377872542,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017572415236702068,
+      "loss": 0.2046,
+      "step": 4789
+    },
+    {
+      "epoch": 0.2269604359156598,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001757144246536997,
+      "loss": 0.9627,
+      "step": 4790
+    },
+    {
+      "epoch": 0.22700781805259418,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001757046952611134,
+      "loss": 0.1513,
+      "step": 4791
+    },
+    {
+      "epoch": 0.22705520018952854,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001756949641894776,
+      "loss": 0.1538,
+      "step": 4792
+    },
+    {
+      "epoch": 0.22710258232646294,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017568523143900815,
+      "loss": 0.9176,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2271499644633973,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017567549700992085,
+      "loss": 0.9416,
+      "step": 4794
+    },
+    {
+      "epoch": 0.22719734660033167,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017566576090243164,
+      "loss": 1.1508,
+      "step": 4795
+    },
+    {
+      "epoch": 0.22724472873726606,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017565602311675647,
+      "loss": 1.1279,
+      "step": 4796
+    },
+    {
+      "epoch": 0.22729211087420043,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017564628365311129,
+      "loss": 0.9059,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2273394930111348,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017563654251171208,
+      "loss": 0.0381,
+      "step": 4798
+    },
+    {
+      "epoch": 0.22738687514806918,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017562679969277496,
+      "loss": 0.2966,
+      "step": 4799
+    },
+    {
+      "epoch": 0.22743425728500355,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017561705519651593,
+      "loss": 1.2171,
+      "step": 4800
+    },
+    {
+      "epoch": 0.22748163942193794,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001756073090231512,
+      "loss": 1.2832,
+      "step": 4801
+    },
+    {
+      "epoch": 0.2275290215588723,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017559756117289688,
+      "loss": 0.7064,
+      "step": 4802
+    },
+    {
+      "epoch": 0.22757640369580667,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017558781164596918,
+      "loss": 0.7603,
+      "step": 4803
+    },
+    {
+      "epoch": 0.22762378583274107,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017557806044258433,
+      "loss": 0.9623,
+      "step": 4804
+    },
+    {
+      "epoch": 0.22767116796967543,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017556830756295858,
+      "loss": 1.2106,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2277185501066098,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017555855300730827,
+      "loss": 1.2195,
+      "step": 4806
+    },
+    {
+      "epoch": 0.2277659322435442,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017554879677584973,
+      "loss": 0.9812,
+      "step": 4807
+    },
+    {
+      "epoch": 0.22781331438047855,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017553903886879937,
+      "loss": 0.0847,
+      "step": 4808
+    },
+    {
+      "epoch": 0.22786069651741295,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001755292792863736,
+      "loss": 1.2341,
+      "step": 4809
+    },
+    {
+      "epoch": 0.2279080786543473,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017551951802878882,
+      "loss": 0.7845,
+      "step": 4810
+    },
+    {
+      "epoch": 0.22795546079128168,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017550975509626163,
+      "loss": 0.1396,
+      "step": 4811
+    },
+    {
+      "epoch": 0.22800284292821607,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017549999048900846,
+      "loss": 1.2626,
+      "step": 4812
+    },
+    {
+      "epoch": 0.22805022506515044,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001754902242072459,
+      "loss": 1.0554,
+      "step": 4813
+    },
+    {
+      "epoch": 0.2280976072020848,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001754804562511906,
+      "loss": 0.6975,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2281449893390192,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001754706866210592,
+      "loss": 0.1331,
+      "step": 4815
+    },
+    {
+      "epoch": 0.22819237147595356,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017546091531706832,
+      "loss": 1.4851,
+      "step": 4816
+    },
+    {
+      "epoch": 0.22823975361288795,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017545114233943472,
+      "loss": 0.8983,
+      "step": 4817
+    },
+    {
+      "epoch": 0.22828713574982232,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001754413676883752,
+      "loss": 1.0767,
+      "step": 4818
+    },
+    {
+      "epoch": 0.22833451788675668,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017543159136410643,
+      "loss": 1.3045,
+      "step": 4819
+    },
+    {
+      "epoch": 0.22838190002369108,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017542181336684535,
+      "loss": 0.7192,
+      "step": 4820
+    },
+    {
+      "epoch": 0.22842928216062544,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017541203369680875,
+      "loss": 0.5849,
+      "step": 4821
+    },
+    {
+      "epoch": 0.2284766642975598,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001754022523542136,
+      "loss": 0.9732,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2285240464344942,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017539246933927682,
+      "loss": 1.1261,
+      "step": 4823
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017538268465221534,
+      "loss": 0.9075,
+      "step": 4824
+    },
+    {
+      "epoch": 0.22861881070836296,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017537289829324624,
+      "loss": 0.134,
+      "step": 4825
+    },
+    {
+      "epoch": 0.22866619284529732,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017536311026258652,
+      "loss": 0.6335,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2287135749822317,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017535332056045332,
+      "loss": 0.5317,
+      "step": 4827
+    },
+    {
+      "epoch": 0.22876095711916608,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001753435291870637,
+      "loss": 0.789,
+      "step": 4828
+    },
+    {
+      "epoch": 0.22880833925610045,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017533373614263487,
+      "loss": 1.2605,
+      "step": 4829
+    },
+    {
+      "epoch": 0.22885572139303484,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000175323941427384,
+      "loss": 0.1808,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2289031035299692,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00017531414504152833,
+      "loss": 0.01,
+      "step": 4831
+    },
+    {
+      "epoch": 0.22895048566690357,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017530434698528516,
+      "loss": 0.9347,
+      "step": 4832
+    },
+    {
+      "epoch": 0.22899786780383796,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017529454725887178,
+      "loss": 0.5792,
+      "step": 4833
+    },
+    {
+      "epoch": 0.22904524994077233,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017528474586250554,
+      "loss": 1.219,
+      "step": 4834
+    },
+    {
+      "epoch": 0.2290926320777067,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017527494279640383,
+      "loss": 0.7787,
+      "step": 4835
+    },
+    {
+      "epoch": 0.22914001421464109,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017526513806078407,
+      "loss": 0.8009,
+      "step": 4836
+    },
+    {
+      "epoch": 0.22918739635157545,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017525533165586374,
+      "loss": 1.0709,
+      "step": 4837
+    },
+    {
+      "epoch": 0.22923477848850984,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017524552358186027,
+      "loss": 0.8568,
+      "step": 4838
+    },
+    {
+      "epoch": 0.2292821606254442,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00017523571383899127,
+      "loss": 0.012,
+      "step": 4839
+    },
+    {
+      "epoch": 0.22932954276237857,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017522590242747426,
+      "loss": 1.3154,
+      "step": 4840
+    },
+    {
+      "epoch": 0.22937692489931297,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017521608934752684,
+      "loss": 0.8839,
+      "step": 4841
+    },
+    {
+      "epoch": 0.22942430703624733,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00017520627459936673,
+      "loss": 0.1537,
+      "step": 4842
+    },
+    {
+      "epoch": 0.2294716891731817,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001751964581832115,
+      "loss": 1.0254,
+      "step": 4843
+    },
+    {
+      "epoch": 0.2295190713101161,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017518664009927895,
+      "loss": 0.2249,
+      "step": 4844
+    },
+    {
+      "epoch": 0.22956645344705046,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001751768203477868,
+      "loss": 1.1852,
+      "step": 4845
+    },
+    {
+      "epoch": 0.22961383558398485,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017516699892895286,
+      "loss": 0.4592,
+      "step": 4846
+    },
+    {
+      "epoch": 0.22966121772091921,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017515717584299493,
+      "loss": 1.0775,
+      "step": 4847
+    },
+    {
+      "epoch": 0.22970859985785358,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001751473510901309,
+      "loss": 0.9325,
+      "step": 4848
+    },
+    {
+      "epoch": 0.22975598199478797,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017513752467057867,
+      "loss": 0.1666,
+      "step": 4849
+    },
+    {
+      "epoch": 0.22980336413172234,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017512769658455617,
+      "loss": 0.8059,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2298507462686567,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0001751178668322814,
+      "loss": 0.1496,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2298981284055911,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017510803541397234,
+      "loss": 1.0282,
+      "step": 4852
+    },
+    {
+      "epoch": 0.22994551054252546,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017509820232984705,
+      "loss": 0.0131,
+      "step": 4853
+    },
+    {
+      "epoch": 0.22999289267945985,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017508836758012365,
+      "loss": 1.2326,
+      "step": 4854
+    },
+    {
+      "epoch": 0.23004027481639422,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017507853116502023,
+      "loss": 1.0969,
+      "step": 4855
+    },
+    {
+      "epoch": 0.23008765695332858,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017506869308475494,
+      "loss": 1.248,
+      "step": 4856
+    },
+    {
+      "epoch": 0.23013503909026298,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000175058853339546,
+      "loss": 0.1875,
+      "step": 4857
+    },
+    {
+      "epoch": 0.23018242122719734,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017504901192961163,
+      "loss": 0.5449,
+      "step": 4858
+    },
+    {
+      "epoch": 0.23022980336413174,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017503916885517015,
+      "loss": 1.0936,
+      "step": 4859
+    },
+    {
+      "epoch": 0.2302771855010661,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0001750293241164398,
+      "loss": 0.0034,
+      "step": 4860
+    },
+    {
+      "epoch": 0.23032456763800047,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017501947771363896,
+      "loss": 0.1314,
+      "step": 4861
+    },
+    {
+      "epoch": 0.23037194977493486,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017500962964698603,
+      "loss": 1.2821,
+      "step": 4862
+    },
+    {
+      "epoch": 0.23041933191186922,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001749997799166994,
+      "loss": 0.0218,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2304667140488036,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017498992852299757,
+      "loss": 0.8939,
+      "step": 4864
+    },
+    {
+      "epoch": 0.23051409618573798,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017498007546609895,
+      "loss": 1.309,
+      "step": 4865
+    },
+    {
+      "epoch": 0.23056147832267235,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017497022074622215,
+      "loss": 0.9838,
+      "step": 4866
+    },
+    {
+      "epoch": 0.23060886045960674,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001749603643635857,
+      "loss": 1.042,
+      "step": 4867
+    },
+    {
+      "epoch": 0.2306562425965411,
+      "grad_norm": 0.0213623046875,
+      "learning_rate": 0.0001749505063184082,
+      "loss": 0.0017,
+      "step": 4868
+    },
+    {
+      "epoch": 0.23070362473347547,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017494064661090835,
+      "loss": 0.8515,
+      "step": 4869
+    },
+    {
+      "epoch": 0.23075100687040986,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017493078524130474,
+      "loss": 1.0213,
+      "step": 4870
+    },
+    {
+      "epoch": 0.23079838900734423,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00017492092220981612,
+      "loss": 0.1837,
+      "step": 4871
+    },
+    {
+      "epoch": 0.2308457711442786,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001749110575166613,
+      "loss": 0.3354,
+      "step": 4872
+    },
+    {
+      "epoch": 0.230893153281213,
+      "grad_norm": 0.010009765625,
+      "learning_rate": 0.00017490119116205896,
+      "loss": 0.0006,
+      "step": 4873
+    },
+    {
+      "epoch": 0.23094053541814735,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000174891323146228,
+      "loss": 0.6591,
+      "step": 4874
+    },
+    {
+      "epoch": 0.23098791755508175,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00017488145346938728,
+      "loss": 0.0333,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2310352996920161,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017487158213175564,
+      "loss": 1.029,
+      "step": 4876
+    },
+    {
+      "epoch": 0.23108268182895048,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017486170913355212,
+      "loss": 0.6199,
+      "step": 4877
+    },
+    {
+      "epoch": 0.23113006396588487,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001748518344749956,
+      "loss": 0.8363,
+      "step": 4878
+    },
+    {
+      "epoch": 0.23117744610281923,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0001748419581563051,
+      "loss": 0.0018,
+      "step": 4879
+    },
+    {
+      "epoch": 0.2312248282397536,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001748320801776997,
+      "loss": 1.0243,
+      "step": 4880
+    },
+    {
+      "epoch": 0.231272210376688,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001748222005393985,
+      "loss": 1.0112,
+      "step": 4881
+    },
+    {
+      "epoch": 0.23131959251362236,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00017481231924162054,
+      "loss": 0.035,
+      "step": 4882
+    },
+    {
+      "epoch": 0.23136697465055675,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00017480243628458504,
+      "loss": 0.0418,
+      "step": 4883
+    },
+    {
+      "epoch": 0.23141435678749112,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001747925516685112,
+      "loss": 0.6955,
+      "step": 4884
+    },
+    {
+      "epoch": 0.23146173892442548,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0001747826653936182,
+      "loss": 0.1441,
+      "step": 4885
+    },
+    {
+      "epoch": 0.23150912106135987,
+      "grad_norm": 0.01904296875,
+      "learning_rate": 0.00017477277746012537,
+      "loss": 0.0011,
+      "step": 4886
+    },
+    {
+      "epoch": 0.23155650319829424,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017476288786825195,
+      "loss": 0.4778,
+      "step": 4887
+    },
+    {
+      "epoch": 0.23160388533522863,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001747529966182173,
+      "loss": 0.5974,
+      "step": 4888
+    },
+    {
+      "epoch": 0.231651267472163,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017474310371024085,
+      "loss": 1.2586,
+      "step": 4889
+    },
+    {
+      "epoch": 0.23169864960909736,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017473320914454193,
+      "loss": 0.9656,
+      "step": 4890
+    },
+    {
+      "epoch": 0.23174603174603176,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001747233129213401,
+      "loss": 1.1482,
+      "step": 4891
+    },
+    {
+      "epoch": 0.23179341388296612,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00017471341504085472,
+      "loss": 0.1904,
+      "step": 4892
+    },
+    {
+      "epoch": 0.23184079601990049,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001747035155033054,
+      "loss": 0.7755,
+      "step": 4893
+    },
+    {
+      "epoch": 0.23188817815683488,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017469361430891167,
+      "loss": 0.0128,
+      "step": 4894
+    },
+    {
+      "epoch": 0.23193556029376924,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017468371145789314,
+      "loss": 0.1607,
+      "step": 4895
+    },
+    {
+      "epoch": 0.23198294243070364,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017467380695046942,
+      "loss": 1.0165,
+      "step": 4896
+    },
+    {
+      "epoch": 0.232030324567638,
+      "grad_norm": 0.028076171875,
+      "learning_rate": 0.00017466390078686023,
+      "loss": 0.0017,
+      "step": 4897
+    },
+    {
+      "epoch": 0.23207770670457237,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017465399296728523,
+      "loss": 0.9253,
+      "step": 4898
+    },
+    {
+      "epoch": 0.23212508884150676,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017464408349196416,
+      "loss": 0.0854,
+      "step": 4899
+    },
+    {
+      "epoch": 0.23217247097844113,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017463417236111684,
+      "loss": 1.1781,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2322198531153755,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001746242595749631,
+      "loss": 0.3372,
+      "step": 4901
+    },
+    {
+      "epoch": 0.23226723525230988,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017461434513372275,
+      "loss": 1.125,
+      "step": 4902
+    },
+    {
+      "epoch": 0.23231461738924425,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00017460442903761573,
+      "loss": 0.0674,
+      "step": 4903
+    },
+    {
+      "epoch": 0.23236199952617864,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017459451128686192,
+      "loss": 1.2133,
+      "step": 4904
+    },
+    {
+      "epoch": 0.232409381663113,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017458459188168132,
+      "loss": 0.3097,
+      "step": 4905
+    },
+    {
+      "epoch": 0.23245676380004737,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017457467082229386,
+      "loss": 0.1143,
+      "step": 4906
+    },
+    {
+      "epoch": 0.23250414593698177,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001745647481089197,
+      "loss": 0.0897,
+      "step": 4907
+    },
+    {
+      "epoch": 0.23255152807391613,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001745548237417788,
+      "loss": 0.6822,
+      "step": 4908
+    },
+    {
+      "epoch": 0.2325989102108505,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017454489772109134,
+      "loss": 1.2888,
+      "step": 4909
+    },
+    {
+      "epoch": 0.2326462923477849,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017453497004707747,
+      "loss": 0.613,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23269367448471925,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017452504071995733,
+      "loss": 0.8391,
+      "step": 4911
+    },
+    {
+      "epoch": 0.23274105662165365,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017451510973995115,
+      "loss": 1.3987,
+      "step": 4912
+    },
+    {
+      "epoch": 0.232788438758588,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017450517710727924,
+      "loss": 0.8413,
+      "step": 4913
+    },
+    {
+      "epoch": 0.23283582089552238,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017449524282216186,
+      "loss": 0.1535,
+      "step": 4914
+    },
+    {
+      "epoch": 0.23288320303245677,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017448530688481934,
+      "loss": 0.9345,
+      "step": 4915
+    },
+    {
+      "epoch": 0.23293058516939114,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017447536929547202,
+      "loss": 0.4542,
+      "step": 4916
+    },
+    {
+      "epoch": 0.23297796730632553,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017446543005434037,
+      "loss": 0.7109,
+      "step": 4917
+    },
+    {
+      "epoch": 0.2330253494432599,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001744554891616448,
+      "loss": 0.0557,
+      "step": 4918
+    },
+    {
+      "epoch": 0.23307273158019426,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017444554661760577,
+      "loss": 1.5583,
+      "step": 4919
+    },
+    {
+      "epoch": 0.23312011371712865,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017443560242244384,
+      "loss": 0.8715,
+      "step": 4920
+    },
+    {
+      "epoch": 0.23316749585406302,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001744256565763795,
+      "loss": 0.6857,
+      "step": 4921
+    },
+    {
+      "epoch": 0.23321487799099738,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001744157090796334,
+      "loss": 1.169,
+      "step": 4922
+    },
+    {
+      "epoch": 0.23326226012793178,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001744057599324261,
+      "loss": 0.8634,
+      "step": 4923
+    },
+    {
+      "epoch": 0.23330964226486614,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017439580913497832,
+      "loss": 1.1702,
+      "step": 4924
+    },
+    {
+      "epoch": 0.23335702440180053,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017438585668751074,
+      "loss": 1.1863,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2334044065387349,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001743759025902441,
+      "loss": 1.2634,
+      "step": 4926
+    },
+    {
+      "epoch": 0.23345178867566926,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00017436594684339912,
+      "loss": 0.0342,
+      "step": 4927
+    },
+    {
+      "epoch": 0.23349917081260366,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001743559894471967,
+      "loss": 0.1795,
+      "step": 4928
+    },
+    {
+      "epoch": 0.23354655294953802,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017434603040185763,
+      "loss": 0.7102,
+      "step": 4929
+    },
+    {
+      "epoch": 0.2335939350864724,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017433606970760276,
+      "loss": 1.11,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23364131722340678,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017432610736465307,
+      "loss": 0.1318,
+      "step": 4931
+    },
+    {
+      "epoch": 0.23368869936034115,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017431614337322948,
+      "loss": 1.1823,
+      "step": 4932
+    },
+    {
+      "epoch": 0.23373608149727554,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017430617773355297,
+      "loss": 0.7964,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2337834636342099,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017429621044584464,
+      "loss": 0.956,
+      "step": 4934
+    },
+    {
+      "epoch": 0.23383084577114427,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017428624151032544,
+      "loss": 0.1908,
+      "step": 4935
+    },
+    {
+      "epoch": 0.23387822790807866,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017427627092721654,
+      "loss": 0.8539,
+      "step": 4936
+    },
+    {
+      "epoch": 0.23392561004501303,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001742662986967391,
+      "loss": 0.4029,
+      "step": 4937
+    },
+    {
+      "epoch": 0.2339729921819474,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00017425632481911423,
+      "loss": 0.0029,
+      "step": 4938
+    },
+    {
+      "epoch": 0.23402037431888179,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001742463492945632,
+      "loss": 1.1076,
+      "step": 4939
+    },
+    {
+      "epoch": 0.23406775645581615,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00017423637212330716,
+      "loss": 0.6563,
+      "step": 4940
+    },
+    {
+      "epoch": 0.23411513859275054,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017422639330556754,
+      "loss": 0.7361,
+      "step": 4941
+    },
+    {
+      "epoch": 0.2341625207296849,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017421641284156553,
+      "loss": 1.0227,
+      "step": 4942
+    },
+    {
+      "epoch": 0.23420990286661927,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017420643073152254,
+      "loss": 0.5419,
+      "step": 4943
+    },
+    {
+      "epoch": 0.23425728500355367,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017419644697565996,
+      "loss": 0.9013,
+      "step": 4944
+    },
+    {
+      "epoch": 0.23430466714048803,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017418646157419922,
+      "loss": 0.0414,
+      "step": 4945
+    },
+    {
+      "epoch": 0.23435204927742243,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017417647452736178,
+      "loss": 0.0804,
+      "step": 4946
+    },
+    {
+      "epoch": 0.2343994314143568,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017416648583536915,
+      "loss": 0.9069,
+      "step": 4947
+    },
+    {
+      "epoch": 0.23444681355129116,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017415649549844286,
+      "loss": 1.0225,
+      "step": 4948
+    },
+    {
+      "epoch": 0.23449419568822555,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017414650351680447,
+      "loss": 1.2589,
+      "step": 4949
+    },
+    {
+      "epoch": 0.2345415778251599,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017413650989067564,
+      "loss": 0.8184,
+      "step": 4950
+    },
+    {
+      "epoch": 0.23458895996209428,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017412651462027798,
+      "loss": 1.5527,
+      "step": 4951
+    },
+    {
+      "epoch": 0.23463634209902867,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017411651770583318,
+      "loss": 0.1233,
+      "step": 4952
+    },
+    {
+      "epoch": 0.23468372423596304,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00017410651914756295,
+      "loss": 1.028,
+      "step": 4953
+    },
+    {
+      "epoch": 0.23473110637289743,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017409651894568907,
+      "loss": 0.8294,
+      "step": 4954
+    },
+    {
+      "epoch": 0.2347784885098318,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017408651710043333,
+      "loss": 0.0641,
+      "step": 4955
+    },
+    {
+      "epoch": 0.23482587064676616,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017407651361201756,
+      "loss": 1.0256,
+      "step": 4956
+    },
+    {
+      "epoch": 0.23487325278370055,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001740665084806636,
+      "loss": 1.0128,
+      "step": 4957
+    },
+    {
+      "epoch": 0.23492063492063492,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017405650170659339,
+      "loss": 1.0555,
+      "step": 4958
+    },
+    {
+      "epoch": 0.23496801705756928,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017404649329002883,
+      "loss": 0.9352,
+      "step": 4959
+    },
+    {
+      "epoch": 0.23501539919450368,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00017403648323119196,
+      "loss": 0.1448,
+      "step": 4960
+    },
+    {
+      "epoch": 0.23506278133143804,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001740264715303047,
+      "loss": 1.1987,
+      "step": 4961
+    },
+    {
+      "epoch": 0.23511016346837244,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00017401645818758917,
+      "loss": 0.0393,
+      "step": 4962
+    },
+    {
+      "epoch": 0.2351575456053068,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017400644320326745,
+      "loss": 0.1885,
+      "step": 4963
+    },
+    {
+      "epoch": 0.23520492774224117,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017399642657756162,
+      "loss": 0.9889,
+      "step": 4964
+    },
+    {
+      "epoch": 0.23525230987917556,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001739864083106939,
+      "loss": 0.9626,
+      "step": 4965
+    },
+    {
+      "epoch": 0.23529969201610992,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017397638840288643,
+      "loss": 0.9734,
+      "step": 4966
+    },
+    {
+      "epoch": 0.2353470741530443,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017396636685436149,
+      "loss": 0.3847,
+      "step": 4967
+    },
+    {
+      "epoch": 0.23539445628997868,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017395634366534131,
+      "loss": 1.0344,
+      "step": 4968
+    },
+    {
+      "epoch": 0.23544183842691305,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00017394631883604818,
+      "loss": 1.1247,
+      "step": 4969
+    },
+    {
+      "epoch": 0.23548922056384744,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00017393629236670446,
+      "loss": 0.0628,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2355366027007818,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017392626425753255,
+      "loss": 0.7047,
+      "step": 4971
+    },
+    {
+      "epoch": 0.23558398483771617,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017391623450875482,
+      "loss": 1.1307,
+      "step": 4972
+    },
+    {
+      "epoch": 0.23563136697465056,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017390620312059376,
+      "loss": 1.0037,
+      "step": 4973
+    },
+    {
+      "epoch": 0.23567874911158493,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00017389617009327184,
+      "loss": 1.0204,
+      "step": 4974
+    },
+    {
+      "epoch": 0.23572613124851932,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017388613542701156,
+      "loss": 0.1655,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2357735133854537,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001738760991220355,
+      "loss": 0.9813,
+      "step": 4976
+    },
+    {
+      "epoch": 0.23582089552238805,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017386606117856626,
+      "loss": 0.0226,
+      "step": 4977
+    },
+    {
+      "epoch": 0.23586827765932245,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017385602159682647,
+      "loss": 1.083,
+      "step": 4978
+    },
+    {
+      "epoch": 0.2359156597962568,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00017384598037703877,
+      "loss": 1.045,
+      "step": 4979
+    },
+    {
+      "epoch": 0.23596304193319118,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001738359375194259,
+      "loss": 1.4046,
+      "step": 4980
+    },
+    {
+      "epoch": 0.23601042407012557,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017382589302421055,
+      "loss": 0.4487,
+      "step": 4981
+    },
+    {
+      "epoch": 0.23605780620705993,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017381584689161555,
+      "loss": 1.0711,
+      "step": 4982
+    },
+    {
+      "epoch": 0.23610518834399433,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001738057991218637,
+      "loss": 0.8805,
+      "step": 4983
+    },
+    {
+      "epoch": 0.2361525704809287,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00017379574971517782,
+      "loss": 0.5025,
+      "step": 4984
+    },
+    {
+      "epoch": 0.23619995261786306,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017378569867178083,
+      "loss": 1.1161,
+      "step": 4985
+    },
+    {
+      "epoch": 0.23624733475479745,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017377564599189562,
+      "loss": 0.0181,
+      "step": 4986
+    },
+    {
+      "epoch": 0.23629471689173182,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00017376559167574517,
+      "loss": 0.1703,
+      "step": 4987
+    },
+    {
+      "epoch": 0.23634209902866618,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017375553572355248,
+      "loss": 0.5635,
+      "step": 4988
+    },
+    {
+      "epoch": 0.23638948116560057,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00017374547813554057,
+      "loss": 1.1397,
+      "step": 4989
+    },
+    {
+      "epoch": 0.23643686330253494,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001737354189119325,
+      "loss": 1.2917,
+      "step": 4990
+    },
+    {
+      "epoch": 0.23648424543946933,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00017372535805295136,
+      "loss": 0.1013,
+      "step": 4991
+    },
+    {
+      "epoch": 0.2365316275764037,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017371529555882032,
+      "loss": 0.8985,
+      "step": 4992
+    },
+    {
+      "epoch": 0.23657900971333806,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017370523142976255,
+      "loss": 0.9875,
+      "step": 4993
+    },
+    {
+      "epoch": 0.23662639185027246,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017369516566600126,
+      "loss": 0.7521,
+      "step": 4994
+    },
+    {
+      "epoch": 0.23667377398720682,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017368509826775968,
+      "loss": 1.1462,
+      "step": 4995
+    },
+    {
+      "epoch": 0.23672115612414119,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017367502923526108,
+      "loss": 1.8729,
+      "step": 4996
+    },
+    {
+      "epoch": 0.23676853826107558,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017366495856872884,
+      "loss": 0.9345,
+      "step": 4997
+    },
+    {
+      "epoch": 0.23681592039800994,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017365488626838632,
+      "loss": 1.3382,
+      "step": 4998
+    },
+    {
+      "epoch": 0.23686330253494434,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001736448123344568,
+      "loss": 0.9335,
+      "step": 4999
+    },
+    {
+      "epoch": 0.2369106846718787,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017363473676716384,
+      "loss": 0.7158,
+      "step": 5000
+    },
+    {
+      "epoch": 0.23695806680881307,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017362465956673078,
+      "loss": 1.3346,
+      "step": 5001
+    },
+    {
+      "epoch": 0.23700544894574746,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017361458073338127,
+      "loss": 1.39,
+      "step": 5002
+    },
+    {
+      "epoch": 0.23705283108268183,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017360450026733873,
+      "loss": 0.6038,
+      "step": 5003
+    },
+    {
+      "epoch": 0.23710021321961622,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001735944181688268,
+      "loss": 0.9036,
+      "step": 5004
+    },
+    {
+      "epoch": 0.23714759535655058,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017358433443806905,
+      "loss": 1.1041,
+      "step": 5005
+    },
+    {
+      "epoch": 0.23719497749348495,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017357424907528914,
+      "loss": 0.152,
+      "step": 5006
+    },
+    {
+      "epoch": 0.23724235963041934,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017356416208071074,
+      "loss": 0.0677,
+      "step": 5007
+    },
+    {
+      "epoch": 0.2372897417673537,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00017355407345455762,
+      "loss": 0.1848,
+      "step": 5008
+    },
+    {
+      "epoch": 0.23733712390428807,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00017354398319705346,
+      "loss": 1.084,
+      "step": 5009
+    },
+    {
+      "epoch": 0.23738450604122247,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001735338913084221,
+      "loss": 1.2232,
+      "step": 5010
+    },
+    {
+      "epoch": 0.23743188817815683,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017352379778888736,
+      "loss": 0.6343,
+      "step": 5011
+    },
+    {
+      "epoch": 0.23747927031509122,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001735137026386731,
+      "loss": 0.811,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2375266524520256,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001735036058580032,
+      "loss": 0.4561,
+      "step": 5013
+    },
+    {
+      "epoch": 0.23757403458895995,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017349350744710163,
+      "loss": 0.6335,
+      "step": 5014
+    },
+    {
+      "epoch": 0.23762141672589435,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017348340740619235,
+      "loss": 0.1175,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2376687988628287,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017347330573549936,
+      "loss": 1.1408,
+      "step": 5016
+    },
+    {
+      "epoch": 0.23771618099976308,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001734632024352467,
+      "loss": 0.99,
+      "step": 5017
+    },
+    {
+      "epoch": 0.23776356313669747,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017345309750565848,
+      "loss": 0.1639,
+      "step": 5018
+    },
+    {
+      "epoch": 0.23781094527363184,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001734429909469588,
+      "loss": 0.7657,
+      "step": 5019
+    },
+    {
+      "epoch": 0.23785832741056623,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00017343288275937176,
+      "loss": 0.1353,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2379057095475006,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00017342277294312165,
+      "loss": 0.1467,
+      "step": 5021
+    },
+    {
+      "epoch": 0.23795309168443496,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017341266149843262,
+      "loss": 0.7441,
+      "step": 5022
+    },
+    {
+      "epoch": 0.23800047382136935,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017340254842552897,
+      "loss": 0.1331,
+      "step": 5023
+    },
+    {
+      "epoch": 0.23804785595830372,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00017339243372463495,
+      "loss": 1.0166,
+      "step": 5024
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017338231739597496,
+      "loss": 0.151,
+      "step": 5025
+    },
+    {
+      "epoch": 0.23814262023217247,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017337219943977332,
+      "loss": 0.0549,
+      "step": 5026
+    },
+    {
+      "epoch": 0.23819000236910684,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00017336207985625443,
+      "loss": 0.4165,
+      "step": 5027
+    },
+    {
+      "epoch": 0.23823738450604123,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017335195864564277,
+      "loss": 1.1232,
+      "step": 5028
+    },
+    {
+      "epoch": 0.2382847666429756,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017334183580816279,
+      "loss": 1.1028,
+      "step": 5029
+    },
+    {
+      "epoch": 0.23833214877990996,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.000173331711344039,
+      "loss": 0.9572,
+      "step": 5030
+    },
+    {
+      "epoch": 0.23837953091684436,
+      "grad_norm": 0.75,
+      "learning_rate": 0.000173321585253496,
+      "loss": 1.3829,
+      "step": 5031
+    },
+    {
+      "epoch": 0.23842691305377872,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0001733114575367583,
+      "loss": 0.0021,
+      "step": 5032
+    },
+    {
+      "epoch": 0.23847429519071311,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017330132819405058,
+      "loss": 0.3454,
+      "step": 5033
+    },
+    {
+      "epoch": 0.23852167732764748,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017329119722559749,
+      "loss": 0.557,
+      "step": 5034
+    },
+    {
+      "epoch": 0.23856905946458185,
+      "grad_norm": 0.017578125,
+      "learning_rate": 0.00017328106463162369,
+      "loss": 0.0011,
+      "step": 5035
+    },
+    {
+      "epoch": 0.23861644160151624,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001732709304123539,
+      "loss": 0.0309,
+      "step": 5036
+    },
+    {
+      "epoch": 0.2386638237384506,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017326079456801298,
+      "loss": 0.9946,
+      "step": 5037
+    },
+    {
+      "epoch": 0.23871120587538497,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017325065709882567,
+      "loss": 0.3698,
+      "step": 5038
+    },
+    {
+      "epoch": 0.23875858801231936,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00017324051800501677,
+      "loss": 0.6622,
+      "step": 5039
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00017323037728681122,
+      "loss": 0.2755,
+      "step": 5040
+    },
+    {
+      "epoch": 0.23885335228618812,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00017322023494443386,
+      "loss": 1.3305,
+      "step": 5041
+    },
+    {
+      "epoch": 0.23890073442312248,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001732100909781097,
+      "loss": 1.0599,
+      "step": 5042
+    },
+    {
+      "epoch": 0.23894811656005685,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00017319994538806372,
+      "loss": 0.2098,
+      "step": 5043
+    },
+    {
+      "epoch": 0.23899549869699124,
+      "grad_norm": 0.01708984375,
+      "learning_rate": 0.00017318979817452091,
+      "loss": 0.0013,
+      "step": 5044
+    },
+    {
+      "epoch": 0.2390428808339256,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00017317964933770633,
+      "loss": 0.483,
+      "step": 5045
+    },
+    {
+      "epoch": 0.23909026297085997,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001731694988778451,
+      "loss": 0.0344,
+      "step": 5046
+    },
+    {
+      "epoch": 0.23913764510779437,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001731593467951623,
+      "loss": 1.0277,
+      "step": 5047
+    },
+    {
+      "epoch": 0.23918502724472873,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001731491930898831,
+      "loss": 1.0551,
+      "step": 5048
+    },
+    {
+      "epoch": 0.23923240938166312,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017313903776223274,
+      "loss": 0.8756,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2392797915185975,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001731288808124364,
+      "loss": 0.1851,
+      "step": 5050
+    },
+    {
+      "epoch": 0.23932717365553186,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017311872224071942,
+      "loss": 0.1446,
+      "step": 5051
+    },
+    {
+      "epoch": 0.23937455579246625,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00017310856204730705,
+      "loss": 0.6632,
+      "step": 5052
+    },
+    {
+      "epoch": 0.2394219379294006,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001730984002324246,
+      "loss": 0.022,
+      "step": 5053
+    },
+    {
+      "epoch": 0.23946932006633498,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017308823679629756,
+      "loss": 1.0228,
+      "step": 5054
+    },
+    {
+      "epoch": 0.23951670220326937,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017307807173915123,
+      "loss": 0.9571,
+      "step": 5055
+    },
+    {
+      "epoch": 0.23956408434020374,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017306790506121114,
+      "loss": 1.0204,
+      "step": 5056
+    },
+    {
+      "epoch": 0.23961146647713813,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001730577367627027,
+      "loss": 0.0199,
+      "step": 5057
+    },
+    {
+      "epoch": 0.2396588486140725,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017304756684385152,
+      "loss": 1.0773,
+      "step": 5058
+    },
+    {
+      "epoch": 0.23970623075100686,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017303739530488308,
+      "loss": 1.0701,
+      "step": 5059
+    },
+    {
+      "epoch": 0.23975361288794125,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017302722214602303,
+      "loss": 0.961,
+      "step": 5060
+    },
+    {
+      "epoch": 0.23980099502487562,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017301704736749697,
+      "loss": 0.0564,
+      "step": 5061
+    },
+    {
+      "epoch": 0.23984837716181,
+      "grad_norm": 0.01483154296875,
+      "learning_rate": 0.0001730068709695306,
+      "loss": 0.001,
+      "step": 5062
+    },
+    {
+      "epoch": 0.23989575929874438,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017299669295234955,
+      "loss": 1.3446,
+      "step": 5063
+    },
+    {
+      "epoch": 0.23994314143567874,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001729865133161796,
+      "loss": 1.3723,
+      "step": 5064
+    },
+    {
+      "epoch": 0.23999052357261313,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017297633206124656,
+      "loss": 0.1451,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2400379057095475,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001729661491877762,
+      "loss": 1.1708,
+      "step": 5066
+    },
+    {
+      "epoch": 0.24008528784648187,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017295596469599437,
+      "loss": 1.1451,
+      "step": 5067
+    },
+    {
+      "epoch": 0.24013266998341626,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017294577858612695,
+      "loss": 0.6699,
+      "step": 5068
+    },
+    {
+      "epoch": 0.24018005212035062,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017293559085839987,
+      "loss": 1.2246,
+      "step": 5069
+    },
+    {
+      "epoch": 0.24022743425728502,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017292540151303903,
+      "loss": 0.2145,
+      "step": 5070
+    },
+    {
+      "epoch": 0.24027481639421938,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017291521055027052,
+      "loss": 0.1655,
+      "step": 5071
+    },
+    {
+      "epoch": 0.24032219853115375,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017290501797032027,
+      "loss": 1.3479,
+      "step": 5072
+    },
+    {
+      "epoch": 0.24036958066808814,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001728948237734144,
+      "loss": 0.1186,
+      "step": 5073
+    },
+    {
+      "epoch": 0.2404169628050225,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017288462795977895,
+      "loss": 0.6707,
+      "step": 5074
+    },
+    {
+      "epoch": 0.24046434494195687,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001728744305296401,
+      "loss": 0.7204,
+      "step": 5075
+    },
+    {
+      "epoch": 0.24051172707889126,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000172864231483224,
+      "loss": 0.9086,
+      "step": 5076
+    },
+    {
+      "epoch": 0.24055910921582563,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017285403082075687,
+      "loss": 1.2283,
+      "step": 5077
+    },
+    {
+      "epoch": 0.24060649135276002,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017284382854246494,
+      "loss": 0.8928,
+      "step": 5078
+    },
+    {
+      "epoch": 0.2406538734896944,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00017283362464857446,
+      "loss": 0.3822,
+      "step": 5079
+    },
+    {
+      "epoch": 0.24070125562662875,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017282341913931178,
+      "loss": 0.8967,
+      "step": 5080
+    },
+    {
+      "epoch": 0.24074863776356314,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001728132120149032,
+      "loss": 1.743,
+      "step": 5081
+    },
+    {
+      "epoch": 0.2407960199004975,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001728030032755752,
+      "loss": 1.162,
+      "step": 5082
+    },
+    {
+      "epoch": 0.24084340203743188,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017279279292155408,
+      "loss": 1.3576,
+      "step": 5083
+    },
+    {
+      "epoch": 0.24089078417436627,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00017278258095306637,
+      "loss": 0.0138,
+      "step": 5084
+    },
+    {
+      "epoch": 0.24093816631130063,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017277236737033854,
+      "loss": 0.1336,
+      "step": 5085
+    },
+    {
+      "epoch": 0.24098554844823503,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001727621521735971,
+      "loss": 0.7657,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2410329305851694,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017275193536306864,
+      "loss": 0.1299,
+      "step": 5087
+    },
+    {
+      "epoch": 0.24108031272210376,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017274171693897975,
+      "loss": 0.9155,
+      "step": 5088
+    },
+    {
+      "epoch": 0.24112769485903815,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00017273149690155703,
+      "loss": 0.6494,
+      "step": 5089
+    },
+    {
+      "epoch": 0.24117507699597251,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017272127525102721,
+      "loss": 0.7286,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2412224591329069,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017271105198761694,
+      "loss": 1.1988,
+      "step": 5091
+    },
+    {
+      "epoch": 0.24126984126984127,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017270082711155302,
+      "loss": 0.1147,
+      "step": 5092
+    },
+    {
+      "epoch": 0.24131722340677564,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017269060062306214,
+      "loss": 1.0916,
+      "step": 5093
+    },
+    {
+      "epoch": 0.24136460554371003,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017268037252237122,
+      "loss": 0.0054,
+      "step": 5094
+    },
+    {
+      "epoch": 0.2414119876806444,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00017267014280970702,
+      "loss": 0.4695,
+      "step": 5095
+    },
+    {
+      "epoch": 0.24145936981757876,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00017265991148529648,
+      "loss": 1.0174,
+      "step": 5096
+    },
+    {
+      "epoch": 0.24150675195451315,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001726496785493665,
+      "loss": 0.1565,
+      "step": 5097
+    },
+    {
+      "epoch": 0.24155413409144752,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000172639444002144,
+      "loss": 1.2032,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2416015162283819,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00017262920784385602,
+      "loss": 1.0722,
+      "step": 5099
+    },
+    {
+      "epoch": 0.24164889836531628,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00017261897007472956,
+      "loss": 0.0169,
+      "step": 5100
+    },
+    {
+      "epoch": 0.24169628050225064,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017260873069499172,
+      "loss": 1.778,
+      "step": 5101
+    },
+    {
+      "epoch": 0.24174366263918504,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017259848970486955,
+      "loss": 0.4495,
+      "step": 5102
+    },
+    {
+      "epoch": 0.2417910447761194,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017258824710459023,
+      "loss": 0.9077,
+      "step": 5103
+    },
+    {
+      "epoch": 0.24183842691305377,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001725780028943809,
+      "loss": 0.0096,
+      "step": 5104
+    },
+    {
+      "epoch": 0.24188580904998816,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017256775707446875,
+      "loss": 0.0437,
+      "step": 5105
+    },
+    {
+      "epoch": 0.24193319118692252,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017255750964508107,
+      "loss": 1.0936,
+      "step": 5106
+    },
+    {
+      "epoch": 0.24198057332385692,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017254726060644512,
+      "loss": 0.0998,
+      "step": 5107
+    },
+    {
+      "epoch": 0.24202795546079128,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017253700995878814,
+      "loss": 1.4343,
+      "step": 5108
+    },
+    {
+      "epoch": 0.24207533759772565,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017252675770233758,
+      "loss": 1.1585,
+      "step": 5109
+    },
+    {
+      "epoch": 0.24212271973466004,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001725165038373208,
+      "loss": 0.9669,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2421701018715944,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001725062483639652,
+      "loss": 0.1645,
+      "step": 5111
+    },
+    {
+      "epoch": 0.24221748400852877,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017249599128249825,
+      "loss": 0.8636,
+      "step": 5112
+    },
+    {
+      "epoch": 0.24226486614546316,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00017248573259314739,
+      "loss": 0.1532,
+      "step": 5113
+    },
+    {
+      "epoch": 0.24231224828239753,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017247547229614022,
+      "loss": 1.4756,
+      "step": 5114
+    },
+    {
+      "epoch": 0.24235963041933192,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017246521039170429,
+      "loss": 0.7939,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2424070125562663,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017245494688006716,
+      "loss": 0.8726,
+      "step": 5116
+    },
+    {
+      "epoch": 0.24245439469320065,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017244468176145648,
+      "loss": 1.7517,
+      "step": 5117
+    },
+    {
+      "epoch": 0.24250177683013505,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017243441503609993,
+      "loss": 0.0335,
+      "step": 5118
+    },
+    {
+      "epoch": 0.2425491589670694,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017242414670422523,
+      "loss": 0.7218,
+      "step": 5119
+    },
+    {
+      "epoch": 0.2425965411040038,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017241387676606004,
+      "loss": 0.6616,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24264392324093817,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017240360522183224,
+      "loss": 0.0204,
+      "step": 5121
+    },
+    {
+      "epoch": 0.24269130537787253,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001723933320717696,
+      "loss": 0.7932,
+      "step": 5122
+    },
+    {
+      "epoch": 0.24273868751480693,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00017238305731609997,
+      "loss": 0.1196,
+      "step": 5123
+    },
+    {
+      "epoch": 0.2427860696517413,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017237278095505118,
+      "loss": 0.6768,
+      "step": 5124
+    },
+    {
+      "epoch": 0.24283345178867566,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017236250298885124,
+      "loss": 0.7022,
+      "step": 5125
+    },
+    {
+      "epoch": 0.24288083392561005,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00017235222341772802,
+      "loss": 1.0216,
+      "step": 5126
+    },
+    {
+      "epoch": 0.24292821606254442,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017234194224190961,
+      "loss": 0.9276,
+      "step": 5127
+    },
+    {
+      "epoch": 0.2429755981994788,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017233165946162394,
+      "loss": 1.2332,
+      "step": 5128
+    },
+    {
+      "epoch": 0.24302298033641317,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00017232137507709912,
+      "loss": 0.0201,
+      "step": 5129
+    },
+    {
+      "epoch": 0.24307036247334754,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001723110890885632,
+      "loss": 0.7718,
+      "step": 5130
+    },
+    {
+      "epoch": 0.24311774461028193,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001723008014962444,
+      "loss": 1.0057,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2431651267472163,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017229051230037082,
+      "loss": 0.9161,
+      "step": 5132
+    },
+    {
+      "epoch": 0.24321250888415066,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017228022150117065,
+      "loss": 1.0776,
+      "step": 5133
+    },
+    {
+      "epoch": 0.24325989102108506,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017226992909887215,
+      "loss": 0.0449,
+      "step": 5134
+    },
+    {
+      "epoch": 0.24330727315801942,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001722596350937036,
+      "loss": 1.0285,
+      "step": 5135
+    },
+    {
+      "epoch": 0.24335465529495381,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017224933948589336,
+      "loss": 0.0313,
+      "step": 5136
+    },
+    {
+      "epoch": 0.24340203743188818,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001722390422756697,
+      "loss": 1.1853,
+      "step": 5137
+    },
+    {
+      "epoch": 0.24344941956882254,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00017222874346326103,
+      "loss": 0.7152,
+      "step": 5138
+    },
+    {
+      "epoch": 0.24349680170575694,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017221844304889577,
+      "loss": 0.1681,
+      "step": 5139
+    },
+    {
+      "epoch": 0.2435441838426913,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017220814103280233,
+      "loss": 0.7481,
+      "step": 5140
+    },
+    {
+      "epoch": 0.24359156597962567,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001721978374152093,
+      "loss": 0.026,
+      "step": 5141
+    },
+    {
+      "epoch": 0.24363894811656006,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001721875321963451,
+      "loss": 1.3398,
+      "step": 5142
+    },
+    {
+      "epoch": 0.24368633025349443,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001721772253764383,
+      "loss": 0.5379,
+      "step": 5143
+    },
+    {
+      "epoch": 0.24373371239042882,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00017216691695571756,
+      "loss": 1.3066,
+      "step": 5144
+    },
+    {
+      "epoch": 0.24378109452736318,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017215660693441147,
+      "loss": 0.4534,
+      "step": 5145
+    },
+    {
+      "epoch": 0.24382847666429755,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00017214629531274865,
+      "loss": 0.2493,
+      "step": 5146
+    },
+    {
+      "epoch": 0.24387585880123194,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017213598209095792,
+      "loss": 0.9154,
+      "step": 5147
+    },
+    {
+      "epoch": 0.2439232409381663,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00017212566726926789,
+      "loss": 0.4233,
+      "step": 5148
+    },
+    {
+      "epoch": 0.24397062307510067,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0001721153508479074,
+      "loss": 0.0072,
+      "step": 5149
+    },
+    {
+      "epoch": 0.24401800521203507,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017210503282710527,
+      "loss": 0.8346,
+      "step": 5150
+    },
+    {
+      "epoch": 0.24406538734896943,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017209471320709025,
+      "loss": 0.2293,
+      "step": 5151
+    },
+    {
+      "epoch": 0.24411276948590382,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017208439198809132,
+      "loss": 0.8109,
+      "step": 5152
+    },
+    {
+      "epoch": 0.2441601516228382,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00017207406917033738,
+      "loss": 0.003,
+      "step": 5153
+    },
+    {
+      "epoch": 0.24420753375977255,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001720637447540573,
+      "loss": 1.1695,
+      "step": 5154
+    },
+    {
+      "epoch": 0.24425491589670695,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017205341873948018,
+      "loss": 0.9309,
+      "step": 5155
+    },
+    {
+      "epoch": 0.2443022980336413,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00017204309112683493,
+      "loss": 0.9097,
+      "step": 5156
+    },
+    {
+      "epoch": 0.2443496801705757,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001720327619163507,
+      "loss": 0.0052,
+      "step": 5157
+    },
+    {
+      "epoch": 0.24439706230751007,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017202243110825652,
+      "loss": 0.0227,
+      "step": 5158
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017201209870278152,
+      "loss": 0.726,
+      "step": 5159
+    },
+    {
+      "epoch": 0.24449182658137883,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017200176470015486,
+      "loss": 0.0155,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2445392087183132,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001719914291006058,
+      "loss": 1.3043,
+      "step": 5161
+    },
+    {
+      "epoch": 0.24458659085524756,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001719810919043635,
+      "loss": 0.9903,
+      "step": 5162
+    },
+    {
+      "epoch": 0.24463397299218195,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017197075311165723,
+      "loss": 0.0655,
+      "step": 5163
+    },
+    {
+      "epoch": 0.24468135512911632,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017196041272271635,
+      "loss": 0.936,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2447287372660507,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00017195007073777014,
+      "loss": 0.0852,
+      "step": 5165
+    },
+    {
+      "epoch": 0.24477611940298508,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000171939727157048,
+      "loss": 0.6782,
+      "step": 5166
+    },
+    {
+      "epoch": 0.24482350153991944,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017192938198077936,
+      "loss": 1.1449,
+      "step": 5167
+    },
+    {
+      "epoch": 0.24487088367685383,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00017191903520919364,
+      "loss": 1.3613,
+      "step": 5168
+    },
+    {
+      "epoch": 0.2449182658137882,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001719086868425203,
+      "loss": 1.1635,
+      "step": 5169
+    },
+    {
+      "epoch": 0.24496564795072256,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001718983368809889,
+      "loss": 0.7599,
+      "step": 5170
+    },
+    {
+      "epoch": 0.24501303008765696,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00017188798532482896,
+      "loss": 0.0246,
+      "step": 5171
+    },
+    {
+      "epoch": 0.24506041222459132,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001718776321742701,
+      "loss": 1.2764,
+      "step": 5172
+    },
+    {
+      "epoch": 0.24510779436152572,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017186727742954188,
+      "loss": 0.1184,
+      "step": 5173
+    },
+    {
+      "epoch": 0.24515517649846008,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017185692109087403,
+      "loss": 0.9347,
+      "step": 5174
+    },
+    {
+      "epoch": 0.24520255863539445,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017184656315849618,
+      "loss": 0.4884,
+      "step": 5175
+    },
+    {
+      "epoch": 0.24524994077232884,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001718362036326381,
+      "loss": 0.1475,
+      "step": 5176
+    },
+    {
+      "epoch": 0.2452973229092632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017182584251352955,
+      "loss": 1.0809,
+      "step": 5177
+    },
+    {
+      "epoch": 0.24534470504619757,
+      "grad_norm": 0.0037689208984375,
+      "learning_rate": 0.00017181547980140032,
+      "loss": 0.0003,
+      "step": 5178
+    },
+    {
+      "epoch": 0.24539208718313196,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00017180511549648024,
+      "loss": 1.2188,
+      "step": 5179
+    },
+    {
+      "epoch": 0.24543946932006633,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00017179474959899918,
+      "loss": 0.0038,
+      "step": 5180
+    },
+    {
+      "epoch": 0.24548685145700072,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00017178438210918703,
+      "loss": 0.0334,
+      "step": 5181
+    },
+    {
+      "epoch": 0.2455342335939351,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00017177401302727376,
+      "loss": 0.0359,
+      "step": 5182
+    },
+    {
+      "epoch": 0.24558161573086945,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017176364235348932,
+      "loss": 1.4391,
+      "step": 5183
+    },
+    {
+      "epoch": 0.24562899786780384,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017175327008806375,
+      "loss": 0.6934,
+      "step": 5184
+    },
+    {
+      "epoch": 0.2456763800047382,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00017174289623122705,
+      "loss": 0.0347,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2457237621416726,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017173252078320935,
+      "loss": 1.4223,
+      "step": 5186
+    },
+    {
+      "epoch": 0.24577114427860697,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017172214374424076,
+      "loss": 0.9328,
+      "step": 5187
+    },
+    {
+      "epoch": 0.24581852641554133,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001717117651145514,
+      "loss": 0.0053,
+      "step": 5188
+    },
+    {
+      "epoch": 0.24586590855247573,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017170138489437146,
+      "loss": 0.4809,
+      "step": 5189
+    },
+    {
+      "epoch": 0.2459132906894101,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001716910030839312,
+      "loss": 0.1269,
+      "step": 5190
+    },
+    {
+      "epoch": 0.24596067282634446,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017168061968346083,
+      "loss": 0.6677,
+      "step": 5191
+    },
+    {
+      "epoch": 0.24600805496327885,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0001716702346931907,
+      "loss": 0.0111,
+      "step": 5192
+    },
+    {
+      "epoch": 0.24605543710021321,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017165984811335106,
+      "loss": 1.6476,
+      "step": 5193
+    },
+    {
+      "epoch": 0.2461028192371476,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00017164945994417233,
+      "loss": 1.1426,
+      "step": 5194
+    },
+    {
+      "epoch": 0.24615020137408197,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017163907018588492,
+      "loss": 0.6152,
+      "step": 5195
+    },
+    {
+      "epoch": 0.24619758351101634,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017162867883871924,
+      "loss": 0.9099,
+      "step": 5196
+    },
+    {
+      "epoch": 0.24624496564795073,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017161828590290572,
+      "loss": 0.0948,
+      "step": 5197
+    },
+    {
+      "epoch": 0.2462923477848851,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00017160789137867495,
+      "loss": 0.5108,
+      "step": 5198
+    },
+    {
+      "epoch": 0.24633972992181946,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001715974952662574,
+      "loss": 0.1785,
+      "step": 5199
+    },
+    {
+      "epoch": 0.24638711205875385,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00017158709756588366,
+      "loss": 0.0049,
+      "step": 5200
+    },
+    {
+      "epoch": 0.24643449419568822,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017157669827778436,
+      "loss": 0.2206,
+      "step": 5201
+    },
+    {
+      "epoch": 0.2464818763326226,
+      "grad_norm": 0.01275634765625,
+      "learning_rate": 0.0001715662974021901,
+      "loss": 0.0009,
+      "step": 5202
+    },
+    {
+      "epoch": 0.24652925846955698,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00017155589493933162,
+      "loss": 0.2739,
+      "step": 5203
+    },
+    {
+      "epoch": 0.24657664060649134,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001715454908894396,
+      "loss": 0.7662,
+      "step": 5204
+    },
+    {
+      "epoch": 0.24662402274342574,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001715350852527448,
+      "loss": 0.7587,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2466714048803601,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017152467802947804,
+      "loss": 1.1468,
+      "step": 5206
+    },
+    {
+      "epoch": 0.24671878701729447,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017151426921987008,
+      "loss": 0.1373,
+      "step": 5207
+    },
+    {
+      "epoch": 0.24676616915422886,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0001715038588241518,
+      "loss": 0.0061,
+      "step": 5208
+    },
+    {
+      "epoch": 0.24681355129116322,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001714934468425541,
+      "loss": 0.9072,
+      "step": 5209
+    },
+    {
+      "epoch": 0.24686093342809762,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017148303327530788,
+      "loss": 0.7494,
+      "step": 5210
+    },
+    {
+      "epoch": 0.24690831556503198,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017147261812264412,
+      "loss": 0.5658,
+      "step": 5211
+    },
+    {
+      "epoch": 0.24695569770196635,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017146220138479384,
+      "loss": 1.052,
+      "step": 5212
+    },
+    {
+      "epoch": 0.24700307983890074,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017145178306198806,
+      "loss": 0.1219,
+      "step": 5213
+    },
+    {
+      "epoch": 0.2470504619758351,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017144136315445783,
+      "loss": 0.7649,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2470978441127695,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017143094166243423,
+      "loss": 1.2744,
+      "step": 5215
+    },
+    {
+      "epoch": 0.24714522624970386,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00017142051858614848,
+      "loss": 0.0192,
+      "step": 5216
+    },
+    {
+      "epoch": 0.24719260838663823,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017141009392583167,
+      "loss": 1.2459,
+      "step": 5217
+    },
+    {
+      "epoch": 0.24723999052357262,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017139966768171504,
+      "loss": 0.5501,
+      "step": 5218
+    },
+    {
+      "epoch": 0.247287372660507,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00017138923985402985,
+      "loss": 0.2089,
+      "step": 5219
+    },
+    {
+      "epoch": 0.24733475479744135,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017137881044300735,
+      "loss": 0.794,
+      "step": 5220
+    },
+    {
+      "epoch": 0.24738213693437575,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017136837944887887,
+      "loss": 0.9825,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2474295190713101,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017135794687187574,
+      "loss": 0.9264,
+      "step": 5222
+    },
+    {
+      "epoch": 0.2474769012082445,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017134751271222936,
+      "loss": 1.1857,
+      "step": 5223
+    },
+    {
+      "epoch": 0.24752428334517887,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017133707697017115,
+      "loss": 0.0222,
+      "step": 5224
+    },
+    {
+      "epoch": 0.24757166548211323,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017132663964593254,
+      "loss": 0.0491,
+      "step": 5225
+    },
+    {
+      "epoch": 0.24761904761904763,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017131620073974503,
+      "loss": 1.0718,
+      "step": 5226
+    },
+    {
+      "epoch": 0.247666429755982,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001713057602518402,
+      "loss": 0.4596,
+      "step": 5227
+    },
+    {
+      "epoch": 0.24771381189291636,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017129531818244954,
+      "loss": 0.0546,
+      "step": 5228
+    },
+    {
+      "epoch": 0.24776119402985075,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00017128487453180462,
+      "loss": 0.8905,
+      "step": 5229
+    },
+    {
+      "epoch": 0.24780857616678512,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017127442930013715,
+      "loss": 0.538,
+      "step": 5230
+    },
+    {
+      "epoch": 0.2478559583037195,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017126398248767875,
+      "loss": 0.144,
+      "step": 5231
+    },
+    {
+      "epoch": 0.24790334044065387,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001712535340946611,
+      "loss": 1.2883,
+      "step": 5232
+    },
+    {
+      "epoch": 0.24795072257758824,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.000171243084121316,
+      "loss": 1.1133,
+      "step": 5233
+    },
+    {
+      "epoch": 0.24799810471452263,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00017123263256787517,
+      "loss": 0.0404,
+      "step": 5234
+    },
+    {
+      "epoch": 0.248045486851457,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001712221794345704,
+      "loss": 0.1245,
+      "step": 5235
+    },
+    {
+      "epoch": 0.24809286898839136,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00017121172472163356,
+      "loss": 0.042,
+      "step": 5236
+    },
+    {
+      "epoch": 0.24814025112532576,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00017120126842929656,
+      "loss": 0.4715,
+      "step": 5237
+    },
+    {
+      "epoch": 0.24818763326226012,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001711908105577912,
+      "loss": 1.1695,
+      "step": 5238
+    },
+    {
+      "epoch": 0.24823501539919451,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00017118035110734954,
+      "loss": 0.9905,
+      "step": 5239
+    },
+    {
+      "epoch": 0.24828239753612888,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0001711698900782035,
+      "loss": 0.0019,
+      "step": 5240
+    },
+    {
+      "epoch": 0.24832977967306324,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001711594274705851,
+      "loss": 0.9222,
+      "step": 5241
+    },
+    {
+      "epoch": 0.24837716180999764,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00017114896328472638,
+      "loss": 0.2917,
+      "step": 5242
+    },
+    {
+      "epoch": 0.248424543946932,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017113849752085946,
+      "loss": 0.5447,
+      "step": 5243
+    },
+    {
+      "epoch": 0.2484719260838664,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001711280301792164,
+      "loss": 1.1684,
+      "step": 5244
+    },
+    {
+      "epoch": 0.24851930822080076,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00017111756126002945,
+      "loss": 1.2022,
+      "step": 5245
+    },
+    {
+      "epoch": 0.24856669035773513,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017110709076353068,
+      "loss": 0.9585,
+      "step": 5246
+    },
+    {
+      "epoch": 0.24861407249466952,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017109661868995244,
+      "loss": 0.932,
+      "step": 5247
+    },
+    {
+      "epoch": 0.24866145463160388,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001710861450395269,
+      "loss": 1.2129,
+      "step": 5248
+    },
+    {
+      "epoch": 0.24870883676853825,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017107566981248637,
+      "loss": 0.0886,
+      "step": 5249
+    },
+    {
+      "epoch": 0.24875621890547264,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001710651930090632,
+      "loss": 0.0481,
+      "step": 5250
+    },
+    {
+      "epoch": 0.248803601042407,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017105471462948975,
+      "loss": 0.1141,
+      "step": 5251
+    },
+    {
+      "epoch": 0.2488509831793414,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00017104423467399838,
+      "loss": 0.9654,
+      "step": 5252
+    },
+    {
+      "epoch": 0.24889836531627577,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001710337531428216,
+      "loss": 0.7203,
+      "step": 5253
+    },
+    {
+      "epoch": 0.24894574745321013,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017102327003619183,
+      "loss": 1.1123,
+      "step": 5254
+    },
+    {
+      "epoch": 0.24899312959014452,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017101278535434155,
+      "loss": 0.857,
+      "step": 5255
+    },
+    {
+      "epoch": 0.2490405117270789,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017100229909750337,
+      "loss": 1.2906,
+      "step": 5256
+    },
+    {
+      "epoch": 0.24908789386401325,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001709918112659098,
+      "loss": 0.7463,
+      "step": 5257
+    },
+    {
+      "epoch": 0.24913527600094765,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 0.00017098132185979346,
+      "loss": 0.0013,
+      "step": 5258
+    },
+    {
+      "epoch": 0.249182658137882,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017097083087938705,
+      "loss": 1.2393,
+      "step": 5259
+    },
+    {
+      "epoch": 0.2492300402748164,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017096033832492317,
+      "loss": 0.1059,
+      "step": 5260
+    },
+    {
+      "epoch": 0.24927742241175077,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017094984419663457,
+      "loss": 1.0911,
+      "step": 5261
+    },
+    {
+      "epoch": 0.24932480454868514,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017093934849475405,
+      "loss": 0.1458,
+      "step": 5262
+    },
+    {
+      "epoch": 0.24937218668561953,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017092885121951427,
+      "loss": 0.7728,
+      "step": 5263
+    },
+    {
+      "epoch": 0.2494195688225539,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017091835237114818,
+      "loss": 1.1178,
+      "step": 5264
+    },
+    {
+      "epoch": 0.24946695095948826,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017090785194988852,
+      "loss": 0.9226,
+      "step": 5265
+    },
+    {
+      "epoch": 0.24951433309642265,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017089734995596824,
+      "loss": 1.1267,
+      "step": 5266
+    },
+    {
+      "epoch": 0.24956171523335702,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017088684638962029,
+      "loss": 0.9419,
+      "step": 5267
+    },
+    {
+      "epoch": 0.2496090973702914,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00017087634125107756,
+      "loss": 0.5949,
+      "step": 5268
+    },
+    {
+      "epoch": 0.24965647950722578,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001708658345405731,
+      "loss": 0.0654,
+      "step": 5269
+    },
+    {
+      "epoch": 0.24970386164416014,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001708553262583399,
+      "loss": 1.1039,
+      "step": 5270
+    },
+    {
+      "epoch": 0.24975124378109453,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017084481640461104,
+      "loss": 0.5722,
+      "step": 5271
+    },
+    {
+      "epoch": 0.2497986259180289,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001708343049796196,
+      "loss": 1.0248,
+      "step": 5272
+    },
+    {
+      "epoch": 0.2498460080549633,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00017082379198359875,
+      "loss": 0.7751,
+      "step": 5273
+    },
+    {
+      "epoch": 0.24989339019189766,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00017081327741678162,
+      "loss": 0.1469,
+      "step": 5274
+    },
+    {
+      "epoch": 0.24994077232883202,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001708027612794014,
+      "loss": 0.9905,
+      "step": 5275
+    },
+    {
+      "epoch": 0.24998815446576642,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017079224357169137,
+      "loss": 0.7281,
+      "step": 5276
+    },
+    {
+      "epoch": 0.2500355366027008,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017078172429388477,
+      "loss": 0.9772,
+      "step": 5277
+    },
+    {
+      "epoch": 0.2500829187396352,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00017077120344621496,
+      "loss": 0.0234,
+      "step": 5278
+    },
+    {
+      "epoch": 0.2501303008765695,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 0.0001707606810289152,
+      "loss": 0.001,
+      "step": 5279
+    },
+    {
+      "epoch": 0.2501776830135039,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001707501570422189,
+      "loss": 1.0387,
+      "step": 5280
+    },
+    {
+      "epoch": 0.2502250651504383,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001707396314863595,
+      "loss": 0.0456,
+      "step": 5281
+    },
+    {
+      "epoch": 0.25027244728737263,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001707291043615704,
+      "loss": 0.9844,
+      "step": 5282
+    },
+    {
+      "epoch": 0.250319829424307,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00017071857566808513,
+      "loss": 0.1872,
+      "step": 5283
+    },
+    {
+      "epoch": 0.2503672115612414,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017070804540613718,
+      "loss": 0.1904,
+      "step": 5284
+    },
+    {
+      "epoch": 0.2504145936981758,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001706975135759601,
+      "loss": 1.0268,
+      "step": 5285
+    },
+    {
+      "epoch": 0.25046197583511015,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001706869801777874,
+      "loss": 1.2751,
+      "step": 5286
+    },
+    {
+      "epoch": 0.25050935797204454,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00017067644521185288,
+      "loss": 1.1897,
+      "step": 5287
+    },
+    {
+      "epoch": 0.25055674010897894,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017066590867838999,
+      "loss": 1.3925,
+      "step": 5288
+    },
+    {
+      "epoch": 0.2506041222459133,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00017065537057763257,
+      "loss": 0.1312,
+      "step": 5289
+    },
+    {
+      "epoch": 0.25065150438284767,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00017064483090981428,
+      "loss": 0.0314,
+      "step": 5290
+    },
+    {
+      "epoch": 0.25069888651978206,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017063428967516888,
+      "loss": 1.44,
+      "step": 5291
+    },
+    {
+      "epoch": 0.2507462686567164,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001706237468739302,
+      "loss": 0.7273,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2507936507936508,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000170613202506332,
+      "loss": 0.0644,
+      "step": 5293
+    },
+    {
+      "epoch": 0.2508410329305852,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017060265657260822,
+      "loss": 1.2131,
+      "step": 5294
+    },
+    {
+      "epoch": 0.2508884150675195,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00017059210907299267,
+      "loss": 0.408,
+      "step": 5295
+    },
+    {
+      "epoch": 0.2509357972044539,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017058156000771937,
+      "loss": 1.3831,
+      "step": 5296
+    },
+    {
+      "epoch": 0.2509831793413883,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00017057100937702222,
+      "loss": 0.134,
+      "step": 5297
+    },
+    {
+      "epoch": 0.25103056147832264,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017056045718113528,
+      "loss": 1.0522,
+      "step": 5298
+    },
+    {
+      "epoch": 0.25107794361525704,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00017054990342029255,
+      "loss": 1.4229,
+      "step": 5299
+    },
+    {
+      "epoch": 0.25112532575219143,
+      "grad_norm": 0.006072998046875,
+      "learning_rate": 0.0001705393480947281,
+      "loss": 0.0003,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2511727078891258,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017052879120467605,
+      "loss": 0.0204,
+      "step": 5301
+    },
+    {
+      "epoch": 0.25122009002606016,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017051823275037053,
+      "loss": 1.1679,
+      "step": 5302
+    },
+    {
+      "epoch": 0.25126747216299455,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00017050767273204574,
+      "loss": 0.9325,
+      "step": 5303
+    },
+    {
+      "epoch": 0.25131485429992895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017049711114993588,
+      "loss": 0.5116,
+      "step": 5304
+    },
+    {
+      "epoch": 0.2513622364368633,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00017048654800427512,
+      "loss": 0.0021,
+      "step": 5305
+    },
+    {
+      "epoch": 0.2514096185737977,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00017047598329529787,
+      "loss": 0.0169,
+      "step": 5306
+    },
+    {
+      "epoch": 0.25145700071073207,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00017046541702323836,
+      "loss": 0.0057,
+      "step": 5307
+    },
+    {
+      "epoch": 0.2515043828476664,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017045484918833093,
+      "loss": 1.2136,
+      "step": 5308
+    },
+    {
+      "epoch": 0.2515517649846008,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017044427979081002,
+      "loss": 0.206,
+      "step": 5309
+    },
+    {
+      "epoch": 0.2515991471215352,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017043370883091,
+      "loss": 0.695,
+      "step": 5310
+    },
+    {
+      "epoch": 0.25164652925846953,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00017042313630886535,
+      "loss": 0.1355,
+      "step": 5311
+    },
+    {
+      "epoch": 0.2516939113954039,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017041256222491056,
+      "loss": 0.4462,
+      "step": 5312
+    },
+    {
+      "epoch": 0.2517412935323383,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017040198657928011,
+      "loss": 0.1635,
+      "step": 5313
+    },
+    {
+      "epoch": 0.2517886756692727,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017039140937220862,
+      "loss": 1.4779,
+      "step": 5314
+    },
+    {
+      "epoch": 0.25183605780620705,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001703808306039306,
+      "loss": 1.2994,
+      "step": 5315
+    },
+    {
+      "epoch": 0.25188343994314144,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001703702502746808,
+      "loss": 1.1394,
+      "step": 5316
+    },
+    {
+      "epoch": 0.25193082208007583,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00017035966838469376,
+      "loss": 0.9651,
+      "step": 5317
+    },
+    {
+      "epoch": 0.25197820421701017,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001703490849342042,
+      "loss": 1.0763,
+      "step": 5318
+    },
+    {
+      "epoch": 0.25202558635394456,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017033849992344687,
+      "loss": 0.9423,
+      "step": 5319
+    },
+    {
+      "epoch": 0.25207296849087896,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017032791335265657,
+      "loss": 0.0622,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2521203506278133,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00017031732522206804,
+      "loss": 0.1328,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2521677327647477,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017030673553191611,
+      "loss": 0.5537,
+      "step": 5322
+    },
+    {
+      "epoch": 0.2522151149016821,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001702961442824357,
+      "loss": 1.0053,
+      "step": 5323
+    },
+    {
+      "epoch": 0.2522624970386164,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017028555147386172,
+      "loss": 0.1488,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2523098791755508,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.000170274957106429,
+      "loss": 0.6858,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2523572613124852,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017026436118037266,
+      "loss": 1.3075,
+      "step": 5326
+    },
+    {
+      "epoch": 0.25240464344941954,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00017025376369592758,
+      "loss": 1.209,
+      "step": 5327
+    },
+    {
+      "epoch": 0.25245202558635393,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017024316465332886,
+      "loss": 0.5749,
+      "step": 5328
+    },
+    {
+      "epoch": 0.2524994077232883,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017023256405281157,
+      "loss": 0.128,
+      "step": 5329
+    },
+    {
+      "epoch": 0.2525467898602227,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001702219618946108,
+      "loss": 0.0648,
+      "step": 5330
+    },
+    {
+      "epoch": 0.25259417199715706,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017021135817896178,
+      "loss": 0.0228,
+      "step": 5331
+    },
+    {
+      "epoch": 0.25264155413409145,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017020075290609957,
+      "loss": 1.1479,
+      "step": 5332
+    },
+    {
+      "epoch": 0.25268893627102584,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00017019014607625943,
+      "loss": 1.4094,
+      "step": 5333
+    },
+    {
+      "epoch": 0.2527363184079602,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017017953768967662,
+      "loss": 1.2996,
+      "step": 5334
+    },
+    {
+      "epoch": 0.2527837005448946,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017016892774658642,
+      "loss": 0.0563,
+      "step": 5335
+    },
+    {
+      "epoch": 0.25283108268182897,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00017015831624722413,
+      "loss": 0.0816,
+      "step": 5336
+    },
+    {
+      "epoch": 0.2528784648187633,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017014770319182513,
+      "loss": 0.0362,
+      "step": 5337
+    },
+    {
+      "epoch": 0.2529258469556977,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001701370885806248,
+      "loss": 0.7215,
+      "step": 5338
+    },
+    {
+      "epoch": 0.2529732290926321,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017012647241385856,
+      "loss": 0.0291,
+      "step": 5339
+    },
+    {
+      "epoch": 0.25302061122956643,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017011585469176184,
+      "loss": 1.0139,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2530679933665008,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00017010523541457015,
+      "loss": 0.8763,
+      "step": 5341
+    },
+    {
+      "epoch": 0.2531153755034352,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.000170094614582519,
+      "loss": 1.1539,
+      "step": 5342
+    },
+    {
+      "epoch": 0.2531627576403696,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00017008399219584398,
+      "loss": 0.7403,
+      "step": 5343
+    },
+    {
+      "epoch": 0.25321013977730394,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017007336825478064,
+      "loss": 0.9521,
+      "step": 5344
+    },
+    {
+      "epoch": 0.25325752191423834,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00017006274275956461,
+      "loss": 0.0444,
+      "step": 5345
+    },
+    {
+      "epoch": 0.25330490405117273,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001700521157104316,
+      "loss": 0.6661,
+      "step": 5346
+    },
+    {
+      "epoch": 0.25335228618810707,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017004148710761732,
+      "loss": 1.1629,
+      "step": 5347
+    },
+    {
+      "epoch": 0.25339966832504146,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017003085695135742,
+      "loss": 1.4518,
+      "step": 5348
+    },
+    {
+      "epoch": 0.25344705046197585,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001700202252418877,
+      "loss": 1.1735,
+      "step": 5349
+    },
+    {
+      "epoch": 0.2534944325989102,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000170009591979444,
+      "loss": 1.2322,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2535418147358446,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016999895716426208,
+      "loss": 0.4126,
+      "step": 5351
+    },
+    {
+      "epoch": 0.253589196872779,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016998832079657787,
+      "loss": 0.0597,
+      "step": 5352
+    },
+    {
+      "epoch": 0.2536365790097133,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016997768287662724,
+      "loss": 0.3836,
+      "step": 5353
+    },
+    {
+      "epoch": 0.2536839611466477,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016996704340464612,
+      "loss": 1.1632,
+      "step": 5354
+    },
+    {
+      "epoch": 0.2537313432835821,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001699564023808706,
+      "loss": 1.2416,
+      "step": 5355
+    },
+    {
+      "epoch": 0.25377872542051644,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001699457598055365,
+      "loss": 0.913,
+      "step": 5356
+    },
+    {
+      "epoch": 0.25382610755745083,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016993511567887996,
+      "loss": 1.0357,
+      "step": 5357
+    },
+    {
+      "epoch": 0.2538734896943852,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016992447000113706,
+      "loss": 0.0206,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2539208718313196,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016991382277254391,
+      "loss": 0.8985,
+      "step": 5359
+    },
+    {
+      "epoch": 0.25396825396825395,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001699031739933366,
+      "loss": 1.1099,
+      "step": 5360
+    },
+    {
+      "epoch": 0.25401563610518835,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016989252366375138,
+      "loss": 0.32,
+      "step": 5361
+    },
+    {
+      "epoch": 0.25406301824212274,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016988187178402443,
+      "loss": 0.3791,
+      "step": 5362
+    },
+    {
+      "epoch": 0.2541104003790571,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016987121835439198,
+      "loss": 1.2845,
+      "step": 5363
+    },
+    {
+      "epoch": 0.25415778251599147,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016986056337509034,
+      "loss": 0.893,
+      "step": 5364
+    },
+    {
+      "epoch": 0.25420516465292586,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016984990684635584,
+      "loss": 0.888,
+      "step": 5365
+    },
+    {
+      "epoch": 0.2542525467898602,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016983924876842478,
+      "loss": 1.352,
+      "step": 5366
+    },
+    {
+      "epoch": 0.2542999289267946,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016982858914153356,
+      "loss": 0.3745,
+      "step": 5367
+    },
+    {
+      "epoch": 0.254347311063729,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016981792796591866,
+      "loss": 1.3169,
+      "step": 5368
+    },
+    {
+      "epoch": 0.2543946932006633,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016980726524181642,
+      "loss": 1.0266,
+      "step": 5369
+    },
+    {
+      "epoch": 0.2544420753375977,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016979660096946343,
+      "loss": 0.8766,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2544894574745321,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001697859351490962,
+      "loss": 1.0002,
+      "step": 5371
+    },
+    {
+      "epoch": 0.2545368396114665,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016977526778095121,
+      "loss": 0.0449,
+      "step": 5372
+    },
+    {
+      "epoch": 0.25458422174840084,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016976459886526514,
+      "loss": 1.2232,
+      "step": 5373
+    },
+    {
+      "epoch": 0.25463160388533523,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016975392840227455,
+      "loss": 0.6692,
+      "step": 5374
+    },
+    {
+      "epoch": 0.2546789860222696,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016974325639221616,
+      "loss": 1.3018,
+      "step": 5375
+    },
+    {
+      "epoch": 0.25472636815920396,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016973258283532657,
+      "loss": 0.1251,
+      "step": 5376
+    },
+    {
+      "epoch": 0.25477375029613836,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001697219077318426,
+      "loss": 0.9578,
+      "step": 5377
+    },
+    {
+      "epoch": 0.25482113243307275,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016971123108200102,
+      "loss": 0.7342,
+      "step": 5378
+    },
+    {
+      "epoch": 0.2548685145700071,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016970055288603858,
+      "loss": 1.1069,
+      "step": 5379
+    },
+    {
+      "epoch": 0.2549158967069415,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016968987314419203,
+      "loss": 0.0442,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2549632788438759,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016967919185669842,
+      "loss": 1.1472,
+      "step": 5381
+    },
+    {
+      "epoch": 0.2550106609808102,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001696685090237945,
+      "loss": 0.598,
+      "step": 5382
+    },
+    {
+      "epoch": 0.2550580431177446,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016965782464571728,
+      "loss": 0.3313,
+      "step": 5383
+    },
+    {
+      "epoch": 0.255105425254679,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001696471387227037,
+      "loss": 0.1466,
+      "step": 5384
+    },
+    {
+      "epoch": 0.25515280739161333,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016963645125499079,
+      "loss": 0.8998,
+      "step": 5385
+    },
+    {
+      "epoch": 0.2552001895285477,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016962576224281554,
+      "loss": 0.1095,
+      "step": 5386
+    },
+    {
+      "epoch": 0.2552475716654821,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00016961507168641507,
+      "loss": 0.0474,
+      "step": 5387
+    },
+    {
+      "epoch": 0.2552949538024165,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016960437958602644,
+      "loss": 0.7683,
+      "step": 5388
+    },
+    {
+      "epoch": 0.25534233593935085,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016959368594188681,
+      "loss": 1.4062,
+      "step": 5389
+    },
+    {
+      "epoch": 0.25538971807628524,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016958299075423336,
+      "loss": 0.9186,
+      "step": 5390
+    },
+    {
+      "epoch": 0.25543710021321964,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00016957229402330329,
+      "loss": 0.0289,
+      "step": 5391
+    },
+    {
+      "epoch": 0.255484482350154,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0001695615957493338,
+      "loss": 0.0273,
+      "step": 5392
+    },
+    {
+      "epoch": 0.25553186448708837,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016955089593256227,
+      "loss": 0.5225,
+      "step": 5393
+    },
+    {
+      "epoch": 0.25557924662402276,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016954019457322595,
+      "loss": 1.5962,
+      "step": 5394
+    },
+    {
+      "epoch": 0.2556266287609571,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016952949167156216,
+      "loss": 0.8025,
+      "step": 5395
+    },
+    {
+      "epoch": 0.2556740108978915,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016951878722780832,
+      "loss": 0.8912,
+      "step": 5396
+    },
+    {
+      "epoch": 0.2557213930348259,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001695080812422018,
+      "loss": 0.4736,
+      "step": 5397
+    },
+    {
+      "epoch": 0.2557687751717602,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016949737371498008,
+      "loss": 0.803,
+      "step": 5398
+    },
+    {
+      "epoch": 0.2558161573086946,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016948666464638068,
+      "loss": 1.0973,
+      "step": 5399
+    },
+    {
+      "epoch": 0.255863539445629,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016947595403664101,
+      "loss": 0.0214,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2559109215825634,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00016946524188599872,
+      "loss": 1.081,
+      "step": 5401
+    },
+    {
+      "epoch": 0.25595830371949774,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016945452819469136,
+      "loss": 0.5477,
+      "step": 5402
+    },
+    {
+      "epoch": 0.25600568585643213,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016944381296295653,
+      "loss": 0.1643,
+      "step": 5403
+    },
+    {
+      "epoch": 0.2560530679933665,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016943309619103192,
+      "loss": 0.2134,
+      "step": 5404
+    },
+    {
+      "epoch": 0.25610045013030086,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00016942237787915515,
+      "loss": 0.1236,
+      "step": 5405
+    },
+    {
+      "epoch": 0.25614783226723525,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016941165802756403,
+      "loss": 0.0374,
+      "step": 5406
+    },
+    {
+      "epoch": 0.25619521440416965,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016940093663649622,
+      "loss": 0.8897,
+      "step": 5407
+    },
+    {
+      "epoch": 0.256242596541104,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001693902137061896,
+      "loss": 1.0275,
+      "step": 5408
+    },
+    {
+      "epoch": 0.2562899786780384,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00016937948923688193,
+      "loss": 0.089,
+      "step": 5409
+    },
+    {
+      "epoch": 0.25633736081497277,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001693687632288111,
+      "loss": 1.2504,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2563847429519071,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000169358035682215,
+      "loss": 0.1348,
+      "step": 5411
+    },
+    {
+      "epoch": 0.2564321250888415,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00016934730659733154,
+      "loss": 0.1577,
+      "step": 5412
+    },
+    {
+      "epoch": 0.2564795072257759,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016933657597439865,
+      "loss": 1.201,
+      "step": 5413
+    },
+    {
+      "epoch": 0.25652688936271023,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016932584381365438,
+      "loss": 0.5596,
+      "step": 5414
+    },
+    {
+      "epoch": 0.2565742714996446,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00016931511011533673,
+      "loss": 0.4878,
+      "step": 5415
+    },
+    {
+      "epoch": 0.256621653636579,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016930437487968378,
+      "loss": 0.0254,
+      "step": 5416
+    },
+    {
+      "epoch": 0.2566690357735134,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001692936381069336,
+      "loss": 0.7329,
+      "step": 5417
+    },
+    {
+      "epoch": 0.25671641791044775,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00016928289979732436,
+      "loss": 0.0926,
+      "step": 5418
+    },
+    {
+      "epoch": 0.25676380004738214,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001692721599510942,
+      "loss": 0.4945,
+      "step": 5419
+    },
+    {
+      "epoch": 0.25681118218431653,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001692614185684813,
+      "loss": 0.0455,
+      "step": 5420
+    },
+    {
+      "epoch": 0.25685856432125087,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016925067564972393,
+      "loss": 1.5666,
+      "step": 5421
+    },
+    {
+      "epoch": 0.25690594645818526,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001692399311950603,
+      "loss": 0.1474,
+      "step": 5422
+    },
+    {
+      "epoch": 0.25695332859511966,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001692291852047288,
+      "loss": 1.1678,
+      "step": 5423
+    },
+    {
+      "epoch": 0.257000710732054,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016921843767896765,
+      "loss": 0.8908,
+      "step": 5424
+    },
+    {
+      "epoch": 0.2570480928689884,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016920768861801533,
+      "loss": 0.6911,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2570954750059228,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001691969380221102,
+      "loss": 0.1496,
+      "step": 5426
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016918618589149064,
+      "loss": 0.8787,
+      "step": 5427
+    },
+    {
+      "epoch": 0.2571902392797915,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001691754322263952,
+      "loss": 1.2739,
+      "step": 5428
+    },
+    {
+      "epoch": 0.2572376214167259,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016916467702706236,
+      "loss": 0.8478,
+      "step": 5429
+    },
+    {
+      "epoch": 0.2572850035536603,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016915392029373064,
+      "loss": 0.8881,
+      "step": 5430
+    },
+    {
+      "epoch": 0.25733238569059463,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016914316202663862,
+      "loss": 1.5181,
+      "step": 5431
+    },
+    {
+      "epoch": 0.257379767827529,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016913240222602493,
+      "loss": 0.5271,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2574271499644634,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001691216408921282,
+      "loss": 0.6417,
+      "step": 5433
+    },
+    {
+      "epoch": 0.25747453210139776,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0001691108780251871,
+      "loss": 0.1801,
+      "step": 5434
+    },
+    {
+      "epoch": 0.25752191423833215,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016910011362544035,
+      "loss": 0.7382,
+      "step": 5435
+    },
+    {
+      "epoch": 0.25756929637526654,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016908934769312666,
+      "loss": 1.7052,
+      "step": 5436
+    },
+    {
+      "epoch": 0.2576166785122009,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00016907858022848483,
+      "loss": 1.3022,
+      "step": 5437
+    },
+    {
+      "epoch": 0.2576640606491353,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016906781123175366,
+      "loss": 0.5217,
+      "step": 5438
+    },
+    {
+      "epoch": 0.25771144278606967,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000169057040703172,
+      "loss": 0.6668,
+      "step": 5439
+    },
+    {
+      "epoch": 0.257758824923004,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016904626864297875,
+      "loss": 1.2352,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2578062070599384,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00016903549505141284,
+      "loss": 0.9384,
+      "step": 5441
+    },
+    {
+      "epoch": 0.2578535891968728,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016902471992871315,
+      "loss": 1.3833,
+      "step": 5442
+    },
+    {
+      "epoch": 0.2579009713338071,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001690139432751187,
+      "loss": 0.7121,
+      "step": 5443
+    },
+    {
+      "epoch": 0.2579483534707415,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016900316509086847,
+      "loss": 1.2244,
+      "step": 5444
+    },
+    {
+      "epoch": 0.2579957356076759,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00016899238537620154,
+      "loss": 0.0141,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2580431177446103,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016898160413135701,
+      "loss": 1.0926,
+      "step": 5446
+    },
+    {
+      "epoch": 0.25809049988154464,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016897082135657399,
+      "loss": 1.0492,
+      "step": 5447
+    },
+    {
+      "epoch": 0.25813788201847904,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00016896003705209157,
+      "loss": 0.1845,
+      "step": 5448
+    },
+    {
+      "epoch": 0.25818526415541343,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016894925121814906,
+      "loss": 1.3174,
+      "step": 5449
+    },
+    {
+      "epoch": 0.25823264629234777,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016893846385498552,
+      "loss": 0.0335,
+      "step": 5450
+    },
+    {
+      "epoch": 0.25828002842928216,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016892767496284034,
+      "loss": 1.2171,
+      "step": 5451
+    },
+    {
+      "epoch": 0.25832741056621655,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016891688454195273,
+      "loss": 1.2409,
+      "step": 5452
+    },
+    {
+      "epoch": 0.2583747927031509,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000168906092592562,
+      "loss": 0.5802,
+      "step": 5453
+    },
+    {
+      "epoch": 0.2584221748400853,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016889529911490753,
+      "loss": 0.9941,
+      "step": 5454
+    },
+    {
+      "epoch": 0.2584695569770197,
+      "grad_norm": 0.002838134765625,
+      "learning_rate": 0.00016888450410922876,
+      "loss": 0.0003,
+      "step": 5455
+    },
+    {
+      "epoch": 0.258516939113954,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.000168873707575765,
+      "loss": 0.2259,
+      "step": 5456
+    },
+    {
+      "epoch": 0.2585643212508884,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016886290951475584,
+      "loss": 0.9563,
+      "step": 5457
+    },
+    {
+      "epoch": 0.2586117033878228,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016885210992644066,
+      "loss": 0.0681,
+      "step": 5458
+    },
+    {
+      "epoch": 0.2586590855247572,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000168841308811059,
+      "loss": 1.1659,
+      "step": 5459
+    },
+    {
+      "epoch": 0.25870646766169153,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016883050616885043,
+      "loss": 1.6042,
+      "step": 5460
+    },
+    {
+      "epoch": 0.2587538497986259,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001688197020000546,
+      "loss": 0.6201,
+      "step": 5461
+    },
+    {
+      "epoch": 0.2588012319355603,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016880889630491104,
+      "loss": 1.339,
+      "step": 5462
+    },
+    {
+      "epoch": 0.25884861407249465,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00016879808908365945,
+      "loss": 1.2221,
+      "step": 5463
+    },
+    {
+      "epoch": 0.25889599620942905,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016878728033653957,
+      "loss": 1.1975,
+      "step": 5464
+    },
+    {
+      "epoch": 0.25894337834636344,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016877647006379104,
+      "loss": 0.8915,
+      "step": 5465
+    },
+    {
+      "epoch": 0.2589907604832978,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016876565826565366,
+      "loss": 0.2214,
+      "step": 5466
+    },
+    {
+      "epoch": 0.25903814262023217,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016875484494236726,
+      "loss": 0.0409,
+      "step": 5467
+    },
+    {
+      "epoch": 0.25908552475716656,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016874403009417162,
+      "loss": 1.0784,
+      "step": 5468
+    },
+    {
+      "epoch": 0.2591329068941009,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001687332137213066,
+      "loss": 0.0318,
+      "step": 5469
+    },
+    {
+      "epoch": 0.2591802890310353,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001687223958240121,
+      "loss": 1.1415,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2592276711679697,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016871157640252807,
+      "loss": 0.7716,
+      "step": 5471
+    },
+    {
+      "epoch": 0.259275053304904,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00016870075545709449,
+      "loss": 0.0613,
+      "step": 5472
+    },
+    {
+      "epoch": 0.2593224354418384,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001686899329879513,
+      "loss": 0.0502,
+      "step": 5473
+    },
+    {
+      "epoch": 0.2593698175787728,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00016867910899533858,
+      "loss": 0.0127,
+      "step": 5474
+    },
+    {
+      "epoch": 0.2594171997157072,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016866828347949638,
+      "loss": 0.7669,
+      "step": 5475
+    },
+    {
+      "epoch": 0.25946458185264154,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001686574564406648,
+      "loss": 0.7168,
+      "step": 5476
+    },
+    {
+      "epoch": 0.25951196398957593,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016864662787908393,
+      "loss": 0.9622,
+      "step": 5477
+    },
+    {
+      "epoch": 0.2595593461265103,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016863579779499398,
+      "loss": 1.2405,
+      "step": 5478
+    },
+    {
+      "epoch": 0.25960672826344466,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016862496618863518,
+      "loss": 0.9796,
+      "step": 5479
+    },
+    {
+      "epoch": 0.25965411040037906,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001686141330602477,
+      "loss": 1.1535,
+      "step": 5480
+    },
+    {
+      "epoch": 0.25970149253731345,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016860329841007182,
+      "loss": 1.0441,
+      "step": 5481
+    },
+    {
+      "epoch": 0.2597488746742478,
+      "grad_norm": 0.030029296875,
+      "learning_rate": 0.0001685924622383479,
+      "loss": 0.0022,
+      "step": 5482
+    },
+    {
+      "epoch": 0.2597962568111822,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016858162454531618,
+      "loss": 1.1805,
+      "step": 5483
+    },
+    {
+      "epoch": 0.2598436389481166,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0001685707853312171,
+      "loss": 0.0136,
+      "step": 5484
+    },
+    {
+      "epoch": 0.2598910210850509,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016855994459629106,
+      "loss": 0.9291,
+      "step": 5485
+    },
+    {
+      "epoch": 0.2599384032219853,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00016854910234077842,
+      "loss": 1.4879,
+      "step": 5486
+    },
+    {
+      "epoch": 0.2599857853589197,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016853825856491972,
+      "loss": 1.1442,
+      "step": 5487
+    },
+    {
+      "epoch": 0.2600331674958541,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001685274132689555,
+      "loss": 0.1475,
+      "step": 5488
+    },
+    {
+      "epoch": 0.2600805496327884,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016851656645312622,
+      "loss": 1.3088,
+      "step": 5489
+    },
+    {
+      "epoch": 0.2601279317697228,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016850571811767245,
+      "loss": 0.2319,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2601753139066572,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016849486826283486,
+      "loss": 1.084,
+      "step": 5491
+    },
+    {
+      "epoch": 0.26022269604359155,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.000168484016888854,
+      "loss": 1.3469,
+      "step": 5492
+    },
+    {
+      "epoch": 0.26027007818052594,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016847316399597065,
+      "loss": 1.4763,
+      "step": 5493
+    },
+    {
+      "epoch": 0.26031746031746034,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016846230958442542,
+      "loss": 0.6448,
+      "step": 5494
+    },
+    {
+      "epoch": 0.2603648424543947,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001684514536544591,
+      "loss": 1.0415,
+      "step": 5495
+    },
+    {
+      "epoch": 0.26041222459132907,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016844059620631244,
+      "loss": 1.5532,
+      "step": 5496
+    },
+    {
+      "epoch": 0.26045960672826346,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016842973724022625,
+      "loss": 0.0721,
+      "step": 5497
+    },
+    {
+      "epoch": 0.2605069888651978,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001684188767564414,
+      "loss": 0.976,
+      "step": 5498
+    },
+    {
+      "epoch": 0.2605543710021322,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0001684080147551987,
+      "loss": 0.0077,
+      "step": 5499
+    },
+    {
+      "epoch": 0.2606017531390666,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016839715123673915,
+      "loss": 1.5189,
+      "step": 5500
+    },
+    {
+      "epoch": 0.2606491352760009,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016838628620130362,
+      "loss": 0.9867,
+      "step": 5501
+    },
+    {
+      "epoch": 0.2606965174129353,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00016837541964913306,
+      "loss": 0.4231,
+      "step": 5502
+    },
+    {
+      "epoch": 0.2607438995498697,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016836455158046856,
+      "loss": 1.5931,
+      "step": 5503
+    },
+    {
+      "epoch": 0.2607912816868041,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016835368199555112,
+      "loss": 1.2388,
+      "step": 5504
+    },
+    {
+      "epoch": 0.26083866382373844,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00016834281089462186,
+      "loss": 0.0145,
+      "step": 5505
+    },
+    {
+      "epoch": 0.26088604596067283,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016833193827792177,
+      "loss": 0.621,
+      "step": 5506
+    },
+    {
+      "epoch": 0.2609334280976072,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016832106414569213,
+      "loss": 0.7694,
+      "step": 5507
+    },
+    {
+      "epoch": 0.26098081023454156,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016831018849817407,
+      "loss": 0.7878,
+      "step": 5508
+    },
+    {
+      "epoch": 0.26102819237147595,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016829931133560875,
+      "loss": 0.0324,
+      "step": 5509
+    },
+    {
+      "epoch": 0.26107557450841035,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016828843265823747,
+      "loss": 0.0147,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2611229566453447,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016827755246630148,
+      "loss": 0.7277,
+      "step": 5511
+    },
+    {
+      "epoch": 0.2611703387822791,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016826667076004213,
+      "loss": 0.5882,
+      "step": 5512
+    },
+    {
+      "epoch": 0.26121772091921347,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001682557875397007,
+      "loss": 0.8579,
+      "step": 5513
+    },
+    {
+      "epoch": 0.2612651030561478,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00016824490280551864,
+      "loss": 0.2415,
+      "step": 5514
+    },
+    {
+      "epoch": 0.2613124851930822,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001682340165577373,
+      "loss": 0.7873,
+      "step": 5515
+    },
+    {
+      "epoch": 0.2613598673300166,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016822312879659817,
+      "loss": 1.012,
+      "step": 5516
+    },
+    {
+      "epoch": 0.261407249466951,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001682122395223427,
+      "loss": 0.0454,
+      "step": 5517
+    },
+    {
+      "epoch": 0.2614546316038853,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001682013487352124,
+      "loss": 0.176,
+      "step": 5518
+    },
+    {
+      "epoch": 0.2615020137408197,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016819045643544885,
+      "loss": 0.254,
+      "step": 5519
+    },
+    {
+      "epoch": 0.2615493958777541,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016817956262329362,
+      "loss": 0.8728,
+      "step": 5520
+    },
+    {
+      "epoch": 0.26159677801468845,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001681686672989883,
+      "loss": 1.3349,
+      "step": 5521
+    },
+    {
+      "epoch": 0.26164416015162284,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016815777046277455,
+      "loss": 0.5688,
+      "step": 5522
+    },
+    {
+      "epoch": 0.26169154228855723,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016814687211489404,
+      "loss": 1.051,
+      "step": 5523
+    },
+    {
+      "epoch": 0.26173892442549157,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001681359722555885,
+      "loss": 0.9618,
+      "step": 5524
+    },
+    {
+      "epoch": 0.26178630656242596,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001681250708850997,
+      "loss": 0.7285,
+      "step": 5525
+    },
+    {
+      "epoch": 0.26183368869936036,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016811416800366935,
+      "loss": 1.0602,
+      "step": 5526
+    },
+    {
+      "epoch": 0.2618810708362947,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001681032636115393,
+      "loss": 0.0395,
+      "step": 5527
+    },
+    {
+      "epoch": 0.2619284529732291,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00016809235770895144,
+      "loss": 0.0628,
+      "step": 5528
+    },
+    {
+      "epoch": 0.2619758351101635,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016808145029614758,
+      "loss": 0.8732,
+      "step": 5529
+    },
+    {
+      "epoch": 0.2620232172470978,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00016807054137336973,
+      "loss": 1.06,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2620705993840322,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0001680596309408597,
+      "loss": 0.0016,
+      "step": 5531
+    },
+    {
+      "epoch": 0.2621179815209666,
+      "grad_norm": 0.025634765625,
+      "learning_rate": 0.0001680487189988596,
+      "loss": 0.0016,
+      "step": 5532
+    },
+    {
+      "epoch": 0.262165363657901,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016803780554761137,
+      "loss": 1.2372,
+      "step": 5533
+    },
+    {
+      "epoch": 0.26221274579483533,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001680268905873571,
+      "loss": 0.8734,
+      "step": 5534
+    },
+    {
+      "epoch": 0.2622601279317697,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00016801597411833885,
+      "loss": 0.0055,
+      "step": 5535
+    },
+    {
+      "epoch": 0.2623075100687041,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016800505614079876,
+      "loss": 1.2974,
+      "step": 5536
+    },
+    {
+      "epoch": 0.26235489220563846,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016799413665497892,
+      "loss": 1.3831,
+      "step": 5537
+    },
+    {
+      "epoch": 0.26240227434257285,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016798321566112158,
+      "loss": 1.1895,
+      "step": 5538
+    },
+    {
+      "epoch": 0.26244965647950724,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001679722931594689,
+      "loss": 1.0775,
+      "step": 5539
+    },
+    {
+      "epoch": 0.2624970386164416,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001679613691502632,
+      "loss": 0.8848,
+      "step": 5540
+    },
+    {
+      "epoch": 0.262544420753376,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016795044363374673,
+      "loss": 0.5862,
+      "step": 5541
+    },
+    {
+      "epoch": 0.26259180289031037,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016793951661016175,
+      "loss": 0.9577,
+      "step": 5542
+    },
+    {
+      "epoch": 0.2626391850272447,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016792858807975066,
+      "loss": 1.3822,
+      "step": 5543
+    },
+    {
+      "epoch": 0.2626865671641791,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001679176580427559,
+      "loss": 0.962,
+      "step": 5544
+    },
+    {
+      "epoch": 0.2627339493011135,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016790672649941976,
+      "loss": 1.5334,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2627813314380479,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001678957934499848,
+      "loss": 0.9279,
+      "step": 5546
+    },
+    {
+      "epoch": 0.2628287135749822,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016788485889469344,
+      "loss": 0.4792,
+      "step": 5547
+    },
+    {
+      "epoch": 0.2628760957119166,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016787392283378822,
+      "loss": 1.0632,
+      "step": 5548
+    },
+    {
+      "epoch": 0.262923477848851,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016786298526751166,
+      "loss": 0.0135,
+      "step": 5549
+    },
+    {
+      "epoch": 0.26297085998578534,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016785204619610642,
+      "loss": 0.8517,
+      "step": 5550
+    },
+    {
+      "epoch": 0.26301824212271974,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016784110561981507,
+      "loss": 0.0653,
+      "step": 5551
+    },
+    {
+      "epoch": 0.26306562425965413,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016783016353888024,
+      "loss": 1.2957,
+      "step": 5552
+    },
+    {
+      "epoch": 0.26311300639658847,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016781921995354462,
+      "loss": 0.7068,
+      "step": 5553
+    },
+    {
+      "epoch": 0.26316038853352286,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016780827486405096,
+      "loss": 1.0027,
+      "step": 5554
+    },
+    {
+      "epoch": 0.26320777067045725,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016779732827064197,
+      "loss": 0.104,
+      "step": 5555
+    },
+    {
+      "epoch": 0.2632551528073916,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016778638017356045,
+      "loss": 0.1015,
+      "step": 5556
+    },
+    {
+      "epoch": 0.263302534944326,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016777543057304922,
+      "loss": 0.1114,
+      "step": 5557
+    },
+    {
+      "epoch": 0.2633499170812604,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016776447946935115,
+      "loss": 0.0839,
+      "step": 5558
+    },
+    {
+      "epoch": 0.2633972992181947,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016775352686270912,
+      "loss": 0.3187,
+      "step": 5559
+    },
+    {
+      "epoch": 0.2634446813551291,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.000167742572753366,
+      "loss": 0.5824,
+      "step": 5560
+    },
+    {
+      "epoch": 0.2634920634920635,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00016773161714156478,
+      "loss": 0.1804,
+      "step": 5561
+    },
+    {
+      "epoch": 0.2635394456289979,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016772066002754846,
+      "loss": 0.0095,
+      "step": 5562
+    },
+    {
+      "epoch": 0.26358682776593223,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016770970141155998,
+      "loss": 1.1152,
+      "step": 5563
+    },
+    {
+      "epoch": 0.2636342099028666,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016769874129384248,
+      "loss": 1.1267,
+      "step": 5564
+    },
+    {
+      "epoch": 0.263681592039801,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.000167687779674639,
+      "loss": 1.0091,
+      "step": 5565
+    },
+    {
+      "epoch": 0.26372897417673535,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016767681655419268,
+      "loss": 0.7512,
+      "step": 5566
+    },
+    {
+      "epoch": 0.26377635631366975,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00016766585193274665,
+      "loss": 0.0189,
+      "step": 5567
+    },
+    {
+      "epoch": 0.26382373845060414,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001676548858105441,
+      "loss": 0.4407,
+      "step": 5568
+    },
+    {
+      "epoch": 0.2638711205875385,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016764391818782824,
+      "loss": 0.8798,
+      "step": 5569
+    },
+    {
+      "epoch": 0.26391850272447287,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016763294906484234,
+      "loss": 1.0353,
+      "step": 5570
+    },
+    {
+      "epoch": 0.26396588486140726,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016762197844182963,
+      "loss": 1.4533,
+      "step": 5571
+    },
+    {
+      "epoch": 0.2640132669983416,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016761100631903348,
+      "loss": 0.1956,
+      "step": 5572
+    },
+    {
+      "epoch": 0.264060649135276,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00016760003269669725,
+      "loss": 0.0535,
+      "step": 5573
+    },
+    {
+      "epoch": 0.2641080312722104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016758905757506426,
+      "loss": 1.2123,
+      "step": 5574
+    },
+    {
+      "epoch": 0.2641554134091448,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000167578080954378,
+      "loss": 0.3087,
+      "step": 5575
+    },
+    {
+      "epoch": 0.2642027955460791,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001675671028348819,
+      "loss": 1.1576,
+      "step": 5576
+    },
+    {
+      "epoch": 0.2642501776830135,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00016755612321681936,
+      "loss": 0.002,
+      "step": 5577
+    },
+    {
+      "epoch": 0.2642975598199479,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016754514210043402,
+      "loss": 0.1876,
+      "step": 5578
+    },
+    {
+      "epoch": 0.26434494195688224,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016753415948596935,
+      "loss": 1.0717,
+      "step": 5579
+    },
+    {
+      "epoch": 0.26439232409381663,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016752317537366897,
+      "loss": 1.1072,
+      "step": 5580
+    },
+    {
+      "epoch": 0.264439706230751,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001675121897637765,
+      "loss": 0.0101,
+      "step": 5581
+    },
+    {
+      "epoch": 0.26448708836768536,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016750120265653554,
+      "loss": 1.4156,
+      "step": 5582
+    },
+    {
+      "epoch": 0.26453447050461976,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0001674902140521898,
+      "loss": 0.0341,
+      "step": 5583
+    },
+    {
+      "epoch": 0.26458185264155415,
+      "grad_norm": 0.0108642578125,
+      "learning_rate": 0.00016747922395098303,
+      "loss": 0.0006,
+      "step": 5584
+    },
+    {
+      "epoch": 0.2646292347784885,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016746823235315895,
+      "loss": 1.0785,
+      "step": 5585
+    },
+    {
+      "epoch": 0.2646766169154229,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001674572392589613,
+      "loss": 0.1082,
+      "step": 5586
+    },
+    {
+      "epoch": 0.2647239990523573,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016744624466863395,
+      "loss": 0.9585,
+      "step": 5587
+    },
+    {
+      "epoch": 0.2647713811892916,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016743524858242075,
+      "loss": 0.853,
+      "step": 5588
+    },
+    {
+      "epoch": 0.264818763326226,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00016742425100056555,
+      "loss": 0.0078,
+      "step": 5589
+    },
+    {
+      "epoch": 0.2648661454631604,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001674132519233123,
+      "loss": 1.1826,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2649135276000948,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00016740225135090491,
+      "loss": 0.1609,
+      "step": 5591
+    },
+    {
+      "epoch": 0.2649609097370291,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001673912492835874,
+      "loss": 0.9908,
+      "step": 5592
+    },
+    {
+      "epoch": 0.2650082918739635,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00016738024572160375,
+      "loss": 0.7295,
+      "step": 5593
+    },
+    {
+      "epoch": 0.2650556740108979,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.000167369240665198,
+      "loss": 0.8419,
+      "step": 5594
+    },
+    {
+      "epoch": 0.26510305614783225,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001673582341146143,
+      "loss": 0.5265,
+      "step": 5595
+    },
+    {
+      "epoch": 0.26515043828476664,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016734722607009665,
+      "loss": 0.8965,
+      "step": 5596
+    },
+    {
+      "epoch": 0.26519782042170104,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016733621653188931,
+      "loss": 0.2007,
+      "step": 5597
+    },
+    {
+      "epoch": 0.2652452025586354,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016732520550023642,
+      "loss": 0.685,
+      "step": 5598
+    },
+    {
+      "epoch": 0.26529258469556977,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016731419297538212,
+      "loss": 0.5059,
+      "step": 5599
+    },
+    {
+      "epoch": 0.26533996683250416,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001673031789575708,
+      "loss": 1.1605,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2653873489694385,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016729216344704662,
+      "loss": 0.6541,
+      "step": 5601
+    },
+    {
+      "epoch": 0.2654347311063729,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016728114644405388,
+      "loss": 1.5109,
+      "step": 5602
+    },
+    {
+      "epoch": 0.2654821132433073,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016727012794883703,
+      "loss": 0.6519,
+      "step": 5603
+    },
+    {
+      "epoch": 0.2655294953802417,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001672591079616404,
+      "loss": 1.0167,
+      "step": 5604
+    },
+    {
+      "epoch": 0.265576877517176,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001672480864827084,
+      "loss": 0.6678,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2656242596541104,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00016723706351228545,
+      "loss": 0.6186,
+      "step": 5606
+    },
+    {
+      "epoch": 0.2656716417910448,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016722603905061604,
+      "loss": 0.7193,
+      "step": 5607
+    },
+    {
+      "epoch": 0.26571902392797914,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016721501309794473,
+      "loss": 0.7566,
+      "step": 5608
+    },
+    {
+      "epoch": 0.26576640606491353,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000167203985654516,
+      "loss": 0.7573,
+      "step": 5609
+    },
+    {
+      "epoch": 0.2658137882018479,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016719295672057443,
+      "loss": 0.5047,
+      "step": 5610
+    },
+    {
+      "epoch": 0.26586117033878226,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00016718192629636467,
+      "loss": 0.5751,
+      "step": 5611
+    },
+    {
+      "epoch": 0.26590855247571665,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016717089438213133,
+      "loss": 0.9762,
+      "step": 5612
+    },
+    {
+      "epoch": 0.26595593461265105,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016715986097811912,
+      "loss": 1.1714,
+      "step": 5613
+    },
+    {
+      "epoch": 0.2660033167495854,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00016714882608457273,
+      "loss": 1.4798,
+      "step": 5614
+    },
+    {
+      "epoch": 0.2660506988865198,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001671377897017369,
+      "loss": 1.0433,
+      "step": 5615
+    },
+    {
+      "epoch": 0.26609808102345417,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001671267518298564,
+      "loss": 0.8132,
+      "step": 5616
+    },
+    {
+      "epoch": 0.2661454631603885,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016711571246917607,
+      "loss": 1.2701,
+      "step": 5617
+    },
+    {
+      "epoch": 0.2661928452973229,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001671046716199407,
+      "loss": 0.7974,
+      "step": 5618
+    },
+    {
+      "epoch": 0.2662402274342573,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00016709362928239515,
+      "loss": 1.0338,
+      "step": 5619
+    },
+    {
+      "epoch": 0.2662876095711917,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016708258545678444,
+      "loss": 0.6452,
+      "step": 5620
+    },
+    {
+      "epoch": 0.266334991708126,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001670715401433534,
+      "loss": 0.0214,
+      "step": 5621
+    },
+    {
+      "epoch": 0.2663823738450604,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016706049334234705,
+      "loss": 1.0348,
+      "step": 5622
+    },
+    {
+      "epoch": 0.2664297559819948,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001670494450540104,
+      "loss": 0.6407,
+      "step": 5623
+    },
+    {
+      "epoch": 0.26647713811892915,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016703839527858848,
+      "loss": 1.1821,
+      "step": 5624
+    },
+    {
+      "epoch": 0.26652452025586354,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016702734401632632,
+      "loss": 0.3451,
+      "step": 5625
+    },
+    {
+      "epoch": 0.26657190239279793,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016701629126746908,
+      "loss": 0.1856,
+      "step": 5626
+    },
+    {
+      "epoch": 0.26661928452973227,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001670052370322619,
+      "loss": 0.0511,
+      "step": 5627
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001669941813109499,
+      "loss": 1.1102,
+      "step": 5628
+    },
+    {
+      "epoch": 0.26671404880360106,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00016698312410377833,
+      "loss": 0.1564,
+      "step": 5629
+    },
+    {
+      "epoch": 0.2667614309405354,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00016697206541099245,
+      "loss": 0.1795,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2668088130774698,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00016696100523283744,
+      "loss": 1.0263,
+      "step": 5631
+    },
+    {
+      "epoch": 0.2668561952144042,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001669499435695587,
+      "loss": 0.5158,
+      "step": 5632
+    },
+    {
+      "epoch": 0.2669035773513386,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016693888042140152,
+      "loss": 1.1575,
+      "step": 5633
+    },
+    {
+      "epoch": 0.2669509594882729,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00016692781578861126,
+      "loss": 1.0503,
+      "step": 5634
+    },
+    {
+      "epoch": 0.2669983416252073,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001669167496714333,
+      "loss": 0.9357,
+      "step": 5635
+    },
+    {
+      "epoch": 0.2670457237621417,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016690568207011313,
+      "loss": 0.9347,
+      "step": 5636
+    },
+    {
+      "epoch": 0.26709310589907603,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016689461298489625,
+      "loss": 1.0529,
+      "step": 5637
+    },
+    {
+      "epoch": 0.2671404880360104,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016688354241602805,
+      "loss": 0.8088,
+      "step": 5638
+    },
+    {
+      "epoch": 0.2671878701729448,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016687247036375414,
+      "loss": 1.321,
+      "step": 5639
+    },
+    {
+      "epoch": 0.26723525230987916,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016686139682832004,
+      "loss": 0.8933,
+      "step": 5640
+    },
+    {
+      "epoch": 0.26728263444681355,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016685032180997143,
+      "loss": 0.8912,
+      "step": 5641
+    },
+    {
+      "epoch": 0.26733001658374794,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016683924530895385,
+      "loss": 0.0123,
+      "step": 5642
+    },
+    {
+      "epoch": 0.2673773987206823,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.000166828167325513,
+      "loss": 1.3312,
+      "step": 5643
+    },
+    {
+      "epoch": 0.2674247808576167,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001668170878598946,
+      "loss": 0.0041,
+      "step": 5644
+    },
+    {
+      "epoch": 0.26747216299455107,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016680600691234437,
+      "loss": 0.8034,
+      "step": 5645
+    },
+    {
+      "epoch": 0.2675195451314854,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00016679492448310804,
+      "loss": 0.4828,
+      "step": 5646
+    },
+    {
+      "epoch": 0.2675669272684198,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016678384057243145,
+      "loss": 0.0189,
+      "step": 5647
+    },
+    {
+      "epoch": 0.2676143094053542,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016677275518056037,
+      "loss": 1.4275,
+      "step": 5648
+    },
+    {
+      "epoch": 0.2676616915422886,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016676166830774074,
+      "loss": 0.9209,
+      "step": 5649
+    },
+    {
+      "epoch": 0.2677090736792229,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016675057995421843,
+      "loss": 0.7652,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2677564558161573,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016673949012023932,
+      "loss": 0.5517,
+      "step": 5651
+    },
+    {
+      "epoch": 0.2678038379530917,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016672839880604943,
+      "loss": 0.9392,
+      "step": 5652
+    },
+    {
+      "epoch": 0.26785122009002604,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016671730601189473,
+      "loss": 1.086,
+      "step": 5653
+    },
+    {
+      "epoch": 0.26789860222696044,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016670621173802124,
+      "loss": 1.0273,
+      "step": 5654
+    },
+    {
+      "epoch": 0.26794598436389483,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00016669511598467503,
+      "loss": 0.0101,
+      "step": 5655
+    },
+    {
+      "epoch": 0.26799336650082917,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001666840187521022,
+      "loss": 1.0378,
+      "step": 5656
+    },
+    {
+      "epoch": 0.26804074863776356,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016667292004054885,
+      "loss": 0.6257,
+      "step": 5657
+    },
+    {
+      "epoch": 0.26808813077469795,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016666181985026113,
+      "loss": 0.6897,
+      "step": 5658
+    },
+    {
+      "epoch": 0.2681355129116323,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016665071818148527,
+      "loss": 0.8365,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2681828950485667,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016663961503446748,
+      "loss": 0.2519,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2682302771855011,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000166628510409454,
+      "loss": 0.4015,
+      "step": 5661
+    },
+    {
+      "epoch": 0.26827765932243547,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016661740430669116,
+      "loss": 0.8856,
+      "step": 5662
+    },
+    {
+      "epoch": 0.2683250414593698,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001666062967264252,
+      "loss": 1.2056,
+      "step": 5663
+    },
+    {
+      "epoch": 0.2683724235963042,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016659518766890257,
+      "loss": 1.3711,
+      "step": 5664
+    },
+    {
+      "epoch": 0.2684198057332386,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016658407713436956,
+      "loss": 0.8598,
+      "step": 5665
+    },
+    {
+      "epoch": 0.26846718787017293,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00016657296512307266,
+      "loss": 1.4123,
+      "step": 5666
+    },
+    {
+      "epoch": 0.2685145700071073,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016656185163525832,
+      "loss": 0.8251,
+      "step": 5667
+    },
+    {
+      "epoch": 0.2685619521440417,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.000166550736671173,
+      "loss": 0.1422,
+      "step": 5668
+    },
+    {
+      "epoch": 0.26860933428097605,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016653962023106323,
+      "loss": 1.0764,
+      "step": 5669
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00016652850231517558,
+      "loss": 1.195,
+      "step": 5670
+    },
+    {
+      "epoch": 0.26870409855484484,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001665173829237566,
+      "loss": 0.7752,
+      "step": 5671
+    },
+    {
+      "epoch": 0.2687514806917792,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001665062620570529,
+      "loss": 0.8655,
+      "step": 5672
+    },
+    {
+      "epoch": 0.26879886282871357,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016649513971531114,
+      "loss": 1.2131,
+      "step": 5673
+    },
+    {
+      "epoch": 0.26884624496564796,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00016648401589877804,
+      "loss": 1.2834,
+      "step": 5674
+    },
+    {
+      "epoch": 0.2688936271025823,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016647289060770027,
+      "loss": 0.1421,
+      "step": 5675
+    },
+    {
+      "epoch": 0.2689410092395167,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016646176384232456,
+      "loss": 1.0725,
+      "step": 5676
+    },
+    {
+      "epoch": 0.2689883913764511,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016645063560289777,
+      "loss": 1.1293,
+      "step": 5677
+    },
+    {
+      "epoch": 0.2690357735133855,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016643950588966663,
+      "loss": 1.3876,
+      "step": 5678
+    },
+    {
+      "epoch": 0.2690831556503198,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016642837470287803,
+      "loss": 0.9904,
+      "step": 5679
+    },
+    {
+      "epoch": 0.2691305377872542,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001664172420427788,
+      "loss": 0.8577,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2691779199241886,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016640610790961591,
+      "loss": 0.2961,
+      "step": 5681
+    },
+    {
+      "epoch": 0.26922530206112294,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001663949723036363,
+      "loss": 1.6089,
+      "step": 5682
+    },
+    {
+      "epoch": 0.26927268419805733,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001663838352250869,
+      "loss": 0.948,
+      "step": 5683
+    },
+    {
+      "epoch": 0.2693200663349917,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001663726966742148,
+      "loss": 0.9098,
+      "step": 5684
+    },
+    {
+      "epoch": 0.26936744847192606,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016636155665126693,
+      "loss": 0.6937,
+      "step": 5685
+    },
+    {
+      "epoch": 0.26941483060886046,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00016635041515649043,
+      "loss": 0.0027,
+      "step": 5686
+    },
+    {
+      "epoch": 0.26946221274579485,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016633927219013238,
+      "loss": 0.2111,
+      "step": 5687
+    },
+    {
+      "epoch": 0.2695095948827292,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016632812775243996,
+      "loss": 0.6147,
+      "step": 5688
+    },
+    {
+      "epoch": 0.2695569770196636,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001663169818436603,
+      "loss": 0.7323,
+      "step": 5689
+    },
+    {
+      "epoch": 0.269604359156598,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016630583446404062,
+      "loss": 0.4665,
+      "step": 5690
+    },
+    {
+      "epoch": 0.26965174129353237,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00016629468561382815,
+      "loss": 0.795,
+      "step": 5691
+    },
+    {
+      "epoch": 0.2696991234304667,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016628353529327022,
+      "loss": 0.5587,
+      "step": 5692
+    },
+    {
+      "epoch": 0.2697465055674011,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016627238350261402,
+      "loss": 1.3128,
+      "step": 5693
+    },
+    {
+      "epoch": 0.2697938877043355,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016626123024210697,
+      "loss": 1.0178,
+      "step": 5694
+    },
+    {
+      "epoch": 0.2698412698412698,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001662500755119964,
+      "loss": 0.2749,
+      "step": 5695
+    },
+    {
+      "epoch": 0.2698886519782042,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00016623891931252972,
+      "loss": 0.1534,
+      "step": 5696
+    },
+    {
+      "epoch": 0.2699360341151386,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00016622776164395436,
+      "loss": 1.0676,
+      "step": 5697
+    },
+    {
+      "epoch": 0.26998341625207295,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0001662166025065178,
+      "loss": 0.0204,
+      "step": 5698
+    },
+    {
+      "epoch": 0.27003079838900734,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016620544190046752,
+      "loss": 1.3422,
+      "step": 5699
+    },
+    {
+      "epoch": 0.27007818052594174,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001661942798260511,
+      "loss": 0.9876,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2701255626628761,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000166183116283516,
+      "loss": 0.018,
+      "step": 5701
+    },
+    {
+      "epoch": 0.27017294479981047,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001661719512731099,
+      "loss": 1.3744,
+      "step": 5702
+    },
+    {
+      "epoch": 0.27022032693674486,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001661607847950804,
+      "loss": 0.0785,
+      "step": 5703
+    },
+    {
+      "epoch": 0.2702677090736792,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0001661496168496752,
+      "loss": 0.0065,
+      "step": 5704
+    },
+    {
+      "epoch": 0.2703150912106136,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001661384474371419,
+      "loss": 1.119,
+      "step": 5705
+    },
+    {
+      "epoch": 0.270362473347548,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016612727655772831,
+      "loss": 1.0204,
+      "step": 5706
+    },
+    {
+      "epoch": 0.2704098554844824,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016611610421168215,
+      "loss": 0.8329,
+      "step": 5707
+    },
+    {
+      "epoch": 0.2704572376214167,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016610493039925125,
+      "loss": 0.9431,
+      "step": 5708
+    },
+    {
+      "epoch": 0.2705046197583511,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.00016609375512068337,
+      "loss": 0.0081,
+      "step": 5709
+    },
+    {
+      "epoch": 0.2705520018952855,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016608257837622646,
+      "loss": 0.2527,
+      "step": 5710
+    },
+    {
+      "epoch": 0.27059938403221984,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016607140016612826,
+      "loss": 1.0873,
+      "step": 5711
+    },
+    {
+      "epoch": 0.27064676616915423,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016606022049063686,
+      "loss": 1.42,
+      "step": 5712
+    },
+    {
+      "epoch": 0.2706941483060886,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001660490393500001,
+      "loss": 0.0884,
+      "step": 5713
+    },
+    {
+      "epoch": 0.27074153044302296,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000166037856744466,
+      "loss": 1.1958,
+      "step": 5714
+    },
+    {
+      "epoch": 0.27078891257995735,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00016602667267428257,
+      "loss": 0.1684,
+      "step": 5715
+    },
+    {
+      "epoch": 0.27083629471689175,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001660154871396979,
+      "loss": 0.5702,
+      "step": 5716
+    },
+    {
+      "epoch": 0.2708836768538261,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016600430014096,
+      "loss": 0.872,
+      "step": 5717
+    },
+    {
+      "epoch": 0.2709310589907605,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016599311167831706,
+      "loss": 1.018,
+      "step": 5718
+    },
+    {
+      "epoch": 0.27097844112769487,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00016598192175201718,
+      "loss": 0.3869,
+      "step": 5719
+    },
+    {
+      "epoch": 0.27102582326462926,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016597073036230854,
+      "loss": 0.7992,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2710732054015636,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0001659595375094394,
+      "loss": 0.0172,
+      "step": 5721
+    },
+    {
+      "epoch": 0.271120587538498,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016594834319365797,
+      "loss": 0.5832,
+      "step": 5722
+    },
+    {
+      "epoch": 0.2711679696754324,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016593714741521253,
+      "loss": 0.6701,
+      "step": 5723
+    },
+    {
+      "epoch": 0.2712153518123667,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001659259501743514,
+      "loss": 0.8962,
+      "step": 5724
+    },
+    {
+      "epoch": 0.2712627339493011,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001659147514713229,
+      "loss": 0.8608,
+      "step": 5725
+    },
+    {
+      "epoch": 0.2713101160862355,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016590355130637546,
+      "loss": 0.9534,
+      "step": 5726
+    },
+    {
+      "epoch": 0.27135749822316985,
+      "grad_norm": 0.0220947265625,
+      "learning_rate": 0.0001658923496797574,
+      "loss": 0.001,
+      "step": 5727
+    },
+    {
+      "epoch": 0.27140488036010424,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00016588114659171722,
+      "loss": 0.0697,
+      "step": 5728
+    },
+    {
+      "epoch": 0.27145226249703863,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016586994204250338,
+      "loss": 1.0821,
+      "step": 5729
+    },
+    {
+      "epoch": 0.27149964463397297,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001658587360323644,
+      "loss": 0.9673,
+      "step": 5730
+    },
+    {
+      "epoch": 0.27154702677090736,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001658475285615488,
+      "loss": 0.4256,
+      "step": 5731
+    },
+    {
+      "epoch": 0.27159440890784176,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016583631963030518,
+      "loss": 0.9804,
+      "step": 5732
+    },
+    {
+      "epoch": 0.2716417910447761,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001658251092388821,
+      "loss": 0.9756,
+      "step": 5733
+    },
+    {
+      "epoch": 0.2716891731817105,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001658138973875282,
+      "loss": 1.2051,
+      "step": 5734
+    },
+    {
+      "epoch": 0.2717365553186449,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016580268407649214,
+      "loss": 0.1696,
+      "step": 5735
+    },
+    {
+      "epoch": 0.27178393745557927,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016579146930602266,
+      "loss": 0.1459,
+      "step": 5736
+    },
+    {
+      "epoch": 0.2718313195925136,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016578025307636846,
+      "loss": 0.9619,
+      "step": 5737
+    },
+    {
+      "epoch": 0.271878701729448,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016576903538777834,
+      "loss": 1.0552,
+      "step": 5738
+    },
+    {
+      "epoch": 0.2719260838663824,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016575781624050104,
+      "loss": 0.7923,
+      "step": 5739
+    },
+    {
+      "epoch": 0.27197346600331673,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016574659563478543,
+      "loss": 0.1956,
+      "step": 5740
+    },
+    {
+      "epoch": 0.2720208481402511,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016573537357088036,
+      "loss": 0.6846,
+      "step": 5741
+    },
+    {
+      "epoch": 0.2720682302771855,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001657241500490347,
+      "loss": 1.2539,
+      "step": 5742
+    },
+    {
+      "epoch": 0.27211561241411986,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016571292506949742,
+      "loss": 0.1037,
+      "step": 5743
+    },
+    {
+      "epoch": 0.27216299455105425,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016570169863251746,
+      "loss": 0.7723,
+      "step": 5744
+    },
+    {
+      "epoch": 0.27221037668798864,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00016569047073834378,
+      "loss": 1.033,
+      "step": 5745
+    },
+    {
+      "epoch": 0.272257758824923,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016567924138722546,
+      "loss": 1.2771,
+      "step": 5746
+    },
+    {
+      "epoch": 0.2723051409618574,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001656680105794115,
+      "loss": 0.0527,
+      "step": 5747
+    },
+    {
+      "epoch": 0.27235252309879177,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016565677831515104,
+      "loss": 1.0547,
+      "step": 5748
+    },
+    {
+      "epoch": 0.27239990523572616,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016564554459469317,
+      "loss": 0.8906,
+      "step": 5749
+    },
+    {
+      "epoch": 0.2724472873726605,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.000165634309418287,
+      "loss": 1.0512,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2724946695095949,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016562307278618178,
+      "loss": 1.1705,
+      "step": 5751
+    },
+    {
+      "epoch": 0.2725420516465293,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00016561183469862673,
+      "loss": 1.1822,
+      "step": 5752
+    },
+    {
+      "epoch": 0.2725894337834636,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016560059515587105,
+      "loss": 0.7487,
+      "step": 5753
+    },
+    {
+      "epoch": 0.272636815920398,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016558935415816403,
+      "loss": 0.6775,
+      "step": 5754
+    },
+    {
+      "epoch": 0.2726841980573324,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00016557811170575504,
+      "loss": 0.9706,
+      "step": 5755
+    },
+    {
+      "epoch": 0.27273158019426674,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016556686779889337,
+      "loss": 0.9756,
+      "step": 5756
+    },
+    {
+      "epoch": 0.27277896233120114,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0001655556224378284,
+      "loss": 0.0017,
+      "step": 5757
+    },
+    {
+      "epoch": 0.27282634446813553,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016554437562280952,
+      "loss": 1.2672,
+      "step": 5758
+    },
+    {
+      "epoch": 0.27287372660506987,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016553312735408625,
+      "loss": 0.1707,
+      "step": 5759
+    },
+    {
+      "epoch": 0.27292110874200426,
+      "grad_norm": 0.0133056640625,
+      "learning_rate": 0.00016552187763190797,
+      "loss": 0.0005,
+      "step": 5760
+    },
+    {
+      "epoch": 0.27296849087893865,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001655106264565243,
+      "loss": 1.0667,
+      "step": 5761
+    },
+    {
+      "epoch": 0.273015873015873,
+      "grad_norm": 0.01031494140625,
+      "learning_rate": 0.00016549937382818466,
+      "loss": 0.0005,
+      "step": 5762
+    },
+    {
+      "epoch": 0.2730632551528074,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001654881197471387,
+      "loss": 0.9846,
+      "step": 5763
+    },
+    {
+      "epoch": 0.2731106372897418,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016547686421363602,
+      "loss": 1.5557,
+      "step": 5764
+    },
+    {
+      "epoch": 0.27315801942667617,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001654656072279262,
+      "loss": 1.2228,
+      "step": 5765
+    },
+    {
+      "epoch": 0.2732054015636105,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016545434879025893,
+      "loss": 0.0552,
+      "step": 5766
+    },
+    {
+      "epoch": 0.2732527837005449,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00016544308890088395,
+      "loss": 0.1073,
+      "step": 5767
+    },
+    {
+      "epoch": 0.2733001658374793,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00016543182756005096,
+      "loss": 1.0415,
+      "step": 5768
+    },
+    {
+      "epoch": 0.27334754797441363,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016542056476800973,
+      "loss": 0.0164,
+      "step": 5769
+    },
+    {
+      "epoch": 0.273394930111348,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016540930052501006,
+      "loss": 0.9602,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2734423122482824,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016539803483130177,
+      "loss": 1.3939,
+      "step": 5771
+    },
+    {
+      "epoch": 0.27348969438521675,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016538676768713476,
+      "loss": 1.0973,
+      "step": 5772
+    },
+    {
+      "epoch": 0.27353707652215115,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016537549909275881,
+      "loss": 0.0422,
+      "step": 5773
+    },
+    {
+      "epoch": 0.27358445865908554,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00016536422904842398,
+      "loss": 0.0074,
+      "step": 5774
+    },
+    {
+      "epoch": 0.2736318407960199,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016535295755438017,
+      "loss": 0.8604,
+      "step": 5775
+    },
+    {
+      "epoch": 0.27367922293295427,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016534168461087738,
+      "loss": 0.2881,
+      "step": 5776
+    },
+    {
+      "epoch": 0.27372660506988866,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001653304102181656,
+      "loss": 0.9498,
+      "step": 5777
+    },
+    {
+      "epoch": 0.27377398720682306,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001653191343764949,
+      "loss": 1.4866,
+      "step": 5778
+    },
+    {
+      "epoch": 0.2738213693437574,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001653078570861154,
+      "loss": 1.2301,
+      "step": 5779
+    },
+    {
+      "epoch": 0.2738687514806918,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001652965783472772,
+      "loss": 0.0687,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2739161336176262,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001652852981602304,
+      "loss": 1.2015,
+      "step": 5781
+    },
+    {
+      "epoch": 0.2739635157545605,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016527401652522528,
+      "loss": 0.6943,
+      "step": 5782
+    },
+    {
+      "epoch": 0.2740108978914949,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00016526273344251197,
+      "loss": 0.0311,
+      "step": 5783
+    },
+    {
+      "epoch": 0.2740582800284293,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00016525144891234078,
+      "loss": 1.2077,
+      "step": 5784
+    },
+    {
+      "epoch": 0.27410566216536364,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0001652401629349619,
+      "loss": 0.0035,
+      "step": 5785
+    },
+    {
+      "epoch": 0.27415304430229803,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001652288755106257,
+      "loss": 0.5082,
+      "step": 5786
+    },
+    {
+      "epoch": 0.2742004264392324,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001652175866395826,
+      "loss": 0.9112,
+      "step": 5787
+    },
+    {
+      "epoch": 0.27424780857616676,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001652062963220828,
+      "loss": 1.5539,
+      "step": 5788
+    },
+    {
+      "epoch": 0.27429519071310116,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016519500455837688,
+      "loss": 1.0714,
+      "step": 5789
+    },
+    {
+      "epoch": 0.27434257285003555,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016518371134871513,
+      "loss": 0.6494,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2743899549869699,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00016517241669334812,
+      "loss": 0.151,
+      "step": 5791
+    },
+    {
+      "epoch": 0.2744373371239043,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016516112059252635,
+      "loss": 0.3159,
+      "step": 5792
+    },
+    {
+      "epoch": 0.27448471926083867,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001651498230465003,
+      "loss": 1.2159,
+      "step": 5793
+    },
+    {
+      "epoch": 0.27453210139777307,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00016513852405552058,
+      "loss": 0.1881,
+      "step": 5794
+    },
+    {
+      "epoch": 0.2745794835347074,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016512722361983778,
+      "loss": 0.9019,
+      "step": 5795
+    },
+    {
+      "epoch": 0.2746268656716418,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016511592173970257,
+      "loss": 1.2556,
+      "step": 5796
+    },
+    {
+      "epoch": 0.2746742478085762,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001651046184153655,
+      "loss": 1.1222,
+      "step": 5797
+    },
+    {
+      "epoch": 0.2747216299455105,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00016509331364707739,
+      "loss": 0.0055,
+      "step": 5798
+    },
+    {
+      "epoch": 0.2747690120824449,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001650820074350889,
+      "loss": 0.8872,
+      "step": 5799
+    },
+    {
+      "epoch": 0.2748163942193793,
+      "grad_norm": 0.01165771484375,
+      "learning_rate": 0.00016507069977965083,
+      "loss": 0.0009,
+      "step": 5800
+    },
+    {
+      "epoch": 0.27486377635631365,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001650593906810139,
+      "loss": 0.1333,
+      "step": 5801
+    },
+    {
+      "epoch": 0.27491115849324804,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000165048080139429,
+      "loss": 0.0043,
+      "step": 5802
+    },
+    {
+      "epoch": 0.27495854063018244,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016503676815514702,
+      "loss": 1.1665,
+      "step": 5803
+    },
+    {
+      "epoch": 0.2750059227671168,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00016502545472841875,
+      "loss": 1.2289,
+      "step": 5804
+    },
+    {
+      "epoch": 0.27505330490405117,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016501413985949514,
+      "loss": 0.8153,
+      "step": 5805
+    },
+    {
+      "epoch": 0.27510068704098556,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001650028235486272,
+      "loss": 0.3612,
+      "step": 5806
+    },
+    {
+      "epoch": 0.27514806917791995,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016499150579606586,
+      "loss": 0.1876,
+      "step": 5807
+    },
+    {
+      "epoch": 0.2751954513148543,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016498018660206213,
+      "loss": 0.4883,
+      "step": 5808
+    },
+    {
+      "epoch": 0.2752428334517887,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00016496886596686707,
+      "loss": 0.0051,
+      "step": 5809
+    },
+    {
+      "epoch": 0.2752902155887231,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 0.00016495754389073182,
+      "loss": 0.0022,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2753375977256574,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001649462203739074,
+      "loss": 0.7888,
+      "step": 5811
+    },
+    {
+      "epoch": 0.2753849798625918,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016493489541664498,
+      "loss": 0.7594,
+      "step": 5812
+    },
+    {
+      "epoch": 0.2754323619995262,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016492356901919575,
+      "loss": 0.8629,
+      "step": 5813
+    },
+    {
+      "epoch": 0.27547974413646054,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001649122411818109,
+      "loss": 1.3489,
+      "step": 5814
+    },
+    {
+      "epoch": 0.27552712627339493,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016490091190474168,
+      "loss": 0.9438,
+      "step": 5815
+    },
+    {
+      "epoch": 0.2755745084103293,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001648895811882394,
+      "loss": 0.8693,
+      "step": 5816
+    },
+    {
+      "epoch": 0.27562189054726366,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001648782490325553,
+      "loss": 0.7939,
+      "step": 5817
+    },
+    {
+      "epoch": 0.27566927268419805,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00016486691543794076,
+      "loss": 0.1437,
+      "step": 5818
+    },
+    {
+      "epoch": 0.27571665482113245,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016485558040464713,
+      "loss": 0.6232,
+      "step": 5819
+    },
+    {
+      "epoch": 0.2757640369580668,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016484424393292576,
+      "loss": 0.5876,
+      "step": 5820
+    },
+    {
+      "epoch": 0.2758114190950012,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00016483290602302818,
+      "loss": 0.0837,
+      "step": 5821
+    },
+    {
+      "epoch": 0.27585880123193557,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016482156667520576,
+      "loss": 0.8046,
+      "step": 5822
+    },
+    {
+      "epoch": 0.27590618336886996,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016481022588971006,
+      "loss": 0.9897,
+      "step": 5823
+    },
+    {
+      "epoch": 0.2759535655058043,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001647988836667926,
+      "loss": 1.3133,
+      "step": 5824
+    },
+    {
+      "epoch": 0.2760009476427387,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00016478754000670489,
+      "loss": 1.4105,
+      "step": 5825
+    },
+    {
+      "epoch": 0.2760483297796731,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016477619490969857,
+      "loss": 0.7777,
+      "step": 5826
+    },
+    {
+      "epoch": 0.2760957119166074,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016476484837602527,
+      "loss": 0.9432,
+      "step": 5827
+    },
+    {
+      "epoch": 0.2761430940535418,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016475350040593656,
+      "loss": 1.1919,
+      "step": 5828
+    },
+    {
+      "epoch": 0.2761904761904762,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00016474215099968422,
+      "loss": 0.1892,
+      "step": 5829
+    },
+    {
+      "epoch": 0.27623785832741055,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001647308001575199,
+      "loss": 0.1605,
+      "step": 5830
+    },
+    {
+      "epoch": 0.27628524046434494,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001647194478796954,
+      "loss": 0.0698,
+      "step": 5831
+    },
+    {
+      "epoch": 0.27633262260127933,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016470809416646248,
+      "loss": 0.8024,
+      "step": 5832
+    },
+    {
+      "epoch": 0.27638000473821367,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016469673901807296,
+      "loss": 1.1148,
+      "step": 5833
+    },
+    {
+      "epoch": 0.27642738687514806,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0001646853824347787,
+      "loss": 0.0053,
+      "step": 5834
+    },
+    {
+      "epoch": 0.27647476901208246,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00016467402441683153,
+      "loss": 0.0064,
+      "step": 5835
+    },
+    {
+      "epoch": 0.27652215114901685,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001646626649644834,
+      "loss": 1.199,
+      "step": 5836
+    },
+    {
+      "epoch": 0.2765695332859512,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001646513040779862,
+      "loss": 0.1704,
+      "step": 5837
+    },
+    {
+      "epoch": 0.2766169154228856,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00016463994175759197,
+      "loss": 0.0157,
+      "step": 5838
+    },
+    {
+      "epoch": 0.27666429755981997,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016462857800355268,
+      "loss": 0.0994,
+      "step": 5839
+    },
+    {
+      "epoch": 0.2767116796967543,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016461721281612037,
+      "loss": 0.7554,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2767590618336887,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001646058461955471,
+      "loss": 0.0174,
+      "step": 5841
+    },
+    {
+      "epoch": 0.2768064439706231,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016459447814208498,
+      "loss": 0.5143,
+      "step": 5842
+    },
+    {
+      "epoch": 0.27685382610755743,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016458310865598617,
+      "loss": 0.8134,
+      "step": 5843
+    },
+    {
+      "epoch": 0.2769012082444918,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016457173773750275,
+      "loss": 0.7829,
+      "step": 5844
+    },
+    {
+      "epoch": 0.2769485903814262,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.000164560365386887,
+      "loss": 0.0016,
+      "step": 5845
+    },
+    {
+      "epoch": 0.27699597251836056,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001645489916043911,
+      "loss": 0.7386,
+      "step": 5846
+    },
+    {
+      "epoch": 0.27704335465529495,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001645376163902673,
+      "loss": 1.1325,
+      "step": 5847
+    },
+    {
+      "epoch": 0.27709073679222934,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00016452623974476795,
+      "loss": 0.4153,
+      "step": 5848
+    },
+    {
+      "epoch": 0.2771381189291637,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001645148616681453,
+      "loss": 0.8578,
+      "step": 5849
+    },
+    {
+      "epoch": 0.2771855010660981,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016450348216065177,
+      "loss": 0.9116,
+      "step": 5850
+    },
+    {
+      "epoch": 0.27723288320303247,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016449210122253968,
+      "loss": 1.5575,
+      "step": 5851
+    },
+    {
+      "epoch": 0.27728026533996686,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00016448071885406148,
+      "loss": 0.1502,
+      "step": 5852
+    },
+    {
+      "epoch": 0.2773276474769012,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016446933505546963,
+      "loss": 1.5635,
+      "step": 5853
+    },
+    {
+      "epoch": 0.2773750296138356,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016445794982701661,
+      "loss": 0.912,
+      "step": 5854
+    },
+    {
+      "epoch": 0.27742241175077,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016444656316895491,
+      "loss": 1.082,
+      "step": 5855
+    },
+    {
+      "epoch": 0.2774697938877043,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016443517508153707,
+      "loss": 0.7714,
+      "step": 5856
+    },
+    {
+      "epoch": 0.2775171760246387,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001644237855650157,
+      "loss": 0.6867,
+      "step": 5857
+    },
+    {
+      "epoch": 0.2775645581615731,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016441239461964337,
+      "loss": 1.5592,
+      "step": 5858
+    },
+    {
+      "epoch": 0.27761194029850744,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016440100224567275,
+      "loss": 0.0292,
+      "step": 5859
+    },
+    {
+      "epoch": 0.27765932243544184,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0001643896084433565,
+      "loss": 0.0025,
+      "step": 5860
+    },
+    {
+      "epoch": 0.27770670457237623,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016437821321294732,
+      "loss": 1.1657,
+      "step": 5861
+    },
+    {
+      "epoch": 0.27775408670931057,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016436681655469793,
+      "loss": 0.7367,
+      "step": 5862
+    },
+    {
+      "epoch": 0.27780146884624496,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001643554184688611,
+      "loss": 1.1748,
+      "step": 5863
+    },
+    {
+      "epoch": 0.27784885098317935,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00016434401895568966,
+      "loss": 1.0877,
+      "step": 5864
+    },
+    {
+      "epoch": 0.27789623312011374,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001643326180154364,
+      "loss": 0.0143,
+      "step": 5865
+    },
+    {
+      "epoch": 0.2779436152570481,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016432121564835422,
+      "loss": 0.7058,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2779909973939825,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016430981185469595,
+      "loss": 1.2879,
+      "step": 5867
+    },
+    {
+      "epoch": 0.27803837953091687,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0001642984066347146,
+      "loss": 0.1743,
+      "step": 5868
+    },
+    {
+      "epoch": 0.2780857616678512,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016428699998866303,
+      "loss": 0.734,
+      "step": 5869
+    },
+    {
+      "epoch": 0.2781331438047856,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001642755919167943,
+      "loss": 1.0882,
+      "step": 5870
+    },
+    {
+      "epoch": 0.27818052594172,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00016426418241936142,
+      "loss": 0.0577,
+      "step": 5871
+    },
+    {
+      "epoch": 0.27822790807865433,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016425277149661736,
+      "loss": 1.1437,
+      "step": 5872
+    },
+    {
+      "epoch": 0.2782752902155887,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016424135914881534,
+      "loss": 1.2461,
+      "step": 5873
+    },
+    {
+      "epoch": 0.2783226723525231,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001642299453762084,
+      "loss": 1.051,
+      "step": 5874
+    },
+    {
+      "epoch": 0.27837005448945745,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016421853017904963,
+      "loss": 0.5847,
+      "step": 5875
+    },
+    {
+      "epoch": 0.27841743662639185,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016420711355759232,
+      "loss": 0.1426,
+      "step": 5876
+    },
+    {
+      "epoch": 0.27846481876332624,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016419569551208958,
+      "loss": 0.0234,
+      "step": 5877
+    },
+    {
+      "epoch": 0.2785122009002606,
+      "grad_norm": 0.0107421875,
+      "learning_rate": 0.00016418427604279471,
+      "loss": 0.0005,
+      "step": 5878
+    },
+    {
+      "epoch": 0.27855958303719497,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016417285514996096,
+      "loss": 1.1206,
+      "step": 5879
+    },
+    {
+      "epoch": 0.27860696517412936,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00016416143283384165,
+      "loss": 0.0053,
+      "step": 5880
+    },
+    {
+      "epoch": 0.27865434731106375,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001641500090946901,
+      "loss": 1.0242,
+      "step": 5881
+    },
+    {
+      "epoch": 0.2787017294479981,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016413858393275968,
+      "loss": 0.9941,
+      "step": 5882
+    },
+    {
+      "epoch": 0.2787491115849325,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016412715734830376,
+      "loss": 0.2387,
+      "step": 5883
+    },
+    {
+      "epoch": 0.2787964937218669,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00016411572934157582,
+      "loss": 0.0203,
+      "step": 5884
+    },
+    {
+      "epoch": 0.2788438758588012,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016410429991282928,
+      "loss": 0.8312,
+      "step": 5885
+    },
+    {
+      "epoch": 0.2788912579957356,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016409286906231765,
+      "loss": 0.8291,
+      "step": 5886
+    },
+    {
+      "epoch": 0.27893864013267,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016408143679029445,
+      "loss": 1.2022,
+      "step": 5887
+    },
+    {
+      "epoch": 0.27898602226960434,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016407000309701323,
+      "loss": 0.826,
+      "step": 5888
+    },
+    {
+      "epoch": 0.27903340440653873,
+      "grad_norm": 0.0247802734375,
+      "learning_rate": 0.0001640585679827276,
+      "loss": 0.0015,
+      "step": 5889
+    },
+    {
+      "epoch": 0.2790807865434731,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001640471314476912,
+      "loss": 0.9617,
+      "step": 5890
+    },
+    {
+      "epoch": 0.27912816868040746,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001640356934921576,
+      "loss": 0.0169,
+      "step": 5891
+    },
+    {
+      "epoch": 0.27917555081734186,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00016402425411638053,
+      "loss": 1.1381,
+      "step": 5892
+    },
+    {
+      "epoch": 0.27922293295427625,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001640128133206137,
+      "loss": 1.4161,
+      "step": 5893
+    },
+    {
+      "epoch": 0.27927031509121064,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016400137110511085,
+      "loss": 0.0422,
+      "step": 5894
+    },
+    {
+      "epoch": 0.279317697228145,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016398992747012578,
+      "loss": 1.1908,
+      "step": 5895
+    },
+    {
+      "epoch": 0.27936507936507937,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00016397848241591224,
+      "loss": 0.0074,
+      "step": 5896
+    },
+    {
+      "epoch": 0.27941246150201376,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016396703594272416,
+      "loss": 1.2606,
+      "step": 5897
+    },
+    {
+      "epoch": 0.2794598436389481,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016395558805081532,
+      "loss": 0.5287,
+      "step": 5898
+    },
+    {
+      "epoch": 0.2795072257758825,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016394413874043967,
+      "loss": 0.6665,
+      "step": 5899
+    },
+    {
+      "epoch": 0.2795546079128169,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016393268801185115,
+      "loss": 0.8474,
+      "step": 5900
+    },
+    {
+      "epoch": 0.2796019900497512,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001639212358653037,
+      "loss": 1.1306,
+      "step": 5901
+    },
+    {
+      "epoch": 0.2796493721866856,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00016390978230105128,
+      "loss": 0.0694,
+      "step": 5902
+    },
+    {
+      "epoch": 0.27969675432362,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016389832731934803,
+      "loss": 0.1758,
+      "step": 5903
+    },
+    {
+      "epoch": 0.27974413646055435,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016388687092044787,
+      "loss": 0.1082,
+      "step": 5904
+    },
+    {
+      "epoch": 0.27979151859748874,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016387541310460506,
+      "loss": 0.6653,
+      "step": 5905
+    },
+    {
+      "epoch": 0.27983890073442313,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016386395387207355,
+      "loss": 0.0071,
+      "step": 5906
+    },
+    {
+      "epoch": 0.2798862828713575,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001638524932231076,
+      "loss": 0.1529,
+      "step": 5907
+    },
+    {
+      "epoch": 0.27993366500829187,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00016384103115796132,
+      "loss": 0.0179,
+      "step": 5908
+    },
+    {
+      "epoch": 0.27998104714522626,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.000163829567676889,
+      "loss": 1.1626,
+      "step": 5909
+    },
+    {
+      "epoch": 0.28002842928216065,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016381810278014484,
+      "loss": 1.0607,
+      "step": 5910
+    },
+    {
+      "epoch": 0.280075811419095,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016380663646798315,
+      "loss": 1.2028,
+      "step": 5911
+    },
+    {
+      "epoch": 0.2801231935560294,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016379516874065825,
+      "loss": 0.6336,
+      "step": 5912
+    },
+    {
+      "epoch": 0.2801705756929638,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016378369959842444,
+      "loss": 0.0943,
+      "step": 5913
+    },
+    {
+      "epoch": 0.2802179578298981,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00016377222904153608,
+      "loss": 1.0451,
+      "step": 5914
+    },
+    {
+      "epoch": 0.2802653399668325,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016376075707024766,
+      "loss": 1.0668,
+      "step": 5915
+    },
+    {
+      "epoch": 0.2803127221037669,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00016374928368481352,
+      "loss": 1.0897,
+      "step": 5916
+    },
+    {
+      "epoch": 0.28036010424070124,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016373780888548817,
+      "loss": 0.1122,
+      "step": 5917
+    },
+    {
+      "epoch": 0.28040748637763563,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016372633267252616,
+      "loss": 0.9279,
+      "step": 5918
+    },
+    {
+      "epoch": 0.28045486851457,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016371485504618191,
+      "loss": 1.1632,
+      "step": 5919
+    },
+    {
+      "epoch": 0.28050225065150436,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016370337600671008,
+      "loss": 0.6464,
+      "step": 5920
+    },
+    {
+      "epoch": 0.28054963278843875,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00016369189555436525,
+      "loss": 0.0027,
+      "step": 5921
+    },
+    {
+      "epoch": 0.28059701492537314,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016368041368940194,
+      "loss": 0.9308,
+      "step": 5922
+    },
+    {
+      "epoch": 0.2806443970623075,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016366893041207492,
+      "loss": 1.0763,
+      "step": 5923
+    },
+    {
+      "epoch": 0.2806917791992419,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00016365744572263886,
+      "loss": 0.135,
+      "step": 5924
+    },
+    {
+      "epoch": 0.28073916133617627,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016364595962134844,
+      "loss": 1.0708,
+      "step": 5925
+    },
+    {
+      "epoch": 0.28078654347311066,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016363447210845843,
+      "loss": 0.6604,
+      "step": 5926
+    },
+    {
+      "epoch": 0.280833925610045,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001636229831842236,
+      "loss": 1.2479,
+      "step": 5927
+    },
+    {
+      "epoch": 0.2808813077469794,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001636114928488988,
+      "loss": 1.0728,
+      "step": 5928
+    },
+    {
+      "epoch": 0.2809286898839138,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001636000011027388,
+      "loss": 0.856,
+      "step": 5929
+    },
+    {
+      "epoch": 0.2809760720208481,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016358850794599857,
+      "loss": 0.0754,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2810234541577825,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016357701337893295,
+      "loss": 0.7193,
+      "step": 5931
+    },
+    {
+      "epoch": 0.2810708362947169,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016356551740179687,
+      "loss": 0.6159,
+      "step": 5932
+    },
+    {
+      "epoch": 0.28111821843165125,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00016355402001484533,
+      "loss": 0.0552,
+      "step": 5933
+    },
+    {
+      "epoch": 0.28116560056858564,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016354252121833333,
+      "loss": 1.3275,
+      "step": 5934
+    },
+    {
+      "epoch": 0.28121298270552003,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00016353102101251588,
+      "loss": 0.8097,
+      "step": 5935
+    },
+    {
+      "epoch": 0.28126036484245437,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016351951939764806,
+      "loss": 0.7637,
+      "step": 5936
+    },
+    {
+      "epoch": 0.28130774697938876,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016350801637398497,
+      "loss": 0.5522,
+      "step": 5937
+    },
+    {
+      "epoch": 0.28135512911632315,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016349651194178173,
+      "loss": 1.2141,
+      "step": 5938
+    },
+    {
+      "epoch": 0.28140251125325755,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016348500610129346,
+      "loss": 1.445,
+      "step": 5939
+    },
+    {
+      "epoch": 0.2814498933901919,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016347349885277538,
+      "loss": 1.3921,
+      "step": 5940
+    },
+    {
+      "epoch": 0.2814972755271263,
+      "grad_norm": 0.00628662109375,
+      "learning_rate": 0.00016346199019648272,
+      "loss": 0.0003,
+      "step": 5941
+    },
+    {
+      "epoch": 0.28154465766406067,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00016345048013267067,
+      "loss": 0.0337,
+      "step": 5942
+    },
+    {
+      "epoch": 0.281592039800995,
+      "grad_norm": 0.03515625,
+      "learning_rate": 0.0001634389686615946,
+      "loss": 0.0011,
+      "step": 5943
+    },
+    {
+      "epoch": 0.2816394219379294,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016342745578350976,
+      "loss": 1.8758,
+      "step": 5944
+    },
+    {
+      "epoch": 0.2816868040748638,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016341594149867154,
+      "loss": 0.994,
+      "step": 5945
+    },
+    {
+      "epoch": 0.28173418621179813,
+      "grad_norm": 0.00262451171875,
+      "learning_rate": 0.00016340442580733527,
+      "loss": 0.0002,
+      "step": 5946
+    },
+    {
+      "epoch": 0.2817815683487325,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016339290870975636,
+      "loss": 1.3652,
+      "step": 5947
+    },
+    {
+      "epoch": 0.2818289504856669,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016338139020619026,
+      "loss": 0.0128,
+      "step": 5948
+    },
+    {
+      "epoch": 0.28187633262260126,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00016336987029689243,
+      "loss": 1.0002,
+      "step": 5949
+    },
+    {
+      "epoch": 0.28192371475953565,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016335834898211838,
+      "loss": 1.2971,
+      "step": 5950
+    },
+    {
+      "epoch": 0.28197109689647004,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016334682626212364,
+      "loss": 1.3942,
+      "step": 5951
+    },
+    {
+      "epoch": 0.2820184790334044,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016333530213716378,
+      "loss": 1.3696,
+      "step": 5952
+    },
+    {
+      "epoch": 0.28206586117033877,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016332377660749437,
+      "loss": 1.3783,
+      "step": 5953
+    },
+    {
+      "epoch": 0.28211324330727316,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016331224967337102,
+      "loss": 1.2957,
+      "step": 5954
+    },
+    {
+      "epoch": 0.28216062544420756,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00016330072133504943,
+      "loss": 0.134,
+      "step": 5955
+    },
+    {
+      "epoch": 0.2822080075811419,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00016328919159278526,
+      "loss": 0.0179,
+      "step": 5956
+    },
+    {
+      "epoch": 0.2822553897180763,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016327766044683424,
+      "loss": 0.594,
+      "step": 5957
+    },
+    {
+      "epoch": 0.2823027718550107,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016326612789745212,
+      "loss": 0.7436,
+      "step": 5958
+    },
+    {
+      "epoch": 0.282350153991945,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016325459394489467,
+      "loss": 1.401,
+      "step": 5959
+    },
+    {
+      "epoch": 0.2823975361288794,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001632430585894177,
+      "loss": 0.1428,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2824449182658138,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016323152183127704,
+      "loss": 0.8121,
+      "step": 5961
+    },
+    {
+      "epoch": 0.28249230040274814,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016321998367072857,
+      "loss": 1.1734,
+      "step": 5962
+    },
+    {
+      "epoch": 0.28253968253968254,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001632084441080282,
+      "loss": 1.1679,
+      "step": 5963
+    },
+    {
+      "epoch": 0.28258706467661693,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001631969031434319,
+      "loss": 0.1729,
+      "step": 5964
+    },
+    {
+      "epoch": 0.28263444681355127,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001631853607771956,
+      "loss": 1.3251,
+      "step": 5965
+    },
+    {
+      "epoch": 0.28268182895048566,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00016317381700957525,
+      "loss": 0.8001,
+      "step": 5966
+    },
+    {
+      "epoch": 0.28272921108742005,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016316227184082698,
+      "loss": 0.8081,
+      "step": 5967
+    },
+    {
+      "epoch": 0.28277659322435444,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001631507252712068,
+      "loss": 1.072,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2828239753612888,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00016313917730097078,
+      "loss": 0.3934,
+      "step": 5969
+    },
+    {
+      "epoch": 0.2828713574982232,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016312762793037507,
+      "loss": 1.1343,
+      "step": 5970
+    },
+    {
+      "epoch": 0.28291873963515757,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00016311607715967577,
+      "loss": 0.0241,
+      "step": 5971
+    },
+    {
+      "epoch": 0.2829661217720919,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016310452498912917,
+      "loss": 1.0106,
+      "step": 5972
+    },
+    {
+      "epoch": 0.2830135039090263,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016309297141899135,
+      "loss": 1.1859,
+      "step": 5973
+    },
+    {
+      "epoch": 0.2830608860459607,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016308141644951867,
+      "loss": 1.3627,
+      "step": 5974
+    },
+    {
+      "epoch": 0.28310826818289503,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001630698600809674,
+      "loss": 0.8214,
+      "step": 5975
+    },
+    {
+      "epoch": 0.2831556503198294,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016305830231359378,
+      "loss": 1.4405,
+      "step": 5976
+    },
+    {
+      "epoch": 0.2832030324567638,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016304674314765417,
+      "loss": 1.1925,
+      "step": 5977
+    },
+    {
+      "epoch": 0.28325041459369815,
+      "grad_norm": 0.01171875,
+      "learning_rate": 0.00016303518258340492,
+      "loss": 0.0006,
+      "step": 5978
+    },
+    {
+      "epoch": 0.28329779673063255,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001630236206211025,
+      "loss": 0.6789,
+      "step": 5979
+    },
+    {
+      "epoch": 0.28334517886756694,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016301205726100332,
+      "loss": 1.5498,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2833925610045013,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016300049250336383,
+      "loss": 0.7349,
+      "step": 5981
+    },
+    {
+      "epoch": 0.28343994314143567,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001629889263484405,
+      "loss": 0.0111,
+      "step": 5982
+    },
+    {
+      "epoch": 0.28348732527837006,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00016297735879648988,
+      "loss": 0.3184,
+      "step": 5983
+    },
+    {
+      "epoch": 0.28353470741530445,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001629657898477685,
+      "loss": 0.8369,
+      "step": 5984
+    },
+    {
+      "epoch": 0.2835820895522388,
+      "grad_norm": 0.00897216796875,
+      "learning_rate": 0.00016295421950253302,
+      "loss": 0.0005,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2836294716891732,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00016294264776103997,
+      "loss": 0.7341,
+      "step": 5986
+    },
+    {
+      "epoch": 0.2836768538261076,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00016293107462354603,
+      "loss": 0.6819,
+      "step": 5987
+    },
+    {
+      "epoch": 0.2837242359630419,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016291950009030793,
+      "loss": 0.9467,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2837716180999763,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016290792416158228,
+      "loss": 0.7839,
+      "step": 5989
+    },
+    {
+      "epoch": 0.2838190002369107,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016289634683762591,
+      "loss": 0.1468,
+      "step": 5990
+    },
+    {
+      "epoch": 0.28386638237384504,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016288476811869557,
+      "loss": 0.7732,
+      "step": 5991
+    },
+    {
+      "epoch": 0.28391376451077943,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00016287318800504804,
+      "loss": 0.0987,
+      "step": 5992
+    },
+    {
+      "epoch": 0.2839611466477138,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001628616064969402,
+      "loss": 0.8688,
+      "step": 5993
+    },
+    {
+      "epoch": 0.28400852878464816,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016285002359462883,
+      "loss": 1.3546,
+      "step": 5994
+    },
+    {
+      "epoch": 0.28405591092158256,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016283843929837093,
+      "loss": 0.8868,
+      "step": 5995
+    },
+    {
+      "epoch": 0.28410329305851695,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016282685360842334,
+      "loss": 0.8705,
+      "step": 5996
+    },
+    {
+      "epoch": 0.28415067519545134,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016281526652504308,
+      "loss": 1.1959,
+      "step": 5997
+    },
+    {
+      "epoch": 0.2841980573323857,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016280367804848711,
+      "loss": 1.3464,
+      "step": 5998
+    },
+    {
+      "epoch": 0.28424543946932007,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016279208817901247,
+      "loss": 0.0082,
+      "step": 5999
+    },
+    {
+      "epoch": 0.28429282160625446,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016278049691687617,
+      "loss": 1.8539,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2843402037431888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00016276890426233535,
+      "loss": 1.017,
+      "step": 6001
+    },
+    {
+      "epoch": 0.2843875858801232,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016275731021564706,
+      "loss": 0.1131,
+      "step": 6002
+    },
+    {
+      "epoch": 0.2844349680170576,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00016274571477706848,
+      "loss": 1.4311,
+      "step": 6003
+    },
+    {
+      "epoch": 0.2844823501539919,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001627341179468568,
+      "loss": 1.0981,
+      "step": 6004
+    },
+    {
+      "epoch": 0.2845297322909263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016272251972526921,
+      "loss": 0.5771,
+      "step": 6005
+    },
+    {
+      "epoch": 0.2845771144278607,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016271092011256293,
+      "loss": 0.0699,
+      "step": 6006
+    },
+    {
+      "epoch": 0.28462449656479505,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00016269931910899526,
+      "loss": 0.1502,
+      "step": 6007
+    },
+    {
+      "epoch": 0.28467187870172944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016268771671482343,
+      "loss": 1.0168,
+      "step": 6008
+    },
+    {
+      "epoch": 0.28471926083866383,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00016267611293030484,
+      "loss": 0.2216,
+      "step": 6009
+    },
+    {
+      "epoch": 0.28476664297559817,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016266450775569684,
+      "loss": 0.639,
+      "step": 6010
+    },
+    {
+      "epoch": 0.28481402511253257,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00016265290119125677,
+      "loss": 1.3192,
+      "step": 6011
+    },
+    {
+      "epoch": 0.28486140724946696,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001626412932372421,
+      "loss": 0.1411,
+      "step": 6012
+    },
+    {
+      "epoch": 0.28490878938640135,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016262968389391027,
+      "loss": 1.1672,
+      "step": 6013
+    },
+    {
+      "epoch": 0.2849561715233357,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001626180731615188,
+      "loss": 0.4654,
+      "step": 6014
+    },
+    {
+      "epoch": 0.2850035536602701,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016260646104032512,
+      "loss": 0.8847,
+      "step": 6015
+    },
+    {
+      "epoch": 0.2850509357972045,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00016259484753058682,
+      "loss": 0.1109,
+      "step": 6016
+    },
+    {
+      "epoch": 0.2850983179341388,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001625832326325615,
+      "loss": 1.344,
+      "step": 6017
+    },
+    {
+      "epoch": 0.2851457000710732,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001625716163465067,
+      "loss": 0.1425,
+      "step": 6018
+    },
+    {
+      "epoch": 0.2851930822080076,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016255999867268012,
+      "loss": 0.0183,
+      "step": 6019
+    },
+    {
+      "epoch": 0.28524046434494194,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016254837961133942,
+      "loss": 0.718,
+      "step": 6020
+    },
+    {
+      "epoch": 0.28528784648187633,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016253675916274226,
+      "loss": 1.0985,
+      "step": 6021
+    },
+    {
+      "epoch": 0.2853352286188107,
+      "grad_norm": 0.00787353515625,
+      "learning_rate": 0.00016252513732714643,
+      "loss": 0.0006,
+      "step": 6022
+    },
+    {
+      "epoch": 0.28538261075574506,
+      "grad_norm": 0.004425048828125,
+      "learning_rate": 0.00016251351410480962,
+      "loss": 0.0002,
+      "step": 6023
+    },
+    {
+      "epoch": 0.28542999289267945,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016250188949598965,
+      "loss": 1.2493,
+      "step": 6024
+    },
+    {
+      "epoch": 0.28547737502961384,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016249026350094437,
+      "loss": 0.1221,
+      "step": 6025
+    },
+    {
+      "epoch": 0.28552475716654824,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016247863611993158,
+      "loss": 0.8118,
+      "step": 6026
+    },
+    {
+      "epoch": 0.2855721393034826,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001624670073532092,
+      "loss": 1.1523,
+      "step": 6027
+    },
+    {
+      "epoch": 0.28561952144041697,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00016245537720103515,
+      "loss": 0.0994,
+      "step": 6028
+    },
+    {
+      "epoch": 0.28566690357735136,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00016244374566366733,
+      "loss": 0.9692,
+      "step": 6029
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016243211274136378,
+      "loss": 0.5362,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2857616678512201,
+      "grad_norm": 0.0086669921875,
+      "learning_rate": 0.00016242047843438245,
+      "loss": 0.0004,
+      "step": 6031
+    },
+    {
+      "epoch": 0.2858090499881545,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001624088427429814,
+      "loss": 1.7455,
+      "step": 6032
+    },
+    {
+      "epoch": 0.2858564321250888,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016239720566741867,
+      "loss": 0.9899,
+      "step": 6033
+    },
+    {
+      "epoch": 0.2859038142620232,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001623855672079524,
+      "loss": 0.0159,
+      "step": 6034
+    },
+    {
+      "epoch": 0.2859511963989576,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001623739273648407,
+      "loss": 0.4276,
+      "step": 6035
+    },
+    {
+      "epoch": 0.28599857853589195,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016236228613834171,
+      "loss": 0.9734,
+      "step": 6036
+    },
+    {
+      "epoch": 0.28604596067282634,
+      "grad_norm": 0.033203125,
+      "learning_rate": 0.00016235064352871365,
+      "loss": 0.0035,
+      "step": 6037
+    },
+    {
+      "epoch": 0.28609334280976073,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00016233899953621475,
+      "loss": 1.3255,
+      "step": 6038
+    },
+    {
+      "epoch": 0.28614072494669507,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00016232735416110323,
+      "loss": 2.0154,
+      "step": 6039
+    },
+    {
+      "epoch": 0.28618810708362946,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016231570740363738,
+      "loss": 1.0311,
+      "step": 6040
+    },
+    {
+      "epoch": 0.28623548922056385,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016230405926407552,
+      "loss": 0.7651,
+      "step": 6041
+    },
+    {
+      "epoch": 0.28628287135749825,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016229240974267596,
+      "loss": 0.9131,
+      "step": 6042
+    },
+    {
+      "epoch": 0.2863302534944326,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016228075883969713,
+      "loss": 0.0416,
+      "step": 6043
+    },
+    {
+      "epoch": 0.286377635631367,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001622691065553974,
+      "loss": 1.0857,
+      "step": 6044
+    },
+    {
+      "epoch": 0.28642501776830137,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016225745289003522,
+      "loss": 0.7592,
+      "step": 6045
+    },
+    {
+      "epoch": 0.2864723999052357,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016224579784386903,
+      "loss": 1.2407,
+      "step": 6046
+    },
+    {
+      "epoch": 0.2865197820421701,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016223414141715737,
+      "loss": 1.2729,
+      "step": 6047
+    },
+    {
+      "epoch": 0.2865671641791045,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016222248361015873,
+      "loss": 1.2148,
+      "step": 6048
+    },
+    {
+      "epoch": 0.28661454631603883,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00016221082442313168,
+      "loss": 0.9701,
+      "step": 6049
+    },
+    {
+      "epoch": 0.2866619284529732,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016219916385633483,
+      "loss": 0.9508,
+      "step": 6050
+    },
+    {
+      "epoch": 0.2867093105899076,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016218750191002675,
+      "loss": 0.0976,
+      "step": 6051
+    },
+    {
+      "epoch": 0.28675669272684196,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001621758385844661,
+      "loss": 0.0691,
+      "step": 6052
+    },
+    {
+      "epoch": 0.28680407486377635,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001621641738799116,
+      "loss": 1.346,
+      "step": 6053
+    },
+    {
+      "epoch": 0.28685145700071074,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00016215250779662193,
+      "loss": 0.1656,
+      "step": 6054
+    },
+    {
+      "epoch": 0.28689883913764513,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001621408403348559,
+      "loss": 0.9364,
+      "step": 6055
+    },
+    {
+      "epoch": 0.28694622127457947,
+      "grad_norm": 0.0032501220703125,
+      "learning_rate": 0.00016212917149487216,
+      "loss": 0.0002,
+      "step": 6056
+    },
+    {
+      "epoch": 0.28699360341151386,
+      "grad_norm": 0.0036468505859375,
+      "learning_rate": 0.00016211750127692956,
+      "loss": 0.0003,
+      "step": 6057
+    },
+    {
+      "epoch": 0.28704098554844826,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016210582968128695,
+      "loss": 0.9902,
+      "step": 6058
+    },
+    {
+      "epoch": 0.2870883676853826,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016209415670820321,
+      "loss": 0.6357,
+      "step": 6059
+    },
+    {
+      "epoch": 0.287135749822317,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016208248235793723,
+      "loss": 1.3255,
+      "step": 6060
+    },
+    {
+      "epoch": 0.2871831319592514,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016207080663074787,
+      "loss": 0.289,
+      "step": 6061
+    },
+    {
+      "epoch": 0.2872305140961857,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001620591295268942,
+      "loss": 0.5338,
+      "step": 6062
+    },
+    {
+      "epoch": 0.2872778962331201,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001620474510466351,
+      "loss": 0.089,
+      "step": 6063
+    },
+    {
+      "epoch": 0.2873252783700545,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001620357711902296,
+      "loss": 0.1438,
+      "step": 6064
+    },
+    {
+      "epoch": 0.28737266050698884,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001620240899579368,
+      "loss": 1.2508,
+      "step": 6065
+    },
+    {
+      "epoch": 0.28742004264392323,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00016201240735001576,
+      "loss": 0.0149,
+      "step": 6066
+    },
+    {
+      "epoch": 0.2874674247808576,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.00016200072336672555,
+      "loss": 0.0066,
+      "step": 6067
+    },
+    {
+      "epoch": 0.28751480691779197,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016198903800832538,
+      "loss": 1.5147,
+      "step": 6068
+    },
+    {
+      "epoch": 0.28756218905472636,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016197735127507435,
+      "loss": 0.776,
+      "step": 6069
+    },
+    {
+      "epoch": 0.28760957119166075,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00016196566316723169,
+      "loss": 0.5439,
+      "step": 6070
+    },
+    {
+      "epoch": 0.28765695332859514,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001619539736850566,
+      "loss": 0.7401,
+      "step": 6071
+    },
+    {
+      "epoch": 0.2877043354655295,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016194228282880838,
+      "loss": 1.199,
+      "step": 6072
+    },
+    {
+      "epoch": 0.2877517176024639,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001619305905987463,
+      "loss": 1.0443,
+      "step": 6073
+    },
+    {
+      "epoch": 0.28779909973939827,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001619188969951297,
+      "loss": 1.0907,
+      "step": 6074
+    },
+    {
+      "epoch": 0.2878464818763326,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016190720201821792,
+      "loss": 0.7844,
+      "step": 6075
+    },
+    {
+      "epoch": 0.287893864013267,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001618955056682703,
+      "loss": 1.1599,
+      "step": 6076
+    },
+    {
+      "epoch": 0.2879412461502014,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016188380794554633,
+      "loss": 1.07,
+      "step": 6077
+    },
+    {
+      "epoch": 0.28798862828713573,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016187210885030542,
+      "loss": 0.3129,
+      "step": 6078
+    },
+    {
+      "epoch": 0.2880360104240701,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016186040838280703,
+      "loss": 1.0747,
+      "step": 6079
+    },
+    {
+      "epoch": 0.2880833925610045,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016184870654331065,
+      "loss": 0.9774,
+      "step": 6080
+    },
+    {
+      "epoch": 0.28813077469793885,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00016183700333207587,
+      "loss": 0.7071,
+      "step": 6081
+    },
+    {
+      "epoch": 0.28817815683487324,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001618252987493622,
+      "loss": 0.4575,
+      "step": 6082
+    },
+    {
+      "epoch": 0.28822553897180764,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016181359279542927,
+      "loss": 0.6988,
+      "step": 6083
+    },
+    {
+      "epoch": 0.28827292110874203,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001618018854705367,
+      "loss": 1.4512,
+      "step": 6084
+    },
+    {
+      "epoch": 0.28832030324567637,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016179017677494414,
+      "loss": 0.4298,
+      "step": 6085
+    },
+    {
+      "epoch": 0.28836768538261076,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016177846670891125,
+      "loss": 0.8053,
+      "step": 6086
+    },
+    {
+      "epoch": 0.28841506751954515,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016176675527269777,
+      "loss": 1.2651,
+      "step": 6087
+    },
+    {
+      "epoch": 0.2884624496564795,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.00016175504246656347,
+      "loss": 0.004,
+      "step": 6088
+    },
+    {
+      "epoch": 0.2885098317934139,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001617433282907681,
+      "loss": 0.7184,
+      "step": 6089
+    },
+    {
+      "epoch": 0.2885572139303483,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00016173161274557146,
+      "loss": 0.0291,
+      "step": 6090
+    },
+    {
+      "epoch": 0.2886045960672826,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00016171989583123341,
+      "loss": 1.0737,
+      "step": 6091
+    },
+    {
+      "epoch": 0.288651978204217,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016170817754801383,
+      "loss": 0.7138,
+      "step": 6092
+    },
+    {
+      "epoch": 0.2886993603411514,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001616964578961726,
+      "loss": 0.7767,
+      "step": 6093
+    },
+    {
+      "epoch": 0.28874674247808574,
+      "grad_norm": 0.0257568359375,
+      "learning_rate": 0.00016168473687596963,
+      "loss": 0.0022,
+      "step": 6094
+    },
+    {
+      "epoch": 0.28879412461502013,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001616730144876649,
+      "loss": 0.6575,
+      "step": 6095
+    },
+    {
+      "epoch": 0.2888415067519545,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016166129073151843,
+      "loss": 0.1054,
+      "step": 6096
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001616495656077902,
+      "loss": 0.9698,
+      "step": 6097
+    },
+    {
+      "epoch": 0.28893627102582325,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016163783911674021,
+      "loss": 1.0205,
+      "step": 6098
+    },
+    {
+      "epoch": 0.28898365316275765,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016162611125862867,
+      "loss": 0.8748,
+      "step": 6099
+    },
+    {
+      "epoch": 0.28903103529969204,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016161438203371562,
+      "loss": 1.3839,
+      "step": 6100
+    },
+    {
+      "epoch": 0.2890784174366264,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001616026514422612,
+      "loss": 0.9761,
+      "step": 6101
+    },
+    {
+      "epoch": 0.28912579957356077,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016159091948452555,
+      "loss": 1.2139,
+      "step": 6102
+    },
+    {
+      "epoch": 0.28917318171049516,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00016157918616076895,
+      "loss": 1.4222,
+      "step": 6103
+    },
+    {
+      "epoch": 0.2892205638474295,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00016156745147125157,
+      "loss": 0.0964,
+      "step": 6104
+    },
+    {
+      "epoch": 0.2892679459843639,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001615557154162337,
+      "loss": 1.0896,
+      "step": 6105
+    },
+    {
+      "epoch": 0.2893153281212983,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016154397799597562,
+      "loss": 0.0228,
+      "step": 6106
+    },
+    {
+      "epoch": 0.2893627102582326,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016153223921073768,
+      "loss": 1.5017,
+      "step": 6107
+    },
+    {
+      "epoch": 0.289410092395167,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001615204990607802,
+      "loss": 0.9594,
+      "step": 6108
+    },
+    {
+      "epoch": 0.2894574745321014,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001615087575463636,
+      "loss": 0.7924,
+      "step": 6109
+    },
+    {
+      "epoch": 0.28950485666903575,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016149701466774827,
+      "loss": 0.9055,
+      "step": 6110
+    },
+    {
+      "epoch": 0.28955223880597014,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016148527042519466,
+      "loss": 0.648,
+      "step": 6111
+    },
+    {
+      "epoch": 0.28959962094290453,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016147352481896322,
+      "loss": 1.648,
+      "step": 6112
+    },
+    {
+      "epoch": 0.2896470030798389,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001614617778493145,
+      "loss": 1.1806,
+      "step": 6113
+    },
+    {
+      "epoch": 0.28969438521677326,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000161450029516509,
+      "loss": 0.5153,
+      "step": 6114
+    },
+    {
+      "epoch": 0.28974176735370766,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001614382798208073,
+      "loss": 1.104,
+      "step": 6115
+    },
+    {
+      "epoch": 0.28978914949064205,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016142652876247,
+      "loss": 0.916,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2898365316275764,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016141477634175772,
+      "loss": 1.2782,
+      "step": 6117
+    },
+    {
+      "epoch": 0.2898839137645108,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016140302255893114,
+      "loss": 1.3098,
+      "step": 6118
+    },
+    {
+      "epoch": 0.2899312959014452,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0001613912674142509,
+      "loss": 0.0161,
+      "step": 6119
+    },
+    {
+      "epoch": 0.2899786780383795,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016137951090797775,
+      "loss": 1.2315,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2900260601753139,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001613677530403724,
+      "loss": 0.7916,
+      "step": 6121
+    },
+    {
+      "epoch": 0.2900734423122483,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00016135599381169566,
+      "loss": 1.5569,
+      "step": 6122
+    },
+    {
+      "epoch": 0.29012082444918263,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016134423322220834,
+      "loss": 0.2116,
+      "step": 6123
+    },
+    {
+      "epoch": 0.29016820658611703,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016133247127217125,
+      "loss": 0.8461,
+      "step": 6124
+    },
+    {
+      "epoch": 0.2902155887230514,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016132070796184532,
+      "loss": 0.9446,
+      "step": 6125
+    },
+    {
+      "epoch": 0.29026297085998576,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016130894329149137,
+      "loss": 1.0857,
+      "step": 6126
+    },
+    {
+      "epoch": 0.29031035299692015,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016129717726137037,
+      "loss": 0.6333,
+      "step": 6127
+    },
+    {
+      "epoch": 0.29035773513385454,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016128540987174326,
+      "loss": 0.0076,
+      "step": 6128
+    },
+    {
+      "epoch": 0.29040511727078894,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00016127364112287107,
+      "loss": 1.0762,
+      "step": 6129
+    },
+    {
+      "epoch": 0.2904524994077233,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016126187101501475,
+      "loss": 1.1474,
+      "step": 6130
+    },
+    {
+      "epoch": 0.29049988154465767,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001612500995484354,
+      "loss": 1.068,
+      "step": 6131
+    },
+    {
+      "epoch": 0.29054726368159206,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016123832672339407,
+      "loss": 0.1338,
+      "step": 6132
+    },
+    {
+      "epoch": 0.2905946458185264,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00016122655254015192,
+      "loss": 0.0309,
+      "step": 6133
+    },
+    {
+      "epoch": 0.2906420279554608,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00016121477699896998,
+      "loss": 0.0521,
+      "step": 6134
+    },
+    {
+      "epoch": 0.2906894100923952,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016120300010010954,
+      "loss": 1.0097,
+      "step": 6135
+    },
+    {
+      "epoch": 0.2907367922293295,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016119122184383175,
+      "loss": 0.8926,
+      "step": 6136
+    },
+    {
+      "epoch": 0.2907841743662639,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001611794422303978,
+      "loss": 1.2017,
+      "step": 6137
+    },
+    {
+      "epoch": 0.2908315565031983,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000161167661260069,
+      "loss": 1.012,
+      "step": 6138
+    },
+    {
+      "epoch": 0.29087893864013264,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00016115587893310665,
+      "loss": 0.0874,
+      "step": 6139
+    },
+    {
+      "epoch": 0.29092632077706704,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00016114409524977202,
+      "loss": 1.116,
+      "step": 6140
+    },
+    {
+      "epoch": 0.29097370291400143,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001611323102103265,
+      "loss": 0.6605,
+      "step": 6141
+    },
+    {
+      "epoch": 0.2910210850509358,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016112052381503147,
+      "loss": 1.1295,
+      "step": 6142
+    },
+    {
+      "epoch": 0.29106846718787016,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001611087360641483,
+      "loss": 0.9839,
+      "step": 6143
+    },
+    {
+      "epoch": 0.29111584932480455,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016109694695793847,
+      "loss": 0.8536,
+      "step": 6144
+    },
+    {
+      "epoch": 0.29116323146173895,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001610851564966634,
+      "loss": 0.048,
+      "step": 6145
+    },
+    {
+      "epoch": 0.2912106135986733,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00016107336468058466,
+      "loss": 0.015,
+      "step": 6146
+    },
+    {
+      "epoch": 0.2912579957356077,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016106157150996375,
+      "loss": 0.9039,
+      "step": 6147
+    },
+    {
+      "epoch": 0.29130537787254207,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0001610497769850622,
+      "loss": 0.2915,
+      "step": 6148
+    },
+    {
+      "epoch": 0.2913527600094764,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016103798110614164,
+      "loss": 0.757,
+      "step": 6149
+    },
+    {
+      "epoch": 0.2914001421464108,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016102618387346367,
+      "loss": 1.1858,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2914475242833452,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00016101438528728993,
+      "loss": 0.0603,
+      "step": 6151
+    },
+    {
+      "epoch": 0.29149490642027953,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016100258534788213,
+      "loss": 0.8641,
+      "step": 6152
+    },
+    {
+      "epoch": 0.2915422885572139,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016099078405550194,
+      "loss": 0.7625,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2915896706941483,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016097898141041115,
+      "loss": 1.667,
+      "step": 6154
+    },
+    {
+      "epoch": 0.29163705283108265,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016096717741287148,
+      "loss": 0.0808,
+      "step": 6155
+    },
+    {
+      "epoch": 0.29168443496801705,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00016095537206314479,
+      "loss": 0.0055,
+      "step": 6156
+    },
+    {
+      "epoch": 0.29173181710495144,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016094356536149284,
+      "loss": 0.2689,
+      "step": 6157
+    },
+    {
+      "epoch": 0.29177919924188583,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016093175730817752,
+      "loss": 0.8178,
+      "step": 6158
+    },
+    {
+      "epoch": 0.29182658137882017,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00016091994790346076,
+      "loss": 0.0233,
+      "step": 6159
+    },
+    {
+      "epoch": 0.29187396351575456,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016090813714760442,
+      "loss": 0.1928,
+      "step": 6160
+    },
+    {
+      "epoch": 0.29192134565268896,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001608963250408705,
+      "loss": 0.1937,
+      "step": 6161
+    },
+    {
+      "epoch": 0.2919687277896233,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001608845115835209,
+      "loss": 0.5616,
+      "step": 6162
+    },
+    {
+      "epoch": 0.2920161099265577,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00016087269677581774,
+      "loss": 0.0031,
+      "step": 6163
+    },
+    {
+      "epoch": 0.2920634920634921,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000160860880618023,
+      "loss": 1.07,
+      "step": 6164
+    },
+    {
+      "epoch": 0.2921108742004264,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016084906311039873,
+      "loss": 1.5735,
+      "step": 6165
+    },
+    {
+      "epoch": 0.2921582563373608,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016083724425320706,
+      "loss": 1.0827,
+      "step": 6166
+    },
+    {
+      "epoch": 0.2922056384742952,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0001608254240467101,
+      "loss": 0.0123,
+      "step": 6167
+    },
+    {
+      "epoch": 0.29225302061122954,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00016081360249117004,
+      "loss": 0.8328,
+      "step": 6168
+    },
+    {
+      "epoch": 0.29230040274816393,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00016080177958684906,
+      "loss": 0.9292,
+      "step": 6169
+    },
+    {
+      "epoch": 0.2923477848850983,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016078995533400935,
+      "loss": 0.1389,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2923951670220327,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001607781297329132,
+      "loss": 0.8424,
+      "step": 6171
+    },
+    {
+      "epoch": 0.29244254915896706,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016076630278382287,
+      "loss": 0.8404,
+      "step": 6172
+    },
+    {
+      "epoch": 0.29248993129590145,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016075447448700067,
+      "loss": 1.0984,
+      "step": 6173
+    },
+    {
+      "epoch": 0.29253731343283584,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016074264484270892,
+      "loss": 0.494,
+      "step": 6174
+    },
+    {
+      "epoch": 0.2925846955697702,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00016073081385121002,
+      "loss": 0.0795,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2926320777067046,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016071898151276637,
+      "loss": 0.9714,
+      "step": 6176
+    },
+    {
+      "epoch": 0.29267945984363897,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016070714782764035,
+      "loss": 0.9679,
+      "step": 6177
+    },
+    {
+      "epoch": 0.2927268419805733,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016069531279609448,
+      "loss": 1.077,
+      "step": 6178
+    },
+    {
+      "epoch": 0.2927742241175077,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001606834764183912,
+      "loss": 0.9103,
+      "step": 6179
+    },
+    {
+      "epoch": 0.2928216062544421,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016067163869479308,
+      "loss": 1.4553,
+      "step": 6180
+    },
+    {
+      "epoch": 0.29286898839137643,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016065979962556263,
+      "loss": 1.0385,
+      "step": 6181
+    },
+    {
+      "epoch": 0.2929163705283108,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016064795921096243,
+      "loss": 0.9758,
+      "step": 6182
+    },
+    {
+      "epoch": 0.2929637526652452,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016063611745125507,
+      "loss": 0.4767,
+      "step": 6183
+    },
+    {
+      "epoch": 0.29301113480217955,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016062427434670329,
+      "loss": 1.1962,
+      "step": 6184
+    },
+    {
+      "epoch": 0.29305851693911394,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001606124298975696,
+      "loss": 1.4655,
+      "step": 6185
+    },
+    {
+      "epoch": 0.29310589907604834,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001606005841041168,
+      "loss": 0.8719,
+      "step": 6186
+    },
+    {
+      "epoch": 0.29315328121298273,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00016058873696660761,
+      "loss": 0.0059,
+      "step": 6187
+    },
+    {
+      "epoch": 0.29320066334991707,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016057688848530475,
+      "loss": 0.9908,
+      "step": 6188
+    },
+    {
+      "epoch": 0.29324804548685146,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016056503866047103,
+      "loss": 1.0236,
+      "step": 6189
+    },
+    {
+      "epoch": 0.29329542762378585,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016055318749236928,
+      "loss": 0.1252,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2933428097607202,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00016054133498126229,
+      "loss": 1.286,
+      "step": 6191
+    },
+    {
+      "epoch": 0.2933901918976546,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00016052948112741302,
+      "loss": 0.0102,
+      "step": 6192
+    },
+    {
+      "epoch": 0.293437574034589,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001605176259310843,
+      "loss": 0.135,
+      "step": 6193
+    },
+    {
+      "epoch": 0.2934849561715233,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016050576939253912,
+      "loss": 1.089,
+      "step": 6194
+    },
+    {
+      "epoch": 0.2935323383084577,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016049391151204043,
+      "loss": 0.9853,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2935797204453921,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00016048205228985117,
+      "loss": 1.1275,
+      "step": 6196
+    },
+    {
+      "epoch": 0.29362710258232644,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00016047019172623448,
+      "loss": 0.0027,
+      "step": 6197
+    },
+    {
+      "epoch": 0.29367448471926083,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001604583298214533,
+      "loss": 0.9025,
+      "step": 6198
+    },
+    {
+      "epoch": 0.2937218668561952,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001604464665757708,
+      "loss": 0.1409,
+      "step": 6199
+    },
+    {
+      "epoch": 0.2937692489931296,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016043460198945,
+      "loss": 0.8453,
+      "step": 6200
+    },
+    {
+      "epoch": 0.29381663113006395,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00016042273606275414,
+      "loss": 1.5202,
+      "step": 6201
+    },
+    {
+      "epoch": 0.29386401326699835,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016041086879594634,
+      "loss": 1.264,
+      "step": 6202
+    },
+    {
+      "epoch": 0.29391139540393274,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001603990001892898,
+      "loss": 0.3838,
+      "step": 6203
+    },
+    {
+      "epoch": 0.2939587775408671,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001603871302430478,
+      "loss": 0.1965,
+      "step": 6204
+    },
+    {
+      "epoch": 0.29400615967780147,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00016037525895748358,
+      "loss": 0.051,
+      "step": 6205
+    },
+    {
+      "epoch": 0.29405354181473586,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001603633863328604,
+      "loss": 0.1407,
+      "step": 6206
+    },
+    {
+      "epoch": 0.2941009239516702,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016035151236944162,
+      "loss": 0.6282,
+      "step": 6207
+    },
+    {
+      "epoch": 0.2941483060886046,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016033963706749057,
+      "loss": 0.3471,
+      "step": 6208
+    },
+    {
+      "epoch": 0.294195688225539,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016032776042727064,
+      "loss": 0.8517,
+      "step": 6209
+    },
+    {
+      "epoch": 0.2942430703624733,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00016031588244904525,
+      "loss": 0.0247,
+      "step": 6210
+    },
+    {
+      "epoch": 0.2942904524994077,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00016030400313307784,
+      "loss": 0.0238,
+      "step": 6211
+    },
+    {
+      "epoch": 0.2943378346363421,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00016029212247963185,
+      "loss": 0.9779,
+      "step": 6212
+    },
+    {
+      "epoch": 0.29438521677327645,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001602802404889708,
+      "loss": 0.1089,
+      "step": 6213
+    },
+    {
+      "epoch": 0.29443259891021084,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016026835716135827,
+      "loss": 1.0606,
+      "step": 6214
+    },
+    {
+      "epoch": 0.29447998104714523,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016025647249705776,
+      "loss": 0.7342,
+      "step": 6215
+    },
+    {
+      "epoch": 0.2945273631840796,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016024458649633282,
+      "loss": 1.044,
+      "step": 6216
+    },
+    {
+      "epoch": 0.29457474532101396,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016023269915944717,
+      "loss": 1.1679,
+      "step": 6217
+    },
+    {
+      "epoch": 0.29462212745794836,
+      "grad_norm": 0.0245361328125,
+      "learning_rate": 0.0001602208104866644,
+      "loss": 0.0018,
+      "step": 6218
+    },
+    {
+      "epoch": 0.29466950959488275,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001602089204782482,
+      "loss": 0.1495,
+      "step": 6219
+    },
+    {
+      "epoch": 0.2947168917318171,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016019702913446226,
+      "loss": 1.0678,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2947642738687515,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016018513645557034,
+      "loss": 0.6447,
+      "step": 6221
+    },
+    {
+      "epoch": 0.2948116560056859,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001601732424418362,
+      "loss": 0.8159,
+      "step": 6222
+    },
+    {
+      "epoch": 0.2948590381426202,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00016016134709352365,
+      "loss": 0.7239,
+      "step": 6223
+    },
+    {
+      "epoch": 0.2949064202795546,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001601494504108965,
+      "loss": 1.0982,
+      "step": 6224
+    },
+    {
+      "epoch": 0.294953802416489,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016013755239421856,
+      "loss": 1.1708,
+      "step": 6225
+    },
+    {
+      "epoch": 0.29500118455342333,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00016012565304375381,
+      "loss": 0.0258,
+      "step": 6226
+    },
+    {
+      "epoch": 0.2950485666903577,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00016011375235976612,
+      "loss": 0.0109,
+      "step": 6227
+    },
+    {
+      "epoch": 0.2950959488272921,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016010185034251944,
+      "loss": 0.8148,
+      "step": 6228
+    },
+    {
+      "epoch": 0.2951433309642265,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016008994699227773,
+      "loss": 0.4518,
+      "step": 6229
+    },
+    {
+      "epoch": 0.29519071310116085,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016007804230930498,
+      "loss": 0.6498,
+      "step": 6230
+    },
+    {
+      "epoch": 0.29523809523809524,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016006613629386527,
+      "loss": 0.9114,
+      "step": 6231
+    },
+    {
+      "epoch": 0.29528547737502964,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016005422894622263,
+      "loss": 0.6616,
+      "step": 6232
+    },
+    {
+      "epoch": 0.295332859511964,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001600423202666411,
+      "loss": 1.1665,
+      "step": 6233
+    },
+    {
+      "epoch": 0.29538024164889837,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016003041025538495,
+      "loss": 0.0348,
+      "step": 6234
+    },
+    {
+      "epoch": 0.29542762378583276,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001600184989127182,
+      "loss": 0.5651,
+      "step": 6235
+    },
+    {
+      "epoch": 0.2954750059227671,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00016000658623890508,
+      "loss": 0.1426,
+      "step": 6236
+    },
+    {
+      "epoch": 0.2955223880597015,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015999467223420979,
+      "loss": 1.3802,
+      "step": 6237
+    },
+    {
+      "epoch": 0.2955697701966359,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015998275689889656,
+      "loss": 0.0291,
+      "step": 6238
+    },
+    {
+      "epoch": 0.2956171523335702,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001599708402332297,
+      "loss": 0.9275,
+      "step": 6239
+    },
+    {
+      "epoch": 0.2956645344705046,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00015995892223747346,
+      "loss": 0.1245,
+      "step": 6240
+    },
+    {
+      "epoch": 0.295711916607439,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001599470029118922,
+      "loss": 0.1849,
+      "step": 6241
+    },
+    {
+      "epoch": 0.29575929874437334,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015993508225675029,
+      "loss": 0.4715,
+      "step": 6242
+    },
+    {
+      "epoch": 0.29580668088130774,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015992316027231204,
+      "loss": 0.1159,
+      "step": 6243
+    },
+    {
+      "epoch": 0.29585406301824213,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00015991123695884197,
+      "loss": 0.0317,
+      "step": 6244
+    },
+    {
+      "epoch": 0.2959014451551765,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015989931231660446,
+      "loss": 0.5017,
+      "step": 6245
+    },
+    {
+      "epoch": 0.29594882729211086,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000159887386345864,
+      "loss": 0.939,
+      "step": 6246
+    },
+    {
+      "epoch": 0.29599620942904525,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015987545904688514,
+      "loss": 1.6365,
+      "step": 6247
+    },
+    {
+      "epoch": 0.29604359156597965,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0001598635304199323,
+      "loss": 0.0365,
+      "step": 6248
+    },
+    {
+      "epoch": 0.296090973702914,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001598516004652702,
+      "loss": 0.0459,
+      "step": 6249
+    },
+    {
+      "epoch": 0.2961383558398484,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001598396691831633,
+      "loss": 0.1533,
+      "step": 6250
+    },
+    {
+      "epoch": 0.29618573797678277,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015982773657387627,
+      "loss": 0.6514,
+      "step": 6251
+    },
+    {
+      "epoch": 0.2962331201137171,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015981580263767383,
+      "loss": 1.3279,
+      "step": 6252
+    },
+    {
+      "epoch": 0.2962805022506515,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015980386737482057,
+      "loss": 1.3035,
+      "step": 6253
+    },
+    {
+      "epoch": 0.2963278843875859,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015979193078558118,
+      "loss": 0.7709,
+      "step": 6254
+    },
+    {
+      "epoch": 0.29637526652452023,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015977999287022053,
+      "loss": 0.9822,
+      "step": 6255
+    },
+    {
+      "epoch": 0.2964226486614546,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015976805362900325,
+      "loss": 0.2037,
+      "step": 6256
+    },
+    {
+      "epoch": 0.296470030798389,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015975611306219423,
+      "loss": 1.2106,
+      "step": 6257
+    },
+    {
+      "epoch": 0.2965174129353234,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001597441711700583,
+      "loss": 0.7787,
+      "step": 6258
+    },
+    {
+      "epoch": 0.29656479507225775,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00015973222795286025,
+      "loss": 1.0039,
+      "step": 6259
+    },
+    {
+      "epoch": 0.29661217720919214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000159720283410865,
+      "loss": 0.9096,
+      "step": 6260
+    },
+    {
+      "epoch": 0.29665955934612653,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015970833754433753,
+      "loss": 0.1478,
+      "step": 6261
+    },
+    {
+      "epoch": 0.29670694148306087,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00015969639035354267,
+      "loss": 0.0838,
+      "step": 6262
+    },
+    {
+      "epoch": 0.29675432361999526,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001596844418387455,
+      "loss": 1.3852,
+      "step": 6263
+    },
+    {
+      "epoch": 0.29680170575692966,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015967249200021094,
+      "loss": 0.9298,
+      "step": 6264
+    },
+    {
+      "epoch": 0.296849087893864,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001596605408382041,
+      "loss": 0.022,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2968964700307984,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015964858835299003,
+      "loss": 0.2375,
+      "step": 6266
+    },
+    {
+      "epoch": 0.2969438521677328,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015963663454483378,
+      "loss": 1.151,
+      "step": 6267
+    },
+    {
+      "epoch": 0.2969912343046671,
+      "grad_norm": 0.02978515625,
+      "learning_rate": 0.00015962467941400048,
+      "loss": 0.0012,
+      "step": 6268
+    },
+    {
+      "epoch": 0.2970386164416015,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001596127229607553,
+      "loss": 1.225,
+      "step": 6269
+    },
+    {
+      "epoch": 0.2970859985785359,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015960076518536345,
+      "loss": 0.9748,
+      "step": 6270
+    },
+    {
+      "epoch": 0.29713338071547024,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001595888060880901,
+      "loss": 1.0452,
+      "step": 6271
+    },
+    {
+      "epoch": 0.29718076285240463,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015957684566920052,
+      "loss": 0.8657,
+      "step": 6272
+    },
+    {
+      "epoch": 0.297228144989339,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015956488392895995,
+      "loss": 0.9778,
+      "step": 6273
+    },
+    {
+      "epoch": 0.2972755271262734,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015955292086763373,
+      "loss": 1.4371,
+      "step": 6274
+    },
+    {
+      "epoch": 0.29732290926320776,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001595409564854871,
+      "loss": 0.6994,
+      "step": 6275
+    },
+    {
+      "epoch": 0.29737029140014215,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001595289907827855,
+      "loss": 0.788,
+      "step": 6276
+    },
+    {
+      "epoch": 0.29741767353707654,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00015951702375979426,
+      "loss": 0.1758,
+      "step": 6277
+    },
+    {
+      "epoch": 0.2974650556740109,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015950505541677886,
+      "loss": 0.1653,
+      "step": 6278
+    },
+    {
+      "epoch": 0.2975124378109453,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00015949308575400473,
+      "loss": 1.3668,
+      "step": 6279
+    },
+    {
+      "epoch": 0.29755981994787967,
+      "grad_norm": 0.0205078125,
+      "learning_rate": 0.0001594811147717373,
+      "loss": 0.0009,
+      "step": 6280
+    },
+    {
+      "epoch": 0.297607202084814,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015946914247024212,
+      "loss": 0.826,
+      "step": 6281
+    },
+    {
+      "epoch": 0.2976545842217484,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001594571688497847,
+      "loss": 0.1798,
+      "step": 6282
+    },
+    {
+      "epoch": 0.2977019663586828,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001594451939106306,
+      "loss": 1.074,
+      "step": 6283
+    },
+    {
+      "epoch": 0.2977493484956171,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015943321765304538,
+      "loss": 0.8172,
+      "step": 6284
+    },
+    {
+      "epoch": 0.2977967306325515,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015942124007729475,
+      "loss": 0.1754,
+      "step": 6285
+    },
+    {
+      "epoch": 0.2978441127694859,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001594092611836443,
+      "loss": 1.321,
+      "step": 6286
+    },
+    {
+      "epoch": 0.2978914949064203,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001593972809723597,
+      "loss": 1.0648,
+      "step": 6287
+    },
+    {
+      "epoch": 0.29793887704335464,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001593852994437067,
+      "loss": 0.0259,
+      "step": 6288
+    },
+    {
+      "epoch": 0.29798625918028904,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015937331659795101,
+      "loss": 1.3141,
+      "step": 6289
+    },
+    {
+      "epoch": 0.29803364131722343,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015936133243535838,
+      "loss": 0.2187,
+      "step": 6290
+    },
+    {
+      "epoch": 0.29808102345415777,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015934934695619463,
+      "loss": 0.1718,
+      "step": 6291
+    },
+    {
+      "epoch": 0.29812840559109216,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00015933736016072558,
+      "loss": 1.254,
+      "step": 6292
+    },
+    {
+      "epoch": 0.29817578772802655,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001593253720492171,
+      "loss": 0.7333,
+      "step": 6293
+    },
+    {
+      "epoch": 0.2982231698649609,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00015931338262193501,
+      "loss": 0.3758,
+      "step": 6294
+    },
+    {
+      "epoch": 0.2982705520018953,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015930139187914533,
+      "loss": 0.9309,
+      "step": 6295
+    },
+    {
+      "epoch": 0.2983179341388297,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001592893998211139,
+      "loss": 1.0219,
+      "step": 6296
+    },
+    {
+      "epoch": 0.298365316275764,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015927740644810677,
+      "loss": 1.0214,
+      "step": 6297
+    },
+    {
+      "epoch": 0.2984126984126984,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001592654117603899,
+      "loss": 0.1297,
+      "step": 6298
+    },
+    {
+      "epoch": 0.2984600805496328,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015925341575822932,
+      "loss": 0.0671,
+      "step": 6299
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015924141844189107,
+      "loss": 0.7515,
+      "step": 6300
+    },
+    {
+      "epoch": 0.29855484482350153,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015922941981164128,
+      "loss": 1.2515,
+      "step": 6301
+    },
+    {
+      "epoch": 0.2986022269604359,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015921741986774603,
+      "loss": 0.9278,
+      "step": 6302
+    },
+    {
+      "epoch": 0.2986496090973703,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001592054186104715,
+      "loss": 1.3154,
+      "step": 6303
+    },
+    {
+      "epoch": 0.29869699123430465,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015919341604008383,
+      "loss": 0.8406,
+      "step": 6304
+    },
+    {
+      "epoch": 0.29874437337123905,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001591814121568493,
+      "loss": 1.1043,
+      "step": 6305
+    },
+    {
+      "epoch": 0.29879175550817344,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000159169406961034,
+      "loss": 0.0178,
+      "step": 6306
+    },
+    {
+      "epoch": 0.2988391376451078,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00015915740045290434,
+      "loss": 0.0587,
+      "step": 6307
+    },
+    {
+      "epoch": 0.29888651978204217,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001591453926327265,
+      "loss": 0.9912,
+      "step": 6308
+    },
+    {
+      "epoch": 0.29893390191897656,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001591333835007669,
+      "loss": 0.0063,
+      "step": 6309
+    },
+    {
+      "epoch": 0.2989812840559109,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015912137305729181,
+      "loss": 1.1696,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2990286661928453,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015910936130256764,
+      "loss": 0.1296,
+      "step": 6311
+    },
+    {
+      "epoch": 0.2990760483297797,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015909734823686081,
+      "loss": 1.0156,
+      "step": 6312
+    },
+    {
+      "epoch": 0.299123430466714,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00015908533386043775,
+      "loss": 0.1437,
+      "step": 6313
+    },
+    {
+      "epoch": 0.2991708126036484,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001590733181735649,
+      "loss": 1.0903,
+      "step": 6314
+    },
+    {
+      "epoch": 0.2992181947405828,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015906130117650878,
+      "loss": 1.1543,
+      "step": 6315
+    },
+    {
+      "epoch": 0.2992655768775172,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00015904928286953593,
+      "loss": 0.0095,
+      "step": 6316
+    },
+    {
+      "epoch": 0.29931295901445154,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015903726325291285,
+      "loss": 0.9702,
+      "step": 6317
+    },
+    {
+      "epoch": 0.29936034115138593,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015902524232690616,
+      "loss": 0.9361,
+      "step": 6318
+    },
+    {
+      "epoch": 0.2994077232883203,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00015901322009178248,
+      "loss": 1.4427,
+      "step": 6319
+    },
+    {
+      "epoch": 0.29945510542525466,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015900119654780842,
+      "loss": 1.0989,
+      "step": 6320
+    },
+    {
+      "epoch": 0.29950248756218906,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001589891716952507,
+      "loss": 0.1506,
+      "step": 6321
+    },
+    {
+      "epoch": 0.29954986969912345,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015897714553437598,
+      "loss": 1.2756,
+      "step": 6322
+    },
+    {
+      "epoch": 0.2995972518360578,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00015896511806545095,
+      "loss": 0.8838,
+      "step": 6323
+    },
+    {
+      "epoch": 0.2996446339729922,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00015895308928874245,
+      "loss": 0.1703,
+      "step": 6324
+    },
+    {
+      "epoch": 0.2996920161099266,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001589410592045172,
+      "loss": 1.1891,
+      "step": 6325
+    },
+    {
+      "epoch": 0.2997393982468609,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00015892902781304203,
+      "loss": 0.0167,
+      "step": 6326
+    },
+    {
+      "epoch": 0.2997867803837953,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015891699511458383,
+      "loss": 1.7041,
+      "step": 6327
+    },
+    {
+      "epoch": 0.2998341625207297,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015890496110940937,
+      "loss": 1.2175,
+      "step": 6328
+    },
+    {
+      "epoch": 0.29988154465766403,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015889292579778568,
+      "loss": 0.6999,
+      "step": 6329
+    },
+    {
+      "epoch": 0.2999289267945984,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015888088917997962,
+      "loss": 0.7467,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2999763089315328,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.00015886885125625813,
+      "loss": 0.0019,
+      "step": 6331
+    },
+    {
+      "epoch": 0.3000236910684672,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001588568120268882,
+      "loss": 1.3148,
+      "step": 6332
+    },
+    {
+      "epoch": 0.30007107320540155,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001588447714921369,
+      "loss": 0.9488,
+      "step": 6333
+    },
+    {
+      "epoch": 0.30011845534233594,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015883272965227125,
+      "loss": 0.1925,
+      "step": 6334
+    },
+    {
+      "epoch": 0.30016583747927034,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015882068650755832,
+      "loss": 0.9735,
+      "step": 6335
+    },
+    {
+      "epoch": 0.3002132196162047,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001588086420582652,
+      "loss": 0.1459,
+      "step": 6336
+    },
+    {
+      "epoch": 0.30026060175313907,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015879659630465905,
+      "loss": 1.1262,
+      "step": 6337
+    },
+    {
+      "epoch": 0.30030798389007346,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00015878454924700706,
+      "loss": 1.0842,
+      "step": 6338
+    },
+    {
+      "epoch": 0.3003553660270078,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00015877250088557635,
+      "loss": 0.9827,
+      "step": 6339
+    },
+    {
+      "epoch": 0.3004027481639422,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015876045122063416,
+      "loss": 1.1656,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3004501303008766,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00015874840025244775,
+      "loss": 0.013,
+      "step": 6341
+    },
+    {
+      "epoch": 0.3004975124378109,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015873634798128442,
+      "loss": 0.7893,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3005448945747453,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015872429440741143,
+      "loss": 0.7876,
+      "step": 6343
+    },
+    {
+      "epoch": 0.3005922767116797,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015871223953109616,
+      "loss": 0.8104,
+      "step": 6344
+    },
+    {
+      "epoch": 0.3006396588486141,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015870018335260597,
+      "loss": 0.6066,
+      "step": 6345
+    },
+    {
+      "epoch": 0.30068704098554844,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001586881258722082,
+      "loss": 0.7675,
+      "step": 6346
+    },
+    {
+      "epoch": 0.30073442312248283,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015867606709017032,
+      "loss": 0.0975,
+      "step": 6347
+    },
+    {
+      "epoch": 0.3007818052594172,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00015866400700675977,
+      "loss": 0.0579,
+      "step": 6348
+    },
+    {
+      "epoch": 0.30082918739635156,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015865194562224402,
+      "loss": 0.0256,
+      "step": 6349
+    },
+    {
+      "epoch": 0.30087656953328595,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015863988293689062,
+      "loss": 0.5278,
+      "step": 6350
+    },
+    {
+      "epoch": 0.30092395167022035,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015862781895096707,
+      "loss": 1.3326,
+      "step": 6351
+    },
+    {
+      "epoch": 0.3009713338071547,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015861575366474094,
+      "loss": 1.1228,
+      "step": 6352
+    },
+    {
+      "epoch": 0.3010187159440891,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015860368707847983,
+      "loss": 0.8816,
+      "step": 6353
+    },
+    {
+      "epoch": 0.30106609808102347,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015859161919245133,
+      "loss": 0.7351,
+      "step": 6354
+    },
+    {
+      "epoch": 0.3011134802179578,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00015857955000692317,
+      "loss": 0.0034,
+      "step": 6355
+    },
+    {
+      "epoch": 0.3011608623548922,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015856747952216297,
+      "loss": 0.9956,
+      "step": 6356
+    },
+    {
+      "epoch": 0.3012082444918266,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015855540773843845,
+      "loss": 0.1596,
+      "step": 6357
+    },
+    {
+      "epoch": 0.30125562662876093,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00015854333465601736,
+      "loss": 0.1357,
+      "step": 6358
+    },
+    {
+      "epoch": 0.3013030087656953,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001585312602751675,
+      "loss": 0.7673,
+      "step": 6359
+    },
+    {
+      "epoch": 0.3013503909026297,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001585191845961566,
+      "loss": 1.2014,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3013977730395641,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015850710761925252,
+      "loss": 1.1184,
+      "step": 6361
+    },
+    {
+      "epoch": 0.30144515517649845,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001584950293447231,
+      "loss": 0.7689,
+      "step": 6362
+    },
+    {
+      "epoch": 0.30149253731343284,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00015848294977283624,
+      "loss": 1.2459,
+      "step": 6363
+    },
+    {
+      "epoch": 0.30153991945036723,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015847086890385988,
+      "loss": 0.8124,
+      "step": 6364
+    },
+    {
+      "epoch": 0.30158730158730157,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015845878673806187,
+      "loss": 0.1512,
+      "step": 6365
+    },
+    {
+      "epoch": 0.30163468372423596,
+      "grad_norm": 0.00567626953125,
+      "learning_rate": 0.00015844670327571029,
+      "loss": 0.0003,
+      "step": 6366
+    },
+    {
+      "epoch": 0.30168206586117036,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015843461851707303,
+      "loss": 0.0213,
+      "step": 6367
+    },
+    {
+      "epoch": 0.3017294479981047,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015842253246241823,
+      "loss": 1.3032,
+      "step": 6368
+    },
+    {
+      "epoch": 0.3017768301350391,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015841044511201387,
+      "loss": 1.1587,
+      "step": 6369
+    },
+    {
+      "epoch": 0.3018242122719735,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00015839835646612804,
+      "loss": 0.1073,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3018715944089078,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015838626652502888,
+      "loss": 0.0134,
+      "step": 6371
+    },
+    {
+      "epoch": 0.3019189765458422,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001583741752889845,
+      "loss": 0.6428,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3019663586827766,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015836208275826313,
+      "loss": 0.8704,
+      "step": 6373
+    },
+    {
+      "epoch": 0.302013740819711,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00015834998893313288,
+      "loss": 0.1387,
+      "step": 6374
+    },
+    {
+      "epoch": 0.30206112295664533,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015833789381386205,
+      "loss": 0.673,
+      "step": 6375
+    },
+    {
+      "epoch": 0.3021085050935797,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001583257974007189,
+      "loss": 0.9997,
+      "step": 6376
+    },
+    {
+      "epoch": 0.3021558872305141,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015831369969397164,
+      "loss": 1.2411,
+      "step": 6377
+    },
+    {
+      "epoch": 0.30220326936744846,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015830160069388866,
+      "loss": 0.7677,
+      "step": 6378
+    },
+    {
+      "epoch": 0.30225065150438285,
+      "grad_norm": 0.0257568359375,
+      "learning_rate": 0.00015828950040073828,
+      "loss": 0.0027,
+      "step": 6379
+    },
+    {
+      "epoch": 0.30229803364131724,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015827739881478888,
+      "loss": 0.0299,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3023454157782516,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00015826529593630881,
+      "loss": 1.3019,
+      "step": 6381
+    },
+    {
+      "epoch": 0.302392797915186,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015825319176556658,
+      "loss": 0.8761,
+      "step": 6382
+    },
+    {
+      "epoch": 0.30244018005212037,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015824108630283057,
+      "loss": 1.1102,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3024875621890547,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015822897954836931,
+      "loss": 1.1638,
+      "step": 6384
+    },
+    {
+      "epoch": 0.3025349443259891,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015821687150245132,
+      "loss": 0.7149,
+      "step": 6385
+    },
+    {
+      "epoch": 0.3025823264629235,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001582047621653451,
+      "loss": 1.0962,
+      "step": 6386
+    },
+    {
+      "epoch": 0.3026297085998578,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001581926515373193,
+      "loss": 1.4139,
+      "step": 6387
+    },
+    {
+      "epoch": 0.3026770907367922,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00015818053961864242,
+      "loss": 0.0104,
+      "step": 6388
+    },
+    {
+      "epoch": 0.3027244728737266,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015816842640958321,
+      "loss": 0.7952,
+      "step": 6389
+    },
+    {
+      "epoch": 0.302771855010661,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001581563119104102,
+      "loss": 1.1018,
+      "step": 6390
+    },
+    {
+      "epoch": 0.30281923714759534,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015814419612139214,
+      "loss": 0.1604,
+      "step": 6391
+    },
+    {
+      "epoch": 0.30286661928452974,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001581320790427978,
+      "loss": 0.9539,
+      "step": 6392
+    },
+    {
+      "epoch": 0.30291400142146413,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001581199606748958,
+      "loss": 0.7522,
+      "step": 6393
+    },
+    {
+      "epoch": 0.30296138355839847,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000158107841017955,
+      "loss": 0.8823,
+      "step": 6394
+    },
+    {
+      "epoch": 0.30300876569533286,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001580957200722442,
+      "loss": 0.034,
+      "step": 6395
+    },
+    {
+      "epoch": 0.30305614783226725,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001580835978380322,
+      "loss": 0.15,
+      "step": 6396
+    },
+    {
+      "epoch": 0.3031035299692016,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00015807147431558786,
+      "loss": 0.1194,
+      "step": 6397
+    },
+    {
+      "epoch": 0.303150912106136,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015805934950518006,
+      "loss": 1.0244,
+      "step": 6398
+    },
+    {
+      "epoch": 0.3031982942430704,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00015804722340707778,
+      "loss": 0.0147,
+      "step": 6399
+    },
+    {
+      "epoch": 0.3032456763800047,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015803509602154984,
+      "loss": 0.7737,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3032930585169391,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015802296734886536,
+      "loss": 0.5766,
+      "step": 6401
+    },
+    {
+      "epoch": 0.3033404406538735,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015801083738929323,
+      "loss": 0.854,
+      "step": 6402
+    },
+    {
+      "epoch": 0.3033878227908079,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00015799870614310253,
+      "loss": 0.1264,
+      "step": 6403
+    },
+    {
+      "epoch": 0.30343520492774223,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015798657361056228,
+      "loss": 1.1678,
+      "step": 6404
+    },
+    {
+      "epoch": 0.3034825870646766,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001579744397919416,
+      "loss": 0.1216,
+      "step": 6405
+    },
+    {
+      "epoch": 0.303529969201611,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015796230468750961,
+      "loss": 1.2662,
+      "step": 6406
+    },
+    {
+      "epoch": 0.30357735133854535,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015795016829753546,
+      "loss": 0.1355,
+      "step": 6407
+    },
+    {
+      "epoch": 0.30362473347547975,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00015793803062228828,
+      "loss": 0.0243,
+      "step": 6408
+    },
+    {
+      "epoch": 0.30367211561241414,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015792589166203729,
+      "loss": 1.1711,
+      "step": 6409
+    },
+    {
+      "epoch": 0.3037194977493485,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001579137514170517,
+      "loss": 1.1089,
+      "step": 6410
+    },
+    {
+      "epoch": 0.30376687988628287,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015790160988760082,
+      "loss": 0.8376,
+      "step": 6411
+    },
+    {
+      "epoch": 0.30381426202321726,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001578894670739539,
+      "loss": 1.0766,
+      "step": 6412
+    },
+    {
+      "epoch": 0.3038616441601516,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00015787732297638027,
+      "loss": 1.198,
+      "step": 6413
+    },
+    {
+      "epoch": 0.303909026297086,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015786517759514926,
+      "loss": 0.8732,
+      "step": 6414
+    },
+    {
+      "epoch": 0.3039564084340204,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015785303093053025,
+      "loss": 0.0758,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3040037905709547,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00015784088298279264,
+      "loss": 0.0796,
+      "step": 6416
+    },
+    {
+      "epoch": 0.3040511727078891,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015782873375220583,
+      "loss": 1.0285,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3040985548448235,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001578165832390393,
+      "loss": 0.0299,
+      "step": 6418
+    },
+    {
+      "epoch": 0.3041459369817579,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001578044314435626,
+      "loss": 1.039,
+      "step": 6419
+    },
+    {
+      "epoch": 0.30419331911869224,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001577922783660451,
+      "loss": 1.2203,
+      "step": 6420
+    },
+    {
+      "epoch": 0.30424070125562663,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015778012400675647,
+      "loss": 1.0667,
+      "step": 6421
+    },
+    {
+      "epoch": 0.304288083392561,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001577679683659662,
+      "loss": 0.1458,
+      "step": 6422
+    },
+    {
+      "epoch": 0.30433546552949536,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015775581144394395,
+      "loss": 0.0251,
+      "step": 6423
+    },
+    {
+      "epoch": 0.30438284766642976,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015774365324095936,
+      "loss": 1.1773,
+      "step": 6424
+    },
+    {
+      "epoch": 0.30443022980336415,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015773149375728198,
+      "loss": 0.9736,
+      "step": 6425
+    },
+    {
+      "epoch": 0.3044776119402985,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001577193329931816,
+      "loss": 0.0459,
+      "step": 6426
+    },
+    {
+      "epoch": 0.3045249940772329,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00015770717094892785,
+      "loss": 0.9453,
+      "step": 6427
+    },
+    {
+      "epoch": 0.3045723762141673,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015769500762479054,
+      "loss": 0.7941,
+      "step": 6428
+    },
+    {
+      "epoch": 0.3046197583511016,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00015768284302103945,
+      "loss": 0.1327,
+      "step": 6429
+    },
+    {
+      "epoch": 0.304667140488036,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001576706771379443,
+      "loss": 0.1174,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3047145226249704,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.000157658509975775,
+      "loss": 1.1515,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00015764634153480134,
+      "loss": 0.0068,
+      "step": 6432
+    },
+    {
+      "epoch": 0.3048092868988391,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001576341718152932,
+      "loss": 0.9539,
+      "step": 6433
+    },
+    {
+      "epoch": 0.3048566690357735,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015762200081752053,
+      "loss": 0.8338,
+      "step": 6434
+    },
+    {
+      "epoch": 0.3049040511727079,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001576098285417533,
+      "loss": 0.8848,
+      "step": 6435
+    },
+    {
+      "epoch": 0.30495143330964225,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015759765498826137,
+      "loss": 1.0667,
+      "step": 6436
+    },
+    {
+      "epoch": 0.30499881544657664,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015758548015731486,
+      "loss": 0.9937,
+      "step": 6437
+    },
+    {
+      "epoch": 0.30504619758351104,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015757330404918372,
+      "loss": 1.1364,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3050935797204454,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000157561126664138,
+      "loss": 1.0133,
+      "step": 6439
+    },
+    {
+      "epoch": 0.30514096185737977,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001575489480024478,
+      "loss": 0.562,
+      "step": 6440
+    },
+    {
+      "epoch": 0.30518834399431416,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015753676806438328,
+      "loss": 0.02,
+      "step": 6441
+    },
+    {
+      "epoch": 0.3052357261312485,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015752458685021448,
+      "loss": 0.138,
+      "step": 6442
+    },
+    {
+      "epoch": 0.3052831082681829,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015751240436021163,
+      "loss": 0.7693,
+      "step": 6443
+    },
+    {
+      "epoch": 0.3053304904051173,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00015750022059464493,
+      "loss": 0.0042,
+      "step": 6444
+    },
+    {
+      "epoch": 0.3053778725420516,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0001574880355537846,
+      "loss": 0.003,
+      "step": 6445
+    },
+    {
+      "epoch": 0.305425254678986,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001574758492379008,
+      "loss": 0.8774,
+      "step": 6446
+    },
+    {
+      "epoch": 0.3054726368159204,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015746366164726396,
+      "loss": 1.1418,
+      "step": 6447
+    },
+    {
+      "epoch": 0.3055200189528548,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015745147278214427,
+      "loss": 1.227,
+      "step": 6448
+    },
+    {
+      "epoch": 0.30556740108978914,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015743928264281212,
+      "loss": 0.0743,
+      "step": 6449
+    },
+    {
+      "epoch": 0.30561478322672353,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015742709122953786,
+      "loss": 0.9859,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3056621653636579,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00015741489854259187,
+      "loss": 1.0521,
+      "step": 6451
+    },
+    {
+      "epoch": 0.30570954750059226,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001574027045822446,
+      "loss": 0.7161,
+      "step": 6452
+    },
+    {
+      "epoch": 0.30575692963752665,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015739050934876652,
+      "loss": 0.6875,
+      "step": 6453
+    },
+    {
+      "epoch": 0.30580431177446105,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015737831284242803,
+      "loss": 0.7807,
+      "step": 6454
+    },
+    {
+      "epoch": 0.3058516939113954,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001573661150634997,
+      "loss": 1.3163,
+      "step": 6455
+    },
+    {
+      "epoch": 0.3058990760483298,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015735391601225202,
+      "loss": 0.319,
+      "step": 6456
+    },
+    {
+      "epoch": 0.30594645818526417,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 0.00015734171568895558,
+      "loss": 0.0014,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3059938403221985,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015732951409388096,
+      "loss": 1.143,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3060412224591329,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015731731122729881,
+      "loss": 0.2536,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3060886045960673,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00015730510708947973,
+      "loss": 0.1477,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3061359867330017,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001572929016806944,
+      "loss": 1.3176,
+      "step": 6461
+    },
+    {
+      "epoch": 0.306183368869936,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015728069500121357,
+      "loss": 0.9559,
+      "step": 6462
+    },
+    {
+      "epoch": 0.3062307510068704,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015726848705130788,
+      "loss": 0.9945,
+      "step": 6463
+    },
+    {
+      "epoch": 0.3062781331438048,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001572562778312482,
+      "loss": 0.1301,
+      "step": 6464
+    },
+    {
+      "epoch": 0.30632551528073915,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00015724406734130524,
+      "loss": 0.0058,
+      "step": 6465
+    },
+    {
+      "epoch": 0.30637289741767354,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00015723185558174984,
+      "loss": 0.1685,
+      "step": 6466
+    },
+    {
+      "epoch": 0.30642027955460793,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015721964255285283,
+      "loss": 0.7259,
+      "step": 6467
+    },
+    {
+      "epoch": 0.30646766169154227,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015720742825488511,
+      "loss": 1.2139,
+      "step": 6468
+    },
+    {
+      "epoch": 0.30651504382847666,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015719521268811758,
+      "loss": 0.6002,
+      "step": 6469
+    },
+    {
+      "epoch": 0.30656242596541106,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015718299585282117,
+      "loss": 0.5995,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3066098081023454,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001571707777492668,
+      "loss": 0.0112,
+      "step": 6471
+    },
+    {
+      "epoch": 0.3066571902392798,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015715855837772547,
+      "loss": 0.9161,
+      "step": 6472
+    },
+    {
+      "epoch": 0.3067045723762142,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015714633773846818,
+      "loss": 0.8095,
+      "step": 6473
+    },
+    {
+      "epoch": 0.3067519545131485,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015713411583176602,
+      "loss": 0.1901,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3067993366500829,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015712189265789002,
+      "loss": 1.2767,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3068467187870173,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015710966821711129,
+      "loss": 1.4565,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3068941009239517,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015709744250970092,
+      "loss": 1.0918,
+      "step": 6477
+    },
+    {
+      "epoch": 0.30694148306088603,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015708521553593012,
+      "loss": 1.4035,
+      "step": 6478
+    },
+    {
+      "epoch": 0.3069888651978204,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015707298729607003,
+      "loss": 0.253,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3070362473347548,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001570607577903919,
+      "loss": 0.7883,
+      "step": 6480
+    },
+    {
+      "epoch": 0.30708362947168916,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001570485270191669,
+      "loss": 0.7736,
+      "step": 6481
+    },
+    {
+      "epoch": 0.30713101160862355,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00015703629498266635,
+      "loss": 0.0134,
+      "step": 6482
+    },
+    {
+      "epoch": 0.30717839374555794,
+      "grad_norm": 0.001922607421875,
+      "learning_rate": 0.00015702406168116153,
+      "loss": 0.0002,
+      "step": 6483
+    },
+    {
+      "epoch": 0.3072257758824923,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015701182711492378,
+      "loss": 1.0878,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3072731580194267,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00015699959128422442,
+      "loss": 0.109,
+      "step": 6485
+    },
+    {
+      "epoch": 0.30732054015636107,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00015698735418933482,
+      "loss": 1.3013,
+      "step": 6486
+    },
+    {
+      "epoch": 0.3073679222932954,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001569751158305264,
+      "loss": 1.0957,
+      "step": 6487
+    },
+    {
+      "epoch": 0.3074153044302298,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015696287620807064,
+      "loss": 1.2234,
+      "step": 6488
+    },
+    {
+      "epoch": 0.3074626865671642,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015695063532223896,
+      "loss": 0.9033,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3075100687040986,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001569383931733028,
+      "loss": 0.7429,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3075574508410329,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015692614976153374,
+      "loss": 1.1825,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3076048329779673,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015691390508720335,
+      "loss": 0.678,
+      "step": 6492
+    },
+    {
+      "epoch": 0.3076522151149017,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015690165915058314,
+      "loss": 1.0472,
+      "step": 6493
+    },
+    {
+      "epoch": 0.30769959725183604,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015688941195194476,
+      "loss": 0.2872,
+      "step": 6494
+    },
+    {
+      "epoch": 0.30774697938877044,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001568771634915598,
+      "loss": 1.4657,
+      "step": 6495
+    },
+    {
+      "epoch": 0.30779436152570483,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015686491376969993,
+      "loss": 0.021,
+      "step": 6496
+    },
+    {
+      "epoch": 0.30784174366263917,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015685266278663685,
+      "loss": 1.0232,
+      "step": 6497
+    },
+    {
+      "epoch": 0.30788912579957356,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001568404105426423,
+      "loss": 1.1927,
+      "step": 6498
+    },
+    {
+      "epoch": 0.30793650793650795,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00015682815703798792,
+      "loss": 0.3838,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3079838900734423,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00015681590227294558,
+      "loss": 0.0158,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3080312722103767,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015680364624778704,
+      "loss": 1.0193,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3080786543473111,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015679138896278417,
+      "loss": 1.3357,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3081260364842454,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015677913041820874,
+      "loss": 0.8807,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3081734186211798,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015676687061433268,
+      "loss": 1.2676,
+      "step": 6504
+    },
+    {
+      "epoch": 0.3082208007581142,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001567546095514279,
+      "loss": 0.62,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3082681828950486,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015674234722976634,
+      "loss": 1.2959,
+      "step": 6506
+    },
+    {
+      "epoch": 0.30831556503198293,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015673008364962,
+      "loss": 0.6548,
+      "step": 6507
+    },
+    {
+      "epoch": 0.3083629471689173,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00015671781881126075,
+      "loss": 0.1654,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3084103293058517,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015670555271496075,
+      "loss": 0.1376,
+      "step": 6509
+    },
+    {
+      "epoch": 0.30845771144278605,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015669328536099198,
+      "loss": 0.6347,
+      "step": 6510
+    },
+    {
+      "epoch": 0.30850509357972045,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015668101674962647,
+      "loss": 0.9348,
+      "step": 6511
+    },
+    {
+      "epoch": 0.30855247571665484,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00015666874688113644,
+      "loss": 0.1464,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3085998578535892,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015665647575579397,
+      "loss": 0.6497,
+      "step": 6513
+    },
+    {
+      "epoch": 0.30864723999052357,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015664420337387118,
+      "loss": 0.9218,
+      "step": 6514
+    },
+    {
+      "epoch": 0.30869462212745796,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015663192973564032,
+      "loss": 0.7704,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3087420042643923,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015661965484137355,
+      "loss": 0.7489,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3087893864013267,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015660737869134322,
+      "loss": 0.5298,
+      "step": 6517
+    },
+    {
+      "epoch": 0.3088367685382611,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015659510128582143,
+      "loss": 0.77,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3088841506751954,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015658282262508063,
+      "loss": 1.2064,
+      "step": 6519
+    },
+    {
+      "epoch": 0.3089315328121298,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015657054270939308,
+      "loss": 0.5916,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3089789149490642,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 0.00015655826153903118,
+      "loss": 0.002,
+      "step": 6521
+    },
+    {
+      "epoch": 0.3090262970859986,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015654597911426722,
+      "loss": 0.9681,
+      "step": 6522
+    },
+    {
+      "epoch": 0.30907367922293294,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015653369543537372,
+      "loss": 0.4623,
+      "step": 6523
+    },
+    {
+      "epoch": 0.30912106135986733,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015652141050262307,
+      "loss": 0.4973,
+      "step": 6524
+    },
+    {
+      "epoch": 0.3091684434968017,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015650912431628772,
+      "loss": 0.2051,
+      "step": 6525
+    },
+    {
+      "epoch": 0.30921582563373606,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015649683687664017,
+      "loss": 1.2065,
+      "step": 6526
+    },
+    {
+      "epoch": 0.30926320777067046,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00015648454818395298,
+      "loss": 0.0663,
+      "step": 6527
+    },
+    {
+      "epoch": 0.30931058990760485,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015647225823849868,
+      "loss": 0.8263,
+      "step": 6528
+    },
+    {
+      "epoch": 0.3093579720445392,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015645996704054984,
+      "loss": 1.0052,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3094053541814736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015644767459037907,
+      "loss": 0.8184,
+      "step": 6530
+    },
+    {
+      "epoch": 0.309452736318408,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000156435380888259,
+      "loss": 0.9012,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3095001184553423,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015642308593446228,
+      "loss": 0.3063,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3095475005922767,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00015641078972926164,
+      "loss": 0.1363,
+      "step": 6533
+    },
+    {
+      "epoch": 0.3095948827292111,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00015639849227292975,
+      "loss": 0.5949,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3096422648661455,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001563861935657394,
+      "loss": 0.7603,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3096896470030798,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001563738936079633,
+      "loss": 0.9269,
+      "step": 6536
+    },
+    {
+      "epoch": 0.3097370291400142,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001563615923998743,
+      "loss": 0.9251,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3097844112769486,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00015634928994174522,
+      "loss": 0.0117,
+      "step": 6538
+    },
+    {
+      "epoch": 0.30983179341388295,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001563369862338489,
+      "loss": 0.9585,
+      "step": 6539
+    },
+    {
+      "epoch": 0.30987917555081734,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015632468127645826,
+      "loss": 1.1298,
+      "step": 6540
+    },
+    {
+      "epoch": 0.30992655768775174,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015631237506984617,
+      "loss": 1.4214,
+      "step": 6541
+    },
+    {
+      "epoch": 0.3099739398246861,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015630006761428557,
+      "loss": 1.0516,
+      "step": 6542
+    },
+    {
+      "epoch": 0.31002132196162047,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015628775891004946,
+      "loss": 0.86,
+      "step": 6543
+    },
+    {
+      "epoch": 0.31006870409855486,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015627544895741076,
+      "loss": 0.4261,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3101160862354892,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001562631377566426,
+      "loss": 1.2106,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3101634683724236,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015625082530801795,
+      "loss": 1.0549,
+      "step": 6546
+    },
+    {
+      "epoch": 0.310210850509358,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015623851161180991,
+      "loss": 0.257,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3102582326462923,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001562261966682916,
+      "loss": 0.5688,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3103056147832267,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015621388047773612,
+      "loss": 0.0285,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3103529969201611,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015620156304041666,
+      "loss": 0.621,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3104003790570955,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001561892443566064,
+      "loss": 0.998,
+      "step": 6551
+    },
+    {
+      "epoch": 0.31044776119402984,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00015617692442657853,
+      "loss": 0.03,
+      "step": 6552
+    },
+    {
+      "epoch": 0.31049514333096423,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015616460325060635,
+      "loss": 0.7481,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3105425254678986,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00015615228082896306,
+      "loss": 0.8258,
+      "step": 6554
+    },
+    {
+      "epoch": 0.31058990760483296,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015613995716192198,
+      "loss": 0.974,
+      "step": 6555
+    },
+    {
+      "epoch": 0.31063728974176735,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015612763224975647,
+      "loss": 0.8936,
+      "step": 6556
+    },
+    {
+      "epoch": 0.31068467187870175,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 0.00015611530609273985,
+      "loss": 0.0016,
+      "step": 6557
+    },
+    {
+      "epoch": 0.3107320540156361,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015610297869114552,
+      "loss": 0.7345,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3107794361525705,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0001560906500452469,
+      "loss": 0.0393,
+      "step": 6559
+    },
+    {
+      "epoch": 0.31082681828950487,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015607832015531736,
+      "loss": 0.1445,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3108742004264392,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015606598902163045,
+      "loss": 0.6681,
+      "step": 6561
+    },
+    {
+      "epoch": 0.3109215825633736,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001560536566444596,
+      "loss": 0.5334,
+      "step": 6562
+    },
+    {
+      "epoch": 0.310968964700308,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015604132302407837,
+      "loss": 1.5135,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3110163468372424,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001560289881607603,
+      "loss": 0.4288,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3110637289741767,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001560166520547789,
+      "loss": 0.4492,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00015600431470640786,
+      "loss": 1.3553,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3111584932480455,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001559919761159208,
+      "loss": 0.6518,
+      "step": 6567
+    },
+    {
+      "epoch": 0.31120587538497985,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001559796362835913,
+      "loss": 1.0372,
+      "step": 6568
+    },
+    {
+      "epoch": 0.31125325752191424,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001559672952096931,
+      "loss": 0.0377,
+      "step": 6569
+    },
+    {
+      "epoch": 0.31130063965884863,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015595495289449994,
+      "loss": 0.8932,
+      "step": 6570
+    },
+    {
+      "epoch": 0.31134802179578297,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001559426093382855,
+      "loss": 1.586,
+      "step": 6571
+    },
+    {
+      "epoch": 0.31139540393271736,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015593026454132359,
+      "loss": 0.9798,
+      "step": 6572
+    },
+    {
+      "epoch": 0.31144278606965176,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015591791850388798,
+      "loss": 0.7761,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3114901682065861,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015590557122625246,
+      "loss": 0.9733,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3115375503435205,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015589322270869095,
+      "loss": 0.6146,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3115849324804549,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001558808729514773,
+      "loss": 1.2899,
+      "step": 6576
+    },
+    {
+      "epoch": 0.3116323146173892,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001558685219548854,
+      "loss": 1.2373,
+      "step": 6577
+    },
+    {
+      "epoch": 0.3116796967543236,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001558561697191892,
+      "loss": 1.3932,
+      "step": 6578
+    },
+    {
+      "epoch": 0.311727078891258,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015584381624466265,
+      "loss": 0.5996,
+      "step": 6579
+    },
+    {
+      "epoch": 0.3117744610281924,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015583146153157974,
+      "loss": 0.6589,
+      "step": 6580
+    },
+    {
+      "epoch": 0.31182184316512673,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015581910558021446,
+      "loss": 1.0168,
+      "step": 6581
+    },
+    {
+      "epoch": 0.3118692253020611,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001558067483908409,
+      "loss": 0.1408,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3119166074389955,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001557943899637331,
+      "loss": 0.093,
+      "step": 6583
+    },
+    {
+      "epoch": 0.31196398957592986,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015578203029916515,
+      "loss": 0.722,
+      "step": 6584
+    },
+    {
+      "epoch": 0.31201137171286425,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001557696693974112,
+      "loss": 0.7425,
+      "step": 6585
+    },
+    {
+      "epoch": 0.31205875384979864,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015575730725874535,
+      "loss": 1.0831,
+      "step": 6586
+    },
+    {
+      "epoch": 0.312106135986733,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015574494388344182,
+      "loss": 1.0582,
+      "step": 6587
+    },
+    {
+      "epoch": 0.3121535181236674,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001557325792717749,
+      "loss": 1.2879,
+      "step": 6588
+    },
+    {
+      "epoch": 0.31220090026060177,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00015572021342401864,
+      "loss": 0.1396,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3122482823975361,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00015570784634044742,
+      "loss": 0.2146,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3122956645344705,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0001556954780213355,
+      "loss": 0.0209,
+      "step": 6591
+    },
+    {
+      "epoch": 0.3123430466714049,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00015568310846695722,
+      "loss": 0.9349,
+      "step": 6592
+    },
+    {
+      "epoch": 0.3123904288083393,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001556707376775869,
+      "loss": 1.2064,
+      "step": 6593
+    },
+    {
+      "epoch": 0.3124378109452736,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015565836565349889,
+      "loss": 0.8765,
+      "step": 6594
+    },
+    {
+      "epoch": 0.312485193082208,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015564599239496765,
+      "loss": 1.4647,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3125325752191424,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001556336179022676,
+      "loss": 0.0322,
+      "step": 6596
+    },
+    {
+      "epoch": 0.31257995735607674,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001556212421756731,
+      "loss": 0.7906,
+      "step": 6597
+    },
+    {
+      "epoch": 0.31262733949301114,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015560886521545866,
+      "loss": 1.4391,
+      "step": 6598
+    },
+    {
+      "epoch": 0.31267472162994553,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015559648702189892,
+      "loss": 0.1864,
+      "step": 6599
+    },
+    {
+      "epoch": 0.31272210376687987,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00015558410759526826,
+      "loss": 0.0391,
+      "step": 6600
+    },
+    {
+      "epoch": 0.31276948590381426,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001555717269358413,
+      "loss": 0.9505,
+      "step": 6601
+    },
+    {
+      "epoch": 0.31281686804074865,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015555934504389262,
+      "loss": 0.1565,
+      "step": 6602
+    },
+    {
+      "epoch": 0.312864250177683,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001555469619196969,
+      "loss": 0.616,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3129116323146174,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001555345775635287,
+      "loss": 1.1719,
+      "step": 6604
+    },
+    {
+      "epoch": 0.3129590144515518,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00015552219197566272,
+      "loss": 2.274,
+      "step": 6605
+    },
+    {
+      "epoch": 0.3130063965884861,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015550980515637367,
+      "loss": 1.172,
+      "step": 6606
+    },
+    {
+      "epoch": 0.3130537787254205,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00015549741710593624,
+      "loss": 0.0183,
+      "step": 6607
+    },
+    {
+      "epoch": 0.3131011608623549,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015548502782462522,
+      "loss": 0.0319,
+      "step": 6608
+    },
+    {
+      "epoch": 0.3131485429992893,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001554726373127154,
+      "loss": 1.1681,
+      "step": 6609
+    },
+    {
+      "epoch": 0.31319592513622363,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015546024557048154,
+      "loss": 1.3657,
+      "step": 6610
+    },
+    {
+      "epoch": 0.313243307273158,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015544785259819855,
+      "loss": 1.2749,
+      "step": 6611
+    },
+    {
+      "epoch": 0.3132906894100924,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015543545839614121,
+      "loss": 0.6629,
+      "step": 6612
+    },
+    {
+      "epoch": 0.31333807154702675,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001554230629645845,
+      "loss": 1.1962,
+      "step": 6613
+    },
+    {
+      "epoch": 0.31338545368396115,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00015541066630380326,
+      "loss": 0.0403,
+      "step": 6614
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015539826841407247,
+      "loss": 0.1714,
+      "step": 6615
+    },
+    {
+      "epoch": 0.3134802179578299,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015538586929566707,
+      "loss": 0.9104,
+      "step": 6616
+    },
+    {
+      "epoch": 0.31352760009476427,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001553734689488621,
+      "loss": 0.7639,
+      "step": 6617
+    },
+    {
+      "epoch": 0.31357498223169866,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.00015536106737393257,
+      "loss": 0.0042,
+      "step": 6618
+    },
+    {
+      "epoch": 0.313622364368633,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015534866457115354,
+      "loss": 0.4751,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3136697465055674,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00015533626054080007,
+      "loss": 0.7211,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3137171286425018,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015532385528314727,
+      "loss": 1.0287,
+      "step": 6621
+    },
+    {
+      "epoch": 0.3137645107794362,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00015531144879847033,
+      "loss": 1.6725,
+      "step": 6622
+    },
+    {
+      "epoch": 0.3138118929163705,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015529904108704435,
+      "loss": 1.1494,
+      "step": 6623
+    },
+    {
+      "epoch": 0.3138592750533049,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015528663214914453,
+      "loss": 0.7874,
+      "step": 6624
+    },
+    {
+      "epoch": 0.3139066571902393,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001552742219850461,
+      "loss": 0.9382,
+      "step": 6625
+    },
+    {
+      "epoch": 0.31395403932717364,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015526181059502428,
+      "loss": 0.8744,
+      "step": 6626
+    },
+    {
+      "epoch": 0.31400142146410803,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015524939797935443,
+      "loss": 0.1217,
+      "step": 6627
+    },
+    {
+      "epoch": 0.3140488036010424,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015523698413831173,
+      "loss": 0.1361,
+      "step": 6628
+    },
+    {
+      "epoch": 0.31409618573797676,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015522456907217155,
+      "loss": 0.8062,
+      "step": 6629
+    },
+    {
+      "epoch": 0.31414356787491116,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001552121527812093,
+      "loss": 0.9175,
+      "step": 6630
+    },
+    {
+      "epoch": 0.31419095001184555,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015519973526570023,
+      "loss": 0.813,
+      "step": 6631
+    },
+    {
+      "epoch": 0.3142383321487799,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001551873165259199,
+      "loss": 1.2004,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015517489656214365,
+      "loss": 0.0825,
+      "step": 6633
+    },
+    {
+      "epoch": 0.3143330964226487,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00015516247537464696,
+      "loss": 0.1309,
+      "step": 6634
+    },
+    {
+      "epoch": 0.314380478559583,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015515005296370534,
+      "loss": 1.126,
+      "step": 6635
+    },
+    {
+      "epoch": 0.3144278606965174,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001551376293295943,
+      "loss": 0.8408,
+      "step": 6636
+    },
+    {
+      "epoch": 0.3144752428334518,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015512520447258932,
+      "loss": 1.3584,
+      "step": 6637
+    },
+    {
+      "epoch": 0.3145226249703862,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015511277839296605,
+      "loss": 1.3213,
+      "step": 6638
+    },
+    {
+      "epoch": 0.3145700071073205,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015510035109100008,
+      "loss": 0.9437,
+      "step": 6639
+    },
+    {
+      "epoch": 0.3146173892442549,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00015508792256696698,
+      "loss": 0.0295,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3146647713811893,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015507549282114244,
+      "loss": 0.3888,
+      "step": 6641
+    },
+    {
+      "epoch": 0.31471215351812365,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00015506306185380213,
+      "loss": 1.0023,
+      "step": 6642
+    },
+    {
+      "epoch": 0.31475953565505804,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001550506296652218,
+      "loss": 0.4771,
+      "step": 6643
+    },
+    {
+      "epoch": 0.31480691779199244,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001550381962556771,
+      "loss": 0.7989,
+      "step": 6644
+    },
+    {
+      "epoch": 0.3148542999289268,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00015502576162544385,
+      "loss": 0.0758,
+      "step": 6645
+    },
+    {
+      "epoch": 0.31490168206586117,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015501332577479778,
+      "loss": 1.0518,
+      "step": 6646
+    },
+    {
+      "epoch": 0.31494906420279556,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00015500088870401476,
+      "loss": 0.0091,
+      "step": 6647
+    },
+    {
+      "epoch": 0.3149964463397299,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015498845041337064,
+      "loss": 1.0809,
+      "step": 6648
+    },
+    {
+      "epoch": 0.3150438284766643,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00015497601090314124,
+      "loss": 0.2188,
+      "step": 6649
+    },
+    {
+      "epoch": 0.3150912106135987,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001549635701736025,
+      "loss": 1.1145,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3151385927505331,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015495112822503027,
+      "loss": 0.5646,
+      "step": 6651
+    },
+    {
+      "epoch": 0.3151859748874674,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001549386850577006,
+      "loss": 1.0729,
+      "step": 6652
+    },
+    {
+      "epoch": 0.3152333570244018,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001549262406718894,
+      "loss": 1.0544,
+      "step": 6653
+    },
+    {
+      "epoch": 0.3152807391613362,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015491379506787264,
+      "loss": 0.781,
+      "step": 6654
+    },
+    {
+      "epoch": 0.31532812129827054,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015490134824592644,
+      "loss": 0.9315,
+      "step": 6655
+    },
+    {
+      "epoch": 0.31537550343520493,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015488890020632677,
+      "loss": 1.2667,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3154228855721393,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00015487645094934983,
+      "loss": 0.0167,
+      "step": 6657
+    },
+    {
+      "epoch": 0.31547026770907366,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015486400047527162,
+      "loss": 0.0131,
+      "step": 6658
+    },
+    {
+      "epoch": 0.31551764984600805,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015485154878436833,
+      "loss": 0.6654,
+      "step": 6659
+    },
+    {
+      "epoch": 0.31556503198294245,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001548390958769161,
+      "loss": 0.4158,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3156124141198768,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015482664175319111,
+      "loss": 1.1609,
+      "step": 6661
+    },
+    {
+      "epoch": 0.3156597962568112,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015481418641346965,
+      "loss": 1.0335,
+      "step": 6662
+    },
+    {
+      "epoch": 0.31570717839374557,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0001548017298580279,
+      "loss": 0.0021,
+      "step": 6663
+    },
+    {
+      "epoch": 0.3157545605306799,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001547892720871422,
+      "loss": 0.7076,
+      "step": 6664
+    },
+    {
+      "epoch": 0.3158019426676143,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015477681310108873,
+      "loss": 0.7821,
+      "step": 6665
+    },
+    {
+      "epoch": 0.3158493248045487,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00015476435290014395,
+      "loss": 0.0123,
+      "step": 6666
+    },
+    {
+      "epoch": 0.3158967069414831,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001547518914845841,
+      "loss": 0.025,
+      "step": 6667
+    },
+    {
+      "epoch": 0.3159440890784174,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001547394288546857,
+      "loss": 1.0853,
+      "step": 6668
+    },
+    {
+      "epoch": 0.3159914712153518,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000154726965010725,
+      "loss": 0.0582,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3160388533522862,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001547144999529785,
+      "loss": 0.0365,
+      "step": 6670
+    },
+    {
+      "epoch": 0.31608623548922055,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001547020336817227,
+      "loss": 0.0178,
+      "step": 6671
+    },
+    {
+      "epoch": 0.31613361762615494,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015468956619723407,
+      "loss": 0.7368,
+      "step": 6672
+    },
+    {
+      "epoch": 0.31618099976308933,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001546770974997891,
+      "loss": 0.0375,
+      "step": 6673
+    },
+    {
+      "epoch": 0.31622838190002367,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015466462758966437,
+      "loss": 0.7151,
+      "step": 6674
+    },
+    {
+      "epoch": 0.31627576403695806,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001546521564671364,
+      "loss": 0.5066,
+      "step": 6675
+    },
+    {
+      "epoch": 0.31632314617389246,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001546396841324818,
+      "loss": 0.011,
+      "step": 6676
+    },
+    {
+      "epoch": 0.3163705283108268,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001546272105859772,
+      "loss": 1.2534,
+      "step": 6677
+    },
+    {
+      "epoch": 0.3164179104477612,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001546147358278993,
+      "loss": 0.9658,
+      "step": 6678
+    },
+    {
+      "epoch": 0.3164652925846956,
+      "grad_norm": 0.00341796875,
+      "learning_rate": 0.0001546022598585247,
+      "loss": 0.0003,
+      "step": 6679
+    },
+    {
+      "epoch": 0.31651267472162997,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015458978267813016,
+      "loss": 0.9939,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3165600568585643,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015457730428699237,
+      "loss": 0.5775,
+      "step": 6681
+    },
+    {
+      "epoch": 0.3166074389954987,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015456482468538814,
+      "loss": 0.9927,
+      "step": 6682
+    },
+    {
+      "epoch": 0.3166548211324331,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015455234387359417,
+      "loss": 0.867,
+      "step": 6683
+    },
+    {
+      "epoch": 0.31670220326936743,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0001545398618518873,
+      "loss": 0.009,
+      "step": 6684
+    },
+    {
+      "epoch": 0.3167495854063018,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015452737862054446,
+      "loss": 1.0679,
+      "step": 6685
+    },
+    {
+      "epoch": 0.3167969675432362,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015451489417984238,
+      "loss": 0.0163,
+      "step": 6686
+    },
+    {
+      "epoch": 0.31684434968017056,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015450240853005805,
+      "loss": 0.973,
+      "step": 6687
+    },
+    {
+      "epoch": 0.31689173181710495,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015448992167146834,
+      "loss": 0.3367,
+      "step": 6688
+    },
+    {
+      "epoch": 0.31693911395403934,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001544774336043502,
+      "loss": 0.1513,
+      "step": 6689
+    },
+    {
+      "epoch": 0.3169864960909737,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015446494432898064,
+      "loss": 1.4223,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3170338782279081,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001544524538456366,
+      "loss": 1.0795,
+      "step": 6691
+    },
+    {
+      "epoch": 0.31708126036484247,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015443996215459515,
+      "loss": 0.4657,
+      "step": 6692
+    },
+    {
+      "epoch": 0.3171286425017768,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001544274692561333,
+      "loss": 0.9374,
+      "step": 6693
+    },
+    {
+      "epoch": 0.3171760246387112,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015441497515052816,
+      "loss": 0.9042,
+      "step": 6694
+    },
+    {
+      "epoch": 0.3172234067756456,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00015440247983805686,
+      "loss": 0.0815,
+      "step": 6695
+    },
+    {
+      "epoch": 0.31727078891258,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001543899833189965,
+      "loss": 1.362,
+      "step": 6696
+    },
+    {
+      "epoch": 0.3173181710495143,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001543774855936242,
+      "loss": 0.6509,
+      "step": 6697
+    },
+    {
+      "epoch": 0.3173655531864487,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00015436498666221725,
+      "loss": 0.014,
+      "step": 6698
+    },
+    {
+      "epoch": 0.3174129353233831,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00015435248652505276,
+      "loss": 0.8763,
+      "step": 6699
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000154339985182408,
+      "loss": 0.7114,
+      "step": 6700
+    },
+    {
+      "epoch": 0.31750769959725184,
+      "grad_norm": 0.0159912109375,
+      "learning_rate": 0.0001543274826345603,
+      "loss": 0.001,
+      "step": 6701
+    },
+    {
+      "epoch": 0.31755508173418623,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00015431497888178687,
+      "loss": 0.0264,
+      "step": 6702
+    },
+    {
+      "epoch": 0.31760246387112057,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00015430247392436507,
+      "loss": 1.4535,
+      "step": 6703
+    },
+    {
+      "epoch": 0.31764984600805496,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00015428996776257225,
+      "loss": 0.6312,
+      "step": 6704
+    },
+    {
+      "epoch": 0.31769722814498935,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00015427746039668575,
+      "loss": 0.1916,
+      "step": 6705
+    },
+    {
+      "epoch": 0.3177446102819237,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015426495182698302,
+      "loss": 0.7129,
+      "step": 6706
+    },
+    {
+      "epoch": 0.3177919924188581,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00015425244205374144,
+      "loss": 0.1759,
+      "step": 6707
+    },
+    {
+      "epoch": 0.3178393745557925,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015423993107723847,
+      "loss": 1.465,
+      "step": 6708
+    },
+    {
+      "epoch": 0.31788675669272687,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001542274188977516,
+      "loss": 0.9762,
+      "step": 6709
+    },
+    {
+      "epoch": 0.3179341388296612,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00015421490551555838,
+      "loss": 0.1516,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3179815209665956,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015420239093093623,
+      "loss": 1.2048,
+      "step": 6711
+    },
+    {
+      "epoch": 0.31802890310353,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015418987514416286,
+      "loss": 1.065,
+      "step": 6712
+    },
+    {
+      "epoch": 0.31807628524046433,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015417735815551574,
+      "loss": 0.2488,
+      "step": 6713
+    },
+    {
+      "epoch": 0.3181236673773987,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015416483996527256,
+      "loss": 1.0395,
+      "step": 6714
+    },
+    {
+      "epoch": 0.3181710495143331,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001541523205737109,
+      "loss": 0.411,
+      "step": 6715
+    },
+    {
+      "epoch": 0.31821843165126745,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015413979998110845,
+      "loss": 0.7368,
+      "step": 6716
+    },
+    {
+      "epoch": 0.31826581378820185,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015412727818774288,
+      "loss": 0.4258,
+      "step": 6717
+    },
+    {
+      "epoch": 0.31831319592513624,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015411475519389197,
+      "loss": 0.8731,
+      "step": 6718
+    },
+    {
+      "epoch": 0.3183605780620706,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00015410223099983342,
+      "loss": 1.1324,
+      "step": 6719
+    },
+    {
+      "epoch": 0.31840796019900497,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.000154089705605845,
+      "loss": 0.0377,
+      "step": 6720
+    },
+    {
+      "epoch": 0.31845534233593936,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015407717901220452,
+      "loss": 0.0648,
+      "step": 6721
+    },
+    {
+      "epoch": 0.3185027244728737,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015406465121918985,
+      "loss": 1.4768,
+      "step": 6722
+    },
+    {
+      "epoch": 0.3185501066098081,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015405212222707875,
+      "loss": 0.0798,
+      "step": 6723
+    },
+    {
+      "epoch": 0.3185974887467425,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015403959203614918,
+      "loss": 0.9283,
+      "step": 6724
+    },
+    {
+      "epoch": 0.3186448708836769,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015402706064667902,
+      "loss": 1.3497,
+      "step": 6725
+    },
+    {
+      "epoch": 0.3186922530206112,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001540145280589462,
+      "loss": 0.6482,
+      "step": 6726
+    },
+    {
+      "epoch": 0.3187396351575456,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015400199427322868,
+      "loss": 1.0577,
+      "step": 6727
+    },
+    {
+      "epoch": 0.31878701729448,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015398945928980445,
+      "loss": 1.4132,
+      "step": 6728
+    },
+    {
+      "epoch": 0.31883439943141434,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001539769231089515,
+      "loss": 0.7271,
+      "step": 6729
+    },
+    {
+      "epoch": 0.31888178156834873,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015396438573094792,
+      "loss": 0.0687,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3189291637052831,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00015395184715607174,
+      "loss": 0.0163,
+      "step": 6731
+    },
+    {
+      "epoch": 0.31897654584221746,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015393930738460103,
+      "loss": 0.6262,
+      "step": 6732
+    },
+    {
+      "epoch": 0.31902392797915186,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015392676641681394,
+      "loss": 0.6423,
+      "step": 6733
+    },
+    {
+      "epoch": 0.31907131011608625,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015391422425298866,
+      "loss": 0.9327,
+      "step": 6734
+    },
+    {
+      "epoch": 0.3191186922530206,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00015390168089340325,
+      "loss": 0.3877,
+      "step": 6735
+    },
+    {
+      "epoch": 0.319166074389955,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015388913633833598,
+      "loss": 0.8933,
+      "step": 6736
+    },
+    {
+      "epoch": 0.31921345652688937,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001538765905880651,
+      "loss": 0.6484,
+      "step": 6737
+    },
+    {
+      "epoch": 0.31926083866382376,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015386404364286884,
+      "loss": 0.1907,
+      "step": 6738
+    },
+    {
+      "epoch": 0.3193082208007581,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015385149550302543,
+      "loss": 1.6144,
+      "step": 6739
+    },
+    {
+      "epoch": 0.3193556029376925,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015383894616881322,
+      "loss": 0.8437,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3194029850746269,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001538263956405105,
+      "loss": 1.4197,
+      "step": 6741
+    },
+    {
+      "epoch": 0.3194503672115612,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001538138439183957,
+      "loss": 1.0557,
+      "step": 6742
+    },
+    {
+      "epoch": 0.3194977493484956,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015380129100274714,
+      "loss": 0.5517,
+      "step": 6743
+    },
+    {
+      "epoch": 0.31954513148543,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015378873689384328,
+      "loss": 1.4355,
+      "step": 6744
+    },
+    {
+      "epoch": 0.31959251362236435,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001537761815919625,
+      "loss": 0.1077,
+      "step": 6745
+    },
+    {
+      "epoch": 0.31963989575929874,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00015376362509738332,
+      "loss": 0.5584,
+      "step": 6746
+    },
+    {
+      "epoch": 0.31968727789623314,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001537510674103842,
+      "loss": 0.9827,
+      "step": 6747
+    },
+    {
+      "epoch": 0.3197346600331675,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00015373850853124362,
+      "loss": 0.7614,
+      "step": 6748
+    },
+    {
+      "epoch": 0.31978204217010187,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00015372594846024018,
+      "loss": 0.6042,
+      "step": 6749
+    },
+    {
+      "epoch": 0.31982942430703626,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015371338719765242,
+      "loss": 0.6404,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3198768064439706,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015370082474375896,
+      "loss": 0.7974,
+      "step": 6751
+    },
+    {
+      "epoch": 0.319924188580905,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00015368826109883837,
+      "loss": 0.0118,
+      "step": 6752
+    },
+    {
+      "epoch": 0.3199715707178394,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001536756962631694,
+      "loss": 0.9106,
+      "step": 6753
+    },
+    {
+      "epoch": 0.3200189528547738,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00015366313023703061,
+      "loss": 0.0626,
+      "step": 6754
+    },
+    {
+      "epoch": 0.3200663349917081,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015365056302070077,
+      "loss": 0.9871,
+      "step": 6755
+    },
+    {
+      "epoch": 0.3201137171286425,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001536379946144586,
+      "loss": 0.7532,
+      "step": 6756
+    },
+    {
+      "epoch": 0.3201610992655769,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0001536254250185828,
+      "loss": 0.0051,
+      "step": 6757
+    },
+    {
+      "epoch": 0.32020848140251124,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00015361285423335223,
+      "loss": 1.1125,
+      "step": 6758
+    },
+    {
+      "epoch": 0.32025586353944563,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015360028225904567,
+      "loss": 0.8966,
+      "step": 6759
+    },
+    {
+      "epoch": 0.32030324567638,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015358770909594188,
+      "loss": 0.189,
+      "step": 6760
+    },
+    {
+      "epoch": 0.32035062781331436,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015357513474431985,
+      "loss": 0.7883,
+      "step": 6761
+    },
+    {
+      "epoch": 0.32039800995024875,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001535625592044584,
+      "loss": 1.0491,
+      "step": 6762
+    },
+    {
+      "epoch": 0.32044539208718315,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001535499824766364,
+      "loss": 0.8326,
+      "step": 6763
+    },
+    {
+      "epoch": 0.3204927742241175,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001535374045611328,
+      "loss": 1.5672,
+      "step": 6764
+    },
+    {
+      "epoch": 0.3205401563610519,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015352482545822666,
+      "loss": 0.678,
+      "step": 6765
+    },
+    {
+      "epoch": 0.32058753849798627,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015351224516819688,
+      "loss": 0.1243,
+      "step": 6766
+    },
+    {
+      "epoch": 0.32063492063492066,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00015349966369132247,
+      "loss": 1.1682,
+      "step": 6767
+    },
+    {
+      "epoch": 0.320682302771855,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015348708102788254,
+      "loss": 1.3193,
+      "step": 6768
+    },
+    {
+      "epoch": 0.3207296849087894,
+      "grad_norm": 0.0240478515625,
+      "learning_rate": 0.0001534744971781561,
+      "loss": 0.0012,
+      "step": 6769
+    },
+    {
+      "epoch": 0.3207770670457238,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001534619121424223,
+      "loss": 1.0788,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3208244491826581,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00015344932592096022,
+      "loss": 0.1021,
+      "step": 6771
+    },
+    {
+      "epoch": 0.3208718313195925,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015343673851404903,
+      "loss": 1.2014,
+      "step": 6772
+    },
+    {
+      "epoch": 0.3209192134565269,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015342414992196787,
+      "loss": 0.716,
+      "step": 6773
+    },
+    {
+      "epoch": 0.32096659559346125,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015341156014499598,
+      "loss": 1.0394,
+      "step": 6774
+    },
+    {
+      "epoch": 0.32101397773039564,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015339896918341257,
+      "loss": 0.9286,
+      "step": 6775
+    },
+    {
+      "epoch": 0.32106135986733003,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00015338637703749694,
+      "loss": 1.0344,
+      "step": 6776
+    },
+    {
+      "epoch": 0.32110874200426437,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015337378370752831,
+      "loss": 0.751,
+      "step": 6777
+    },
+    {
+      "epoch": 0.32115612414119876,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015336118919378601,
+      "loss": 1.0953,
+      "step": 6778
+    },
+    {
+      "epoch": 0.32120350627813316,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015334859349654936,
+      "loss": 0.9351,
+      "step": 6779
+    },
+    {
+      "epoch": 0.3212508884150675,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00015333599661609775,
+      "loss": 0.0137,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3212982705520019,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00015332339855271052,
+      "loss": 0.0023,
+      "step": 6781
+    },
+    {
+      "epoch": 0.3213456526889363,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001533107993066671,
+      "loss": 1.0246,
+      "step": 6782
+    },
+    {
+      "epoch": 0.32139303482587067,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015329819887824697,
+      "loss": 0.8105,
+      "step": 6783
+    },
+    {
+      "epoch": 0.321440416962805,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015328559726772957,
+      "loss": 0.0319,
+      "step": 6784
+    },
+    {
+      "epoch": 0.3214877990997394,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00015327299447539435,
+      "loss": 0.1445,
+      "step": 6785
+    },
+    {
+      "epoch": 0.3215351812366738,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015326039050152086,
+      "loss": 0.6866,
+      "step": 6786
+    },
+    {
+      "epoch": 0.32158256337360813,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015324778534638863,
+      "loss": 1.0162,
+      "step": 6787
+    },
+    {
+      "epoch": 0.3216299455105425,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00015323517901027727,
+      "loss": 1.3541,
+      "step": 6788
+    },
+    {
+      "epoch": 0.3216773276474769,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001532225714934663,
+      "loss": 0.7535,
+      "step": 6789
+    },
+    {
+      "epoch": 0.32172470978441126,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00015320996279623544,
+      "loss": 0.1705,
+      "step": 6790
+    },
+    {
+      "epoch": 0.32177209192134565,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015319735291886423,
+      "loss": 1.1241,
+      "step": 6791
+    },
+    {
+      "epoch": 0.32181947405828004,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00015318474186163243,
+      "loss": 1.0389,
+      "step": 6792
+    },
+    {
+      "epoch": 0.3218668561952144,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00015317212962481967,
+      "loss": 1.0484,
+      "step": 6793
+    },
+    {
+      "epoch": 0.32191423833214877,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015315951620870573,
+      "loss": 1.4084,
+      "step": 6794
+    },
+    {
+      "epoch": 0.32196162046908317,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015314690161357037,
+      "loss": 0.9202,
+      "step": 6795
+    },
+    {
+      "epoch": 0.32200900260601756,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001531342858396933,
+      "loss": 0.7585,
+      "step": 6796
+    },
+    {
+      "epoch": 0.3220563847429519,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015312166888735437,
+      "loss": 1.2135,
+      "step": 6797
+    },
+    {
+      "epoch": 0.3221037668798863,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015310905075683343,
+      "loss": 0.1361,
+      "step": 6798
+    },
+    {
+      "epoch": 0.3221511490168207,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00015309643144841034,
+      "loss": 1.1331,
+      "step": 6799
+    },
+    {
+      "epoch": 0.322198531153755,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001530838109623649,
+      "loss": 0.0115,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3222459132906894,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001530711892989771,
+      "loss": 0.2296,
+      "step": 6801
+    },
+    {
+      "epoch": 0.3222932954276238,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015305856645852686,
+      "loss": 1.0299,
+      "step": 6802
+    },
+    {
+      "epoch": 0.32234067756455814,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015304594244129412,
+      "loss": 0.1335,
+      "step": 6803
+    },
+    {
+      "epoch": 0.32238805970149254,
+      "grad_norm": 0.004669189453125,
+      "learning_rate": 0.00015303331724755885,
+      "loss": 0.0002,
+      "step": 6804
+    },
+    {
+      "epoch": 0.32243544183842693,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001530206908776011,
+      "loss": 1.3244,
+      "step": 6805
+    },
+    {
+      "epoch": 0.32248282397536127,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00015300806333170094,
+      "loss": 0.1052,
+      "step": 6806
+    },
+    {
+      "epoch": 0.32253020611229566,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015299543461013836,
+      "loss": 0.8976,
+      "step": 6807
+    },
+    {
+      "epoch": 0.32257758824923005,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015298280471319353,
+      "loss": 0.1655,
+      "step": 6808
+    },
+    {
+      "epoch": 0.3226249703861644,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00015297017364114649,
+      "loss": 0.0132,
+      "step": 6809
+    },
+    {
+      "epoch": 0.3226723525230988,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015295754139427742,
+      "loss": 0.7851,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3227197346600332,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001529449079728665,
+      "loss": 1.1728,
+      "step": 6811
+    },
+    {
+      "epoch": 0.32276711679696757,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015293227337719395,
+      "loss": 0.7953,
+      "step": 6812
+    },
+    {
+      "epoch": 0.3228144989339019,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015291963760753988,
+      "loss": 0.4337,
+      "step": 6813
+    },
+    {
+      "epoch": 0.3228618810708363,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001529070006641847,
+      "loss": 0.0239,
+      "step": 6814
+    },
+    {
+      "epoch": 0.3229092632077707,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015289436254740853,
+      "loss": 1.0518,
+      "step": 6815
+    },
+    {
+      "epoch": 0.32295664534470503,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015288172325749177,
+      "loss": 0.6293,
+      "step": 6816
+    },
+    {
+      "epoch": 0.3230040274816394,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015286908279471472,
+      "loss": 0.7351,
+      "step": 6817
+    },
+    {
+      "epoch": 0.3230514096185738,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015285644115935773,
+      "loss": 0.9188,
+      "step": 6818
+    },
+    {
+      "epoch": 0.32309879175550815,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00015284379835170118,
+      "loss": 0.0971,
+      "step": 6819
+    },
+    {
+      "epoch": 0.32314617389244255,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00015283115437202542,
+      "loss": 0.5676,
+      "step": 6820
+    },
+    {
+      "epoch": 0.32319355602937694,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015281850922061095,
+      "loss": 0.9162,
+      "step": 6821
+    },
+    {
+      "epoch": 0.3232409381663113,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015280586289773823,
+      "loss": 1.3716,
+      "step": 6822
+    },
+    {
+      "epoch": 0.32328832030324567,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001527932154036877,
+      "loss": 0.3771,
+      "step": 6823
+    },
+    {
+      "epoch": 0.32333570244018006,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001527805667387399,
+      "loss": 0.4614,
+      "step": 6824
+    },
+    {
+      "epoch": 0.32338308457711445,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00015276791690317534,
+      "loss": 0.5965,
+      "step": 6825
+    },
+    {
+      "epoch": 0.3234304667140488,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015275526589727463,
+      "loss": 1.106,
+      "step": 6826
+    },
+    {
+      "epoch": 0.3234778488509832,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015274261372131824,
+      "loss": 1.1501,
+      "step": 6827
+    },
+    {
+      "epoch": 0.3235252309879176,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00015272996037558688,
+      "loss": 0.0227,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3235726131248519,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015271730586036118,
+      "loss": 0.9243,
+      "step": 6829
+    },
+    {
+      "epoch": 0.3236199952617863,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001527046501759218,
+      "loss": 0.1782,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3236673773987207,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001526919933225494,
+      "loss": 0.1422,
+      "step": 6831
+    },
+    {
+      "epoch": 0.32371475953565504,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001526793353005247,
+      "loss": 1.4339,
+      "step": 6832
+    },
+    {
+      "epoch": 0.32376214167258943,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015266667611012847,
+      "loss": 0.9213,
+      "step": 6833
+    },
+    {
+      "epoch": 0.3238095238095238,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015265401575164148,
+      "loss": 0.6695,
+      "step": 6834
+    },
+    {
+      "epoch": 0.32385690594645816,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015264135422534447,
+      "loss": 1.099,
+      "step": 6835
+    },
+    {
+      "epoch": 0.32390428808339256,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001526286915315183,
+      "loss": 1.2299,
+      "step": 6836
+    },
+    {
+      "epoch": 0.32395167022032695,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015261602767044382,
+      "loss": 0.6742,
+      "step": 6837
+    },
+    {
+      "epoch": 0.3239990523572613,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015260336264240188,
+      "loss": 1.0091,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3240464344941957,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001525906964476734,
+      "loss": 0.1714,
+      "step": 6839
+    },
+    {
+      "epoch": 0.32409381663113007,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015257802908653928,
+      "loss": 0.1054,
+      "step": 6840
+    },
+    {
+      "epoch": 0.32414119876806446,
+      "grad_norm": 0.00323486328125,
+      "learning_rate": 0.0001525653605592805,
+      "loss": 0.0003,
+      "step": 6841
+    },
+    {
+      "epoch": 0.3241885809049988,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000152552690866178,
+      "loss": 0.1259,
+      "step": 6842
+    },
+    {
+      "epoch": 0.3242359630419332,
+      "grad_norm": 0.0028228759765625,
+      "learning_rate": 0.0001525400200075128,
+      "loss": 0.0002,
+      "step": 6843
+    },
+    {
+      "epoch": 0.3242833451788676,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00015252734798356588,
+      "loss": 1.4238,
+      "step": 6844
+    },
+    {
+      "epoch": 0.3243307273158019,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.00015251467479461837,
+      "loss": 0.0015,
+      "step": 6845
+    },
+    {
+      "epoch": 0.3243781094527363,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001525020004409513,
+      "loss": 0.3023,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3244254915896707,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001524893249228458,
+      "loss": 1.0292,
+      "step": 6847
+    },
+    {
+      "epoch": 0.32447287372660505,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015247664824058295,
+      "loss": 0.2902,
+      "step": 6848
+    },
+    {
+      "epoch": 0.32452025586353944,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015246397039444398,
+      "loss": 0.9553,
+      "step": 6849
+    },
+    {
+      "epoch": 0.32456763800047383,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015245129138471,
+      "loss": 0.8417,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3246150201374082,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015243861121166222,
+      "loss": 0.9548,
+      "step": 6851
+    },
+    {
+      "epoch": 0.32466240227434257,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015242592987558192,
+      "loss": 0.9847,
+      "step": 6852
+    },
+    {
+      "epoch": 0.32470978441127696,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00015241324737675038,
+      "loss": 1.4285,
+      "step": 6853
+    },
+    {
+      "epoch": 0.32475716654821135,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015240056371544877,
+      "loss": 1.125,
+      "step": 6854
+    },
+    {
+      "epoch": 0.3248045486851457,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001523878788919585,
+      "loss": 0.7167,
+      "step": 6855
+    },
+    {
+      "epoch": 0.3248519308220801,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001523751929065609,
+      "loss": 0.7513,
+      "step": 6856
+    },
+    {
+      "epoch": 0.3248993129590145,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001523625057595373,
+      "loss": 1.1208,
+      "step": 6857
+    },
+    {
+      "epoch": 0.3249466950959488,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001523498174511691,
+      "loss": 0.7778,
+      "step": 6858
+    },
+    {
+      "epoch": 0.3249940772328832,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0001523371279817377,
+      "loss": 0.0056,
+      "step": 6859
+    },
+    {
+      "epoch": 0.3250414593698176,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00015232443735152456,
+      "loss": 0.9183,
+      "step": 6860
+    },
+    {
+      "epoch": 0.32508884150675194,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015231174556081109,
+      "loss": 0.7987,
+      "step": 6861
+    },
+    {
+      "epoch": 0.32513622364368633,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00015229905260987886,
+      "loss": 0.1747,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3251836057806207,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00015228635849900935,
+      "loss": 1.0784,
+      "step": 6863
+    },
+    {
+      "epoch": 0.32523098791755506,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015227366322848407,
+      "loss": 0.0737,
+      "step": 6864
+    },
+    {
+      "epoch": 0.32527837005448945,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00015226096679858467,
+      "loss": 0.0539,
+      "step": 6865
+    },
+    {
+      "epoch": 0.32532575219142384,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015224826920959264,
+      "loss": 1.331,
+      "step": 6866
+    },
+    {
+      "epoch": 0.3253731343283582,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015223557046178967,
+      "loss": 0.8021,
+      "step": 6867
+    },
+    {
+      "epoch": 0.3254205164652926,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001522228705554574,
+      "loss": 0.6568,
+      "step": 6868
+    },
+    {
+      "epoch": 0.32546789860222697,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00015221016949087748,
+      "loss": 1.5958,
+      "step": 6869
+    },
+    {
+      "epoch": 0.32551528073916136,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001521974672683316,
+      "loss": 1.0054,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3255626628760957,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00015218476388810151,
+      "loss": 0.1415,
+      "step": 6871
+    },
+    {
+      "epoch": 0.3256100450130301,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015217205935046896,
+      "loss": 0.9692,
+      "step": 6872
+    },
+    {
+      "epoch": 0.3256574271499645,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015215935365571568,
+      "loss": 0.6272,
+      "step": 6873
+    },
+    {
+      "epoch": 0.3257048092868988,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001521466468041235,
+      "loss": 1.0321,
+      "step": 6874
+    },
+    {
+      "epoch": 0.3257521914238332,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001521339387959742,
+      "loss": 1.1456,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3257995735607676,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00015212122963154974,
+      "loss": 0.2301,
+      "step": 6876
+    },
+    {
+      "epoch": 0.32584695569770195,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00015210851931113186,
+      "loss": 0.0103,
+      "step": 6877
+    },
+    {
+      "epoch": 0.32589433783463634,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015209580783500255,
+      "loss": 1.1929,
+      "step": 6878
+    },
+    {
+      "epoch": 0.32594171997157073,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00015208309520344372,
+      "loss": 1.2416,
+      "step": 6879
+    },
+    {
+      "epoch": 0.32598910210850507,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001520703814167373,
+      "loss": 1.2499,
+      "step": 6880
+    },
+    {
+      "epoch": 0.32603648424543946,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001520576664751653,
+      "loss": 0.8114,
+      "step": 6881
+    },
+    {
+      "epoch": 0.32608386638237385,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001520449503790097,
+      "loss": 0.1669,
+      "step": 6882
+    },
+    {
+      "epoch": 0.32613124851930825,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015203223312855254,
+      "loss": 0.5363,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3261786306562426,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00015201951472407584,
+      "loss": 1.1362,
+      "step": 6884
+    },
+    {
+      "epoch": 0.326226012793177,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015200679516586173,
+      "loss": 0.0285,
+      "step": 6885
+    },
+    {
+      "epoch": 0.32627339493011137,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015199407445419235,
+      "loss": 1.0288,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3263207770670457,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015198135258934968,
+      "loss": 1.0881,
+      "step": 6887
+    },
+    {
+      "epoch": 0.3263681592039801,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00015196862957161605,
+      "loss": 0.0375,
+      "step": 6888
+    },
+    {
+      "epoch": 0.3264155413409145,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00015195590540127357,
+      "loss": 0.0011,
+      "step": 6889
+    },
+    {
+      "epoch": 0.32646292347784883,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015194318007860443,
+      "loss": 0.862,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3265103056147832,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015193045360389088,
+      "loss": 1.0593,
+      "step": 6891
+    },
+    {
+      "epoch": 0.3265576877517176,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001519177259774152,
+      "loss": 0.0673,
+      "step": 6892
+    },
+    {
+      "epoch": 0.32660506988865196,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015190499719945963,
+      "loss": 0.0542,
+      "step": 6893
+    },
+    {
+      "epoch": 0.32665245202558635,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015189226727030652,
+      "loss": 1.0707,
+      "step": 6894
+    },
+    {
+      "epoch": 0.32669983416252074,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015187953619023823,
+      "loss": 1.0912,
+      "step": 6895
+    },
+    {
+      "epoch": 0.3267472162994551,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00015186680395953706,
+      "loss": 0.0081,
+      "step": 6896
+    },
+    {
+      "epoch": 0.32679459843638947,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001518540705784854,
+      "loss": 1.0111,
+      "step": 6897
+    },
+    {
+      "epoch": 0.32684198057332386,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015184133604736571,
+      "loss": 1.0445,
+      "step": 6898
+    },
+    {
+      "epoch": 0.32688936271025826,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015182860036646041,
+      "loss": 0.9122,
+      "step": 6899
+    },
+    {
+      "epoch": 0.3269367448471926,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00015181586353605196,
+      "loss": 1.3648,
+      "step": 6900
+    },
+    {
+      "epoch": 0.326984126984127,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001518031255564228,
+      "loss": 0.7088,
+      "step": 6901
+    },
+    {
+      "epoch": 0.3270315091210614,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015179038642785557,
+      "loss": 1.4338,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3270788912579957,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001517776461506327,
+      "loss": 0.8348,
+      "step": 6903
+    },
+    {
+      "epoch": 0.3271262733949301,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015176490472503678,
+      "loss": 0.9255,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3271736555318645,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00015175216215135043,
+      "loss": 0.035,
+      "step": 6905
+    },
+    {
+      "epoch": 0.32722103766879884,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015173941842985627,
+      "loss": 1.0767,
+      "step": 6906
+    },
+    {
+      "epoch": 0.32726841980573323,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015172667356083686,
+      "loss": 0.8763,
+      "step": 6907
+    },
+    {
+      "epoch": 0.32731580194266763,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015171392754457494,
+      "loss": 0.6437,
+      "step": 6908
+    },
+    {
+      "epoch": 0.32736318407960197,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00015170118038135318,
+      "loss": 0.1531,
+      "step": 6909
+    },
+    {
+      "epoch": 0.32741056621653636,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015168843207145435,
+      "loss": 0.0325,
+      "step": 6910
+    },
+    {
+      "epoch": 0.32745794835347075,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015167568261516116,
+      "loss": 1.0991,
+      "step": 6911
+    },
+    {
+      "epoch": 0.32750533049040514,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015166293201275633,
+      "loss": 0.8243,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3275527126273395,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001516501802645227,
+      "loss": 1.4578,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3276000947642739,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015163742737074314,
+      "loss": 0.8581,
+      "step": 6914
+    },
+    {
+      "epoch": 0.32764747690120827,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001516246733317004,
+      "loss": 0.9694,
+      "step": 6915
+    },
+    {
+      "epoch": 0.3276948590381426,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001516119181476774,
+      "loss": 0.5541,
+      "step": 6916
+    },
+    {
+      "epoch": 0.327742241175077,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.000151599161818957,
+      "loss": 1.2393,
+      "step": 6917
+    },
+    {
+      "epoch": 0.3277896233120114,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001515864043458222,
+      "loss": 1.134,
+      "step": 6918
+    },
+    {
+      "epoch": 0.32783700544894573,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001515736457285559,
+      "loss": 0.6624,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3278843875858801,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015156088596744103,
+      "loss": 0.6075,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3279317697228145,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015154812506276065,
+      "loss": 1.0984,
+      "step": 6921
+    },
+    {
+      "epoch": 0.32797915185974885,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015153536301479775,
+      "loss": 0.9776,
+      "step": 6922
+    },
+    {
+      "epoch": 0.32802653399668324,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015152259982383542,
+      "loss": 1.2544,
+      "step": 6923
+    },
+    {
+      "epoch": 0.32807391613361764,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00015150983549015666,
+      "loss": 0.0135,
+      "step": 6924
+    },
+    {
+      "epoch": 0.328121298270552,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015149707001404464,
+      "loss": 1.0113,
+      "step": 6925
+    },
+    {
+      "epoch": 0.32816868040748637,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001514843033957825,
+      "loss": 1.1702,
+      "step": 6926
+    },
+    {
+      "epoch": 0.32821606254442076,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00015147153563565332,
+      "loss": 0.0092,
+      "step": 6927
+    },
+    {
+      "epoch": 0.32826344468135515,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00015145876673394027,
+      "loss": 0.017,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3283108268182895,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00015144599669092662,
+      "loss": 1.0963,
+      "step": 6929
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015143322550689554,
+      "loss": 1.3804,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3284055910921583,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015142045318213031,
+      "loss": 0.2335,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3284529732290926,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001514076797169142,
+      "loss": 0.9878,
+      "step": 6932
+    },
+    {
+      "epoch": 0.328500355366027,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015139490511153055,
+      "loss": 0.0259,
+      "step": 6933
+    },
+    {
+      "epoch": 0.3285477375029614,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015138212936626258,
+      "loss": 0.8695,
+      "step": 6934
+    },
+    {
+      "epoch": 0.32859511963989574,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00015136935248139376,
+      "loss": 1.6176,
+      "step": 6935
+    },
+    {
+      "epoch": 0.32864250177683013,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015135657445720742,
+      "loss": 0.122,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3286898839137645,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015134379529398693,
+      "loss": 0.8374,
+      "step": 6937
+    },
+    {
+      "epoch": 0.32873726605069886,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015133101499201576,
+      "loss": 0.9414,
+      "step": 6938
+    },
+    {
+      "epoch": 0.32878464818763325,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015131823355157736,
+      "loss": 0.7317,
+      "step": 6939
+    },
+    {
+      "epoch": 0.32883203032456765,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015130545097295518,
+      "loss": 1.0873,
+      "step": 6940
+    },
+    {
+      "epoch": 0.32887941246150204,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001512926672564328,
+      "loss": 0.0427,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3289267945984364,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015127988240229364,
+      "loss": 0.9747,
+      "step": 6942
+    },
+    {
+      "epoch": 0.32897417673537077,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015126709641082132,
+      "loss": 1.2964,
+      "step": 6943
+    },
+    {
+      "epoch": 0.32902155887230516,
+      "grad_norm": 0.0018157958984375,
+      "learning_rate": 0.00015125430928229942,
+      "loss": 0.0002,
+      "step": 6944
+    },
+    {
+      "epoch": 0.3290689410092395,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015124152101701155,
+      "loss": 1.0356,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3291163231461739,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015122873161524126,
+      "loss": 1.1814,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3291637052831083,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001512159410772723,
+      "loss": 0.0803,
+      "step": 6947
+    },
+    {
+      "epoch": 0.3292110874200426,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015120314940338835,
+      "loss": 0.8302,
+      "step": 6948
+    },
+    {
+      "epoch": 0.329258469556977,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015119035659387308,
+      "loss": 1.1037,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3293058516939114,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001511775626490102,
+      "loss": 0.5201,
+      "step": 6950
+    },
+    {
+      "epoch": 0.32935323383084575,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00015116476756908353,
+      "loss": 0.1235,
+      "step": 6951
+    },
+    {
+      "epoch": 0.32940061596778014,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0001511519713543768,
+      "loss": 0.0051,
+      "step": 6952
+    },
+    {
+      "epoch": 0.32944799810471453,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001511391740051738,
+      "loss": 0.0045,
+      "step": 6953
+    },
+    {
+      "epoch": 0.32949538024164887,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001511263755217584,
+      "loss": 0.3869,
+      "step": 6954
+    },
+    {
+      "epoch": 0.32954276237858326,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015111357590441444,
+      "loss": 1.262,
+      "step": 6955
+    },
+    {
+      "epoch": 0.32959014451551766,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015110077515342586,
+      "loss": 0.0431,
+      "step": 6956
+    },
+    {
+      "epoch": 0.32963752665245205,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001510879732690765,
+      "loss": 1.046,
+      "step": 6957
+    },
+    {
+      "epoch": 0.3296849087893864,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001510751702516503,
+      "loss": 0.0331,
+      "step": 6958
+    },
+    {
+      "epoch": 0.3297322909263208,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00015106236610143122,
+      "loss": 0.7115,
+      "step": 6959
+    },
+    {
+      "epoch": 0.3297796730632552,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015104956081870324,
+      "loss": 0.5726,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3298270552001895,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001510367544037504,
+      "loss": 0.2715,
+      "step": 6961
+    },
+    {
+      "epoch": 0.3298744373371239,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015102394685685667,
+      "loss": 0.0544,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3299218194740583,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001510111381783062,
+      "loss": 0.0501,
+      "step": 6963
+    },
+    {
+      "epoch": 0.32996920161099264,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.000150998328368383,
+      "loss": 0.9406,
+      "step": 6964
+    },
+    {
+      "epoch": 0.33001658374792703,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001509855174273712,
+      "loss": 0.8347,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3300639658848614,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015097270535555487,
+      "loss": 0.9301,
+      "step": 6966
+    },
+    {
+      "epoch": 0.33011134802179576,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015095989215321832,
+      "loss": 0.7732,
+      "step": 6967
+    },
+    {
+      "epoch": 0.33015873015873015,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001509470778206456,
+      "loss": 1.3104,
+      "step": 6968
+    },
+    {
+      "epoch": 0.33020611229566454,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00015093426235812096,
+      "loss": 0.162,
+      "step": 6969
+    },
+    {
+      "epoch": 0.33025349443259894,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0001509214457659286,
+      "loss": 1.7801,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3303008765695333,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015090862804435288,
+      "loss": 1.7799,
+      "step": 6971
+    },
+    {
+      "epoch": 0.33034825870646767,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015089580919367798,
+      "loss": 0.642,
+      "step": 6972
+    },
+    {
+      "epoch": 0.33039564084340206,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00015088298921418826,
+      "loss": 1.085,
+      "step": 6973
+    },
+    {
+      "epoch": 0.3304430229803364,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00015087016810616803,
+      "loss": 1.5359,
+      "step": 6974
+    },
+    {
+      "epoch": 0.3304904051172708,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00015085734586990166,
+      "loss": 1.1087,
+      "step": 6975
+    },
+    {
+      "epoch": 0.3305377872542052,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00015084452250567352,
+      "loss": 0.0126,
+      "step": 6976
+    },
+    {
+      "epoch": 0.3305851693911395,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00015083169801376802,
+      "loss": 0.0923,
+      "step": 6977
+    },
+    {
+      "epoch": 0.3306325515280739,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001508188723944696,
+      "loss": 1.0601,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3306799336650083,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015080604564806274,
+      "loss": 1.2822,
+      "step": 6979
+    },
+    {
+      "epoch": 0.33072731580194265,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001507932177748319,
+      "loss": 0.8857,
+      "step": 6980
+    },
+    {
+      "epoch": 0.33077469793887704,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015078038877506159,
+      "loss": 0.0921,
+      "step": 6981
+    },
+    {
+      "epoch": 0.33082208007581143,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015076755864903632,
+      "loss": 1.1313,
+      "step": 6982
+    },
+    {
+      "epoch": 0.33086946221274577,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001507547273970407,
+      "loss": 1.3182,
+      "step": 6983
+    },
+    {
+      "epoch": 0.33091684434968016,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001507418950193593,
+      "loss": 0.1749,
+      "step": 6984
+    },
+    {
+      "epoch": 0.33096422648661455,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001507290615162767,
+      "loss": 0.5609,
+      "step": 6985
+    },
+    {
+      "epoch": 0.33101160862354895,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00015071622688807757,
+      "loss": 0.0124,
+      "step": 6986
+    },
+    {
+      "epoch": 0.3310589907604833,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001507033911350465,
+      "loss": 0.259,
+      "step": 6987
+    },
+    {
+      "epoch": 0.3311063728974177,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015069055425746827,
+      "loss": 0.8591,
+      "step": 6988
+    },
+    {
+      "epoch": 0.33115375503435207,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015067771625562753,
+      "loss": 1.0551,
+      "step": 6989
+    },
+    {
+      "epoch": 0.3312011371712864,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015066487712980898,
+      "loss": 0.8592,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3312485193082208,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001506520368802975,
+      "loss": 1.0344,
+      "step": 6991
+    },
+    {
+      "epoch": 0.3312959014451552,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00015063919550737772,
+      "loss": 0.0456,
+      "step": 6992
+    },
+    {
+      "epoch": 0.33134328358208953,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015062635301133455,
+      "loss": 0.9304,
+      "step": 6993
+    },
+    {
+      "epoch": 0.3313906657190239,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00015061350939245277,
+      "loss": 0.0861,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3314380478559583,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015060066465101733,
+      "loss": 0.0486,
+      "step": 6995
+    },
+    {
+      "epoch": 0.33148542999289266,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000150587818787313,
+      "loss": 0.8601,
+      "step": 6996
+    },
+    {
+      "epoch": 0.33153281212982705,
+      "grad_norm": 0.00244140625,
+      "learning_rate": 0.0001505749718016247,
+      "loss": 0.0002,
+      "step": 6997
+    },
+    {
+      "epoch": 0.33158019426676144,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015056212369423747,
+      "loss": 1.1144,
+      "step": 6998
+    },
+    {
+      "epoch": 0.33162757640369583,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015054927446543615,
+      "loss": 0.8029,
+      "step": 6999
+    },
+    {
+      "epoch": 0.33167495854063017,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00015053642411550576,
+      "loss": 0.0089,
+      "step": 7000
+    },
+    {
+      "epoch": 0.33172234067756456,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001505235726447313,
+      "loss": 0.497,
+      "step": 7001
+    },
+    {
+      "epoch": 0.33176972281449896,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00015051072005339787,
+      "loss": 1.1711,
+      "step": 7002
+    },
+    {
+      "epoch": 0.3318171049514333,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001504978663417904,
+      "loss": 0.8463,
+      "step": 7003
+    },
+    {
+      "epoch": 0.3318644870883677,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015048501151019412,
+      "loss": 0.9828,
+      "step": 7004
+    },
+    {
+      "epoch": 0.3319118692253021,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000150472155558894,
+      "loss": 1.1101,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3319592513622364,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015045929848817526,
+      "loss": 0.008,
+      "step": 7006
+    },
+    {
+      "epoch": 0.3320066334991708,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.000150446440298323,
+      "loss": 1.2718,
+      "step": 7007
+    },
+    {
+      "epoch": 0.3320540156361052,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00015043358098962246,
+      "loss": 0.179,
+      "step": 7008
+    },
+    {
+      "epoch": 0.33210139777303954,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015042072056235876,
+      "loss": 1.006,
+      "step": 7009
+    },
+    {
+      "epoch": 0.33214877990997393,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.00015040785901681723,
+      "loss": 0.0034,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3321961620469083,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015039499635328306,
+      "loss": 0.2602,
+      "step": 7011
+    },
+    {
+      "epoch": 0.33224354418384267,
+      "grad_norm": 0.01025390625,
+      "learning_rate": 0.0001503821325720416,
+      "loss": 0.0003,
+      "step": 7012
+    },
+    {
+      "epoch": 0.33229092632077706,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00015036926767337803,
+      "loss": 0.096,
+      "step": 7013
+    },
+    {
+      "epoch": 0.33233830845771145,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001503564016575778,
+      "loss": 0.8911,
+      "step": 7014
+    },
+    {
+      "epoch": 0.33238569059464584,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001503435345249262,
+      "loss": 0.9765,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3324330727315802,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00015033066627570863,
+      "loss": 0.9499,
+      "step": 7016
+    },
+    {
+      "epoch": 0.3324804548685146,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015031779691021047,
+      "loss": 0.7292,
+      "step": 7017
+    },
+    {
+      "epoch": 0.33252783700544897,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015030492642871722,
+      "loss": 1.1517,
+      "step": 7018
+    },
+    {
+      "epoch": 0.3325752191423833,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00015029205483151425,
+      "loss": 1.2147,
+      "step": 7019
+    },
+    {
+      "epoch": 0.3326226012793177,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001502791821188871,
+      "loss": 0.977,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3326699834162521,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015026630829112123,
+      "loss": 1.0967,
+      "step": 7021
+    },
+    {
+      "epoch": 0.33271736555318643,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015025343334850217,
+      "loss": 0.9472,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3327647476901208,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015024055729131547,
+      "loss": 0.5209,
+      "step": 7023
+    },
+    {
+      "epoch": 0.3328121298270552,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015022768011984676,
+      "loss": 1.2798,
+      "step": 7024
+    },
+    {
+      "epoch": 0.33285951196398955,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0001502148018343816,
+      "loss": 0.0117,
+      "step": 7025
+    },
+    {
+      "epoch": 0.33290689410092394,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015020192243520562,
+      "loss": 0.8227,
+      "step": 7026
+    },
+    {
+      "epoch": 0.33295427623785834,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00015018904192260445,
+      "loss": 0.1,
+      "step": 7027
+    },
+    {
+      "epoch": 0.33300165837479273,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00015017616029686383,
+      "loss": 0.7774,
+      "step": 7028
+    },
+    {
+      "epoch": 0.33304904051172707,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001501632775582694,
+      "loss": 1.027,
+      "step": 7029
+    },
+    {
+      "epoch": 0.33309642264866146,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00015015039370710692,
+      "loss": 1.0763,
+      "step": 7030
+    },
+    {
+      "epoch": 0.33314380478559585,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015013750874366208,
+      "loss": 0.3201,
+      "step": 7031
+    },
+    {
+      "epoch": 0.3331911869225302,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015012462266822074,
+      "loss": 0.0487,
+      "step": 7032
+    },
+    {
+      "epoch": 0.3332385690594646,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015011173548106865,
+      "loss": 0.0261,
+      "step": 7033
+    },
+    {
+      "epoch": 0.333285951196399,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00015009884718249162,
+      "loss": 0.4067,
+      "step": 7034
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015008595777277556,
+      "loss": 0.0487,
+      "step": 7035
+    },
+    {
+      "epoch": 0.3333807154702677,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0001500730672522063,
+      "loss": 0.0181,
+      "step": 7036
+    },
+    {
+      "epoch": 0.3334280976072021,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00015006017562106973,
+      "loss": 0.772,
+      "step": 7037
+    },
+    {
+      "epoch": 0.33347547974413644,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001500472828796518,
+      "loss": 0.8465,
+      "step": 7038
+    },
+    {
+      "epoch": 0.33352286188107083,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001500343890282384,
+      "loss": 0.7087,
+      "step": 7039
+    },
+    {
+      "epoch": 0.3335702440180052,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015002149406711558,
+      "loss": 0.0318,
+      "step": 7040
+    },
+    {
+      "epoch": 0.33361762615493956,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001500085979965693,
+      "loss": 0.1506,
+      "step": 7041
+    },
+    {
+      "epoch": 0.33366500829187395,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00014999570081688558,
+      "loss": 1.1756,
+      "step": 7042
+    },
+    {
+      "epoch": 0.33371239042880835,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001499828025283505,
+      "loss": 1.231,
+      "step": 7043
+    },
+    {
+      "epoch": 0.33375977256574274,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014996990313125008,
+      "loss": 0.0667,
+      "step": 7044
+    },
+    {
+      "epoch": 0.3338071547026771,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014995700262587043,
+      "loss": 0.8659,
+      "step": 7045
+    },
+    {
+      "epoch": 0.33385453683961147,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00014994410101249766,
+      "loss": 0.8055,
+      "step": 7046
+    },
+    {
+      "epoch": 0.33390191897654586,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014993119829141794,
+      "loss": 0.7001,
+      "step": 7047
+    },
+    {
+      "epoch": 0.3339493011134802,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014991829446291743,
+      "loss": 0.0552,
+      "step": 7048
+    },
+    {
+      "epoch": 0.3339966832504146,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001499053895272823,
+      "loss": 0.6531,
+      "step": 7049
+    },
+    {
+      "epoch": 0.334044065387349,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001498924834847988,
+      "loss": 0.1737,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3340914475242833,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014987957633575316,
+      "loss": 0.5999,
+      "step": 7051
+    },
+    {
+      "epoch": 0.3341388296612177,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00014986666808043165,
+      "loss": 0.149,
+      "step": 7052
+    },
+    {
+      "epoch": 0.3341862117981521,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00014985375871912055,
+      "loss": 0.7358,
+      "step": 7053
+    },
+    {
+      "epoch": 0.33423359393508645,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014984084825210619,
+      "loss": 0.6134,
+      "step": 7054
+    },
+    {
+      "epoch": 0.33428097607202084,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001498279366796749,
+      "loss": 0.8201,
+      "step": 7055
+    },
+    {
+      "epoch": 0.33432835820895523,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00014981502400211304,
+      "loss": 0.0103,
+      "step": 7056
+    },
+    {
+      "epoch": 0.3343757403458896,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014980211021970702,
+      "loss": 1.4085,
+      "step": 7057
+    },
+    {
+      "epoch": 0.33442312248282396,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014978919533274322,
+      "loss": 1.0654,
+      "step": 7058
+    },
+    {
+      "epoch": 0.33447050461975836,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00014977627934150816,
+      "loss": 0.9682,
+      "step": 7059
+    },
+    {
+      "epoch": 0.33451788675669275,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014976336224628822,
+      "loss": 0.8443,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3345652688936271,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014975044404736987,
+      "loss": 0.4145,
+      "step": 7061
+    },
+    {
+      "epoch": 0.3346126510305615,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014973752474503968,
+      "loss": 1.3228,
+      "step": 7062
+    },
+    {
+      "epoch": 0.3346600331674959,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014972460433958419,
+      "loss": 0.8524,
+      "step": 7063
+    },
+    {
+      "epoch": 0.3347074153044302,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00014971168283128993,
+      "loss": 0.8595,
+      "step": 7064
+    },
+    {
+      "epoch": 0.3347547974413646,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00014969876022044346,
+      "loss": 0.6905,
+      "step": 7065
+    },
+    {
+      "epoch": 0.334802179578299,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001496858365073315,
+      "loss": 0.9029,
+      "step": 7066
+    },
+    {
+      "epoch": 0.33484956171523333,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014967291169224058,
+      "loss": 0.9739,
+      "step": 7067
+    },
+    {
+      "epoch": 0.33489694385216773,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00014965998577545734,
+      "loss": 0.0078,
+      "step": 7068
+    },
+    {
+      "epoch": 0.3349443259891021,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014964705875726857,
+      "loss": 0.0475,
+      "step": 7069
+    },
+    {
+      "epoch": 0.33499170812603646,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00014963413063796087,
+      "loss": 0.0081,
+      "step": 7070
+    },
+    {
+      "epoch": 0.33503909026297085,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00014962120141782104,
+      "loss": 0.3946,
+      "step": 7071
+    },
+    {
+      "epoch": 0.33508647239990524,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001496082710971358,
+      "loss": 1.0272,
+      "step": 7072
+    },
+    {
+      "epoch": 0.33513385453683964,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.000149595339676192,
+      "loss": 0.0063,
+      "step": 7073
+    },
+    {
+      "epoch": 0.335181236673774,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014958240715527636,
+      "loss": 1.2715,
+      "step": 7074
+    },
+    {
+      "epoch": 0.33522861881070837,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.00014956947353467578,
+      "loss": 0.0058,
+      "step": 7075
+    },
+    {
+      "epoch": 0.33527600094764276,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014955653881467703,
+      "loss": 1.415,
+      "step": 7076
+    },
+    {
+      "epoch": 0.3353233830845771,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00014954360299556703,
+      "loss": 0.0029,
+      "step": 7077
+    },
+    {
+      "epoch": 0.3353707652215115,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014953066607763268,
+      "loss": 0.0616,
+      "step": 7078
+    },
+    {
+      "epoch": 0.3354181473584459,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00014951772806116095,
+      "loss": 1.2487,
+      "step": 7079
+    },
+    {
+      "epoch": 0.3354655294953802,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014950478894643873,
+      "loss": 1.2876,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3355129116323146,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000149491848733753,
+      "loss": 0.9256,
+      "step": 7081
+    },
+    {
+      "epoch": 0.335560293769249,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014947890742339086,
+      "loss": 0.678,
+      "step": 7082
+    },
+    {
+      "epoch": 0.33560767590618334,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001494659650156392,
+      "loss": 1.1603,
+      "step": 7083
+    },
+    {
+      "epoch": 0.33565505804311774,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014945302151078512,
+      "loss": 0.7622,
+      "step": 7084
+    },
+    {
+      "epoch": 0.33570244018005213,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014944007690911572,
+      "loss": 0.8894,
+      "step": 7085
+    },
+    {
+      "epoch": 0.3357498223169865,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0001494271312109181,
+      "loss": 0.001,
+      "step": 7086
+    },
+    {
+      "epoch": 0.33579720445392086,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001494141844164793,
+      "loss": 1.2383,
+      "step": 7087
+    },
+    {
+      "epoch": 0.33584458659085525,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00014940123652608652,
+      "loss": 0.1373,
+      "step": 7088
+    },
+    {
+      "epoch": 0.33589196872778965,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014938828754002697,
+      "loss": 0.508,
+      "step": 7089
+    },
+    {
+      "epoch": 0.335939350864724,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014937533745858783,
+      "loss": 1.205,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3359867330016584,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014936238628205625,
+      "loss": 1.1114,
+      "step": 7091
+    },
+    {
+      "epoch": 0.33603411513859277,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014934943401071954,
+      "loss": 0.8703,
+      "step": 7092
+    },
+    {
+      "epoch": 0.3360814972755271,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00014933648064486494,
+      "loss": 0.1389,
+      "step": 7093
+    },
+    {
+      "epoch": 0.3361288794124615,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00014932352618477976,
+      "loss": 1.1597,
+      "step": 7094
+    },
+    {
+      "epoch": 0.3361762615493959,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001493105706307513,
+      "loss": 0.1503,
+      "step": 7095
+    },
+    {
+      "epoch": 0.33622364368633023,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001492976139830669,
+      "loss": 0.5618,
+      "step": 7096
+    },
+    {
+      "epoch": 0.3362710258232646,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014928465624201395,
+      "loss": 0.1363,
+      "step": 7097
+    },
+    {
+      "epoch": 0.336318407960199,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014927169740787982,
+      "loss": 0.0229,
+      "step": 7098
+    },
+    {
+      "epoch": 0.33636579009713335,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014925873748095192,
+      "loss": 1.0774,
+      "step": 7099
+    },
+    {
+      "epoch": 0.33641317223406775,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014924577646151767,
+      "loss": 0.7773,
+      "step": 7100
+    },
+    {
+      "epoch": 0.33646055437100214,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014923281434986457,
+      "loss": 0.811,
+      "step": 7101
+    },
+    {
+      "epoch": 0.33650793650793653,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0001492198511462801,
+      "loss": 0.0017,
+      "step": 7102
+    },
+    {
+      "epoch": 0.33655531864487087,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00014920688685105172,
+      "loss": 0.0042,
+      "step": 7103
+    },
+    {
+      "epoch": 0.33660270078180526,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014919392146446703,
+      "loss": 1.039,
+      "step": 7104
+    },
+    {
+      "epoch": 0.33665008291873966,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014918095498681356,
+      "loss": 1.2575,
+      "step": 7105
+    },
+    {
+      "epoch": 0.336697465055674,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001491679874183789,
+      "loss": 0.6664,
+      "step": 7106
+    },
+    {
+      "epoch": 0.3367448471926084,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014915501875945064,
+      "loss": 0.9582,
+      "step": 7107
+    },
+    {
+      "epoch": 0.3367922293295428,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001491420490103164,
+      "loss": 0.6622,
+      "step": 7108
+    },
+    {
+      "epoch": 0.3368396114664771,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001491290781712639,
+      "loss": 1.2484,
+      "step": 7109
+    },
+    {
+      "epoch": 0.3368869936034115,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014911610624258076,
+      "loss": 1.2539,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3369343757403459,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014910313322455466,
+      "loss": 0.5703,
+      "step": 7111
+    },
+    {
+      "epoch": 0.33698175787728024,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00014909015911747343,
+      "loss": 0.0237,
+      "step": 7112
+    },
+    {
+      "epoch": 0.33702914001421463,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014907718392162474,
+      "loss": 1.2957,
+      "step": 7113
+    },
+    {
+      "epoch": 0.337076522151149,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014906420763729638,
+      "loss": 0.5867,
+      "step": 7114
+    },
+    {
+      "epoch": 0.3371239042880834,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00014905123026477614,
+      "loss": 0.0653,
+      "step": 7115
+    },
+    {
+      "epoch": 0.33717128642501776,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014903825180435186,
+      "loss": 1.0758,
+      "step": 7116
+    },
+    {
+      "epoch": 0.33721866856195215,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001490252722563114,
+      "loss": 0.4744,
+      "step": 7117
+    },
+    {
+      "epoch": 0.33726605069888654,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014901229162094263,
+      "loss": 0.1012,
+      "step": 7118
+    },
+    {
+      "epoch": 0.3373134328358209,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014899930989853343,
+      "loss": 0.8182,
+      "step": 7119
+    },
+    {
+      "epoch": 0.3373608149727553,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014898632708937171,
+      "loss": 1.108,
+      "step": 7120
+    },
+    {
+      "epoch": 0.33740819710968967,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014897334319374545,
+      "loss": 0.9846,
+      "step": 7121
+    },
+    {
+      "epoch": 0.337455579246624,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014896035821194262,
+      "loss": 0.0365,
+      "step": 7122
+    },
+    {
+      "epoch": 0.3375029613835584,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014894737214425117,
+      "loss": 0.9856,
+      "step": 7123
+    },
+    {
+      "epoch": 0.3375503435204928,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014893438499095915,
+      "loss": 0.0867,
+      "step": 7124
+    },
+    {
+      "epoch": 0.33759772565742713,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014892139675235462,
+      "loss": 0.9966,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3376451077943615,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00014890840742872555,
+      "loss": 0.0097,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3376924899312959,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014889541702036013,
+      "loss": 1.1499,
+      "step": 7127
+    },
+    {
+      "epoch": 0.33773987206823025,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014888242552754647,
+      "loss": 0.6494,
+      "step": 7128
+    },
+    {
+      "epoch": 0.33778725420516464,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.00014886943295057265,
+      "loss": 0.0051,
+      "step": 7129
+    },
+    {
+      "epoch": 0.33783463634209904,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001488564392897269,
+      "loss": 0.1142,
+      "step": 7130
+    },
+    {
+      "epoch": 0.33788201847903343,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014884344454529734,
+      "loss": 0.6924,
+      "step": 7131
+    },
+    {
+      "epoch": 0.33792940061596777,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014883044871757218,
+      "loss": 0.8983,
+      "step": 7132
+    },
+    {
+      "epoch": 0.33797678275290216,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001488174518068397,
+      "loss": 0.622,
+      "step": 7133
+    },
+    {
+      "epoch": 0.33802416488983655,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014880445381338815,
+      "loss": 1.1508,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3380715470267709,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00014879145473750577,
+      "loss": 0.0232,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3381189291637053,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014877845457948093,
+      "loss": 0.7931,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3381663113006397,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001487654533396019,
+      "loss": 1.0485,
+      "step": 7137
+    },
+    {
+      "epoch": 0.338213693437574,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014875245101815708,
+      "loss": 0.4988,
+      "step": 7138
+    },
+    {
+      "epoch": 0.3382610755745084,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001487394476154348,
+      "loss": 0.9374,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3383084577114428,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001487264431317235,
+      "loss": 1.1444,
+      "step": 7140
+    },
+    {
+      "epoch": 0.33835583984837714,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00014871343756731156,
+      "loss": 0.5691,
+      "step": 7141
+    },
+    {
+      "epoch": 0.33840322198531153,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014870043092248748,
+      "loss": 1.2617,
+      "step": 7142
+    },
+    {
+      "epoch": 0.3384506041222459,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014868742319753975,
+      "loss": 1.0233,
+      "step": 7143
+    },
+    {
+      "epoch": 0.33849798625918026,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001486744143927568,
+      "loss": 1.2125,
+      "step": 7144
+    },
+    {
+      "epoch": 0.33854536839611465,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014866140450842718,
+      "loss": 1.0453,
+      "step": 7145
+    },
+    {
+      "epoch": 0.33859275053304905,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014864839354483946,
+      "loss": 0.8343,
+      "step": 7146
+    },
+    {
+      "epoch": 0.33864013266998344,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014863538150228217,
+      "loss": 0.6355,
+      "step": 7147
+    },
+    {
+      "epoch": 0.3386875148069178,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014862236838104396,
+      "loss": 1.1156,
+      "step": 7148
+    },
+    {
+      "epoch": 0.33873489694385217,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014860935418141338,
+      "loss": 1.0201,
+      "step": 7149
+    },
+    {
+      "epoch": 0.33878227908078656,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014859633890367907,
+      "loss": 0.5542,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3388296612177209,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 0.0001485833225481298,
+      "loss": 0.0045,
+      "step": 7151
+    },
+    {
+      "epoch": 0.3388770433546553,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014857030511505412,
+      "loss": 1.0131,
+      "step": 7152
+    },
+    {
+      "epoch": 0.3389244254915897,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00014855728660474084,
+      "loss": 0.0299,
+      "step": 7153
+    },
+    {
+      "epoch": 0.338971807628524,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014854426701747865,
+      "loss": 0.1357,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3390191897654584,
+      "grad_norm": 0.031494140625,
+      "learning_rate": 0.00014853124635355632,
+      "loss": 0.0032,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3390665719023928,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014851822461326266,
+      "loss": 0.3549,
+      "step": 7156
+    },
+    {
+      "epoch": 0.33911395403932715,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00014850520179688644,
+      "loss": 0.4451,
+      "step": 7157
+    },
+    {
+      "epoch": 0.33916133617626154,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001484921779047165,
+      "loss": 0.0741,
+      "step": 7158
+    },
+    {
+      "epoch": 0.33920871831319593,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014847915293704172,
+      "loss": 1.5142,
+      "step": 7159
+    },
+    {
+      "epoch": 0.3392561004501303,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 0.000148466126894151,
+      "loss": 0.0014,
+      "step": 7160
+    },
+    {
+      "epoch": 0.33930348258706466,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014845309977633316,
+      "loss": 1.29,
+      "step": 7161
+    },
+    {
+      "epoch": 0.33935086472399906,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00014844007158387718,
+      "loss": 0.0123,
+      "step": 7162
+    },
+    {
+      "epoch": 0.33939824686093345,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014842704231707204,
+      "loss": 0.9652,
+      "step": 7163
+    },
+    {
+      "epoch": 0.3394456289978678,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014841401197620665,
+      "loss": 1.042,
+      "step": 7164
+    },
+    {
+      "epoch": 0.3394930111348022,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00014840098056157007,
+      "loss": 0.9504,
+      "step": 7165
+    },
+    {
+      "epoch": 0.3395403932717366,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00014838794807345128,
+      "loss": 1.1345,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3395877754086709,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014837491451213933,
+      "loss": 0.0718,
+      "step": 7167
+    },
+    {
+      "epoch": 0.3396351575456053,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014836187987792333,
+      "loss": 0.5195,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3396825396825397,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014834884417109235,
+      "loss": 0.9032,
+      "step": 7169
+    },
+    {
+      "epoch": 0.33972992181947403,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014833580739193546,
+      "loss": 1.1558,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3397773039564084,
+      "grad_norm": 0.004119873046875,
+      "learning_rate": 0.00014832276954074191,
+      "loss": 0.0002,
+      "step": 7171
+    },
+    {
+      "epoch": 0.3398246860933428,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014830973061780075,
+      "loss": 0.5184,
+      "step": 7172
+    },
+    {
+      "epoch": 0.33987206823027716,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014829669062340123,
+      "loss": 0.9264,
+      "step": 7173
+    },
+    {
+      "epoch": 0.33991945036721155,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014828364955783257,
+      "loss": 0.8504,
+      "step": 7174
+    },
+    {
+      "epoch": 0.33996683250414594,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014827060742138399,
+      "loss": 1.2405,
+      "step": 7175
+    },
+    {
+      "epoch": 0.34001421464108034,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00014825756421434476,
+      "loss": 0.007,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3400615967780147,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014824451993700416,
+      "loss": 0.8505,
+      "step": 7177
+    },
+    {
+      "epoch": 0.34010897891494907,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001482314745896515,
+      "loss": 0.6473,
+      "step": 7178
+    },
+    {
+      "epoch": 0.34015636105188346,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00014821842817257607,
+      "loss": 1.3722,
+      "step": 7179
+    },
+    {
+      "epoch": 0.3402037431888178,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014820538068606727,
+      "loss": 0.9416,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3402511253257522,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014819233213041451,
+      "loss": 0.0131,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3402985074626866,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00014817928250590714,
+      "loss": 0.0351,
+      "step": 7182
+    },
+    {
+      "epoch": 0.3403458895996209,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014816623181283458,
+      "loss": 1.075,
+      "step": 7183
+    },
+    {
+      "epoch": 0.3403932717365553,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001481531800514863,
+      "loss": 1.0367,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3404406538734897,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014814012722215181,
+      "loss": 0.8159,
+      "step": 7185
+    },
+    {
+      "epoch": 0.34048803601042404,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001481270733251206,
+      "loss": 0.8916,
+      "step": 7186
+    },
+    {
+      "epoch": 0.34053541814735844,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00014811401836068211,
+      "loss": 0.7287,
+      "step": 7187
+    },
+    {
+      "epoch": 0.34058280028429283,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014810096232912594,
+      "loss": 0.1533,
+      "step": 7188
+    },
+    {
+      "epoch": 0.3406301824212272,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001480879052307417,
+      "loss": 0.0744,
+      "step": 7189
+    },
+    {
+      "epoch": 0.34067756455816156,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001480748470658189,
+      "loss": 0.9296,
+      "step": 7190
+    },
+    {
+      "epoch": 0.34072494669509595,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014806178783464722,
+      "loss": 0.6243,
+      "step": 7191
+    },
+    {
+      "epoch": 0.34077232883203035,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014804872753751625,
+      "loss": 0.5635,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3408197109689647,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00014803566617471573,
+      "loss": 0.1587,
+      "step": 7193
+    },
+    {
+      "epoch": 0.3408670931058991,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001480226037465353,
+      "loss": 0.0075,
+      "step": 7194
+    },
+    {
+      "epoch": 0.34091447524283347,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00014800954025326465,
+      "loss": 0.1212,
+      "step": 7195
+    },
+    {
+      "epoch": 0.3409618573797678,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00014799647569519353,
+      "loss": 1.1877,
+      "step": 7196
+    },
+    {
+      "epoch": 0.3410092395167022,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00014798341007261171,
+      "loss": 1.0534,
+      "step": 7197
+    },
+    {
+      "epoch": 0.3410566216536366,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014797034338580897,
+      "loss": 1.1359,
+      "step": 7198
+    },
+    {
+      "epoch": 0.34110400379057093,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001479572756350751,
+      "loss": 0.2054,
+      "step": 7199
+    },
+    {
+      "epoch": 0.3411513859275053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014794420682069995,
+      "loss": 0.3547,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3411987680644397,
+      "grad_norm": 0.0225830078125,
+      "learning_rate": 0.00014793113694297336,
+      "loss": 0.0009,
+      "step": 7201
+    },
+    {
+      "epoch": 0.34124615020137405,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001479180660021852,
+      "loss": 1.2018,
+      "step": 7202
+    },
+    {
+      "epoch": 0.34129353233830845,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014790499399862535,
+      "loss": 0.0145,
+      "step": 7203
+    },
+    {
+      "epoch": 0.34134091447524284,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00014789192093258378,
+      "loss": 0.9923,
+      "step": 7204
+    },
+    {
+      "epoch": 0.34138829661217723,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014787884680435043,
+      "loss": 0.5752,
+      "step": 7205
+    },
+    {
+      "epoch": 0.34143567874911157,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014786577161421524,
+      "loss": 1.3934,
+      "step": 7206
+    },
+    {
+      "epoch": 0.34148306088604596,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014785269536246823,
+      "loss": 1.1021,
+      "step": 7207
+    },
+    {
+      "epoch": 0.34153044302298036,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001478396180493994,
+      "loss": 1.4513,
+      "step": 7208
+    },
+    {
+      "epoch": 0.3415778251599147,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014782653967529882,
+      "loss": 0.8498,
+      "step": 7209
+    },
+    {
+      "epoch": 0.3416252072968491,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001478134602404565,
+      "loss": 0.5001,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3416725894337835,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00014780037974516258,
+      "loss": 0.8522,
+      "step": 7211
+    },
+    {
+      "epoch": 0.3417199715707178,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014778729818970714,
+      "loss": 0.0263,
+      "step": 7212
+    },
+    {
+      "epoch": 0.3417673537076522,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014777421557438033,
+      "loss": 1.4596,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3418147358445866,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001477611318994723,
+      "loss": 0.9247,
+      "step": 7214
+    },
+    {
+      "epoch": 0.34186211798152094,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014774804716527324,
+      "loss": 1.1011,
+      "step": 7215
+    },
+    {
+      "epoch": 0.34190950011845533,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00014773496137207337,
+      "loss": 0.4848,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3419568822553897,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001477218745201629,
+      "loss": 0.1409,
+      "step": 7217
+    },
+    {
+      "epoch": 0.3420042643923241,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014770878660983207,
+      "loss": 1.2408,
+      "step": 7218
+    },
+    {
+      "epoch": 0.34205164652925846,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00014769569764137117,
+      "loss": 0.1452,
+      "step": 7219
+    },
+    {
+      "epoch": 0.34209902866619285,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001476826076150705,
+      "loss": 0.0404,
+      "step": 7220
+    },
+    {
+      "epoch": 0.34214641080312724,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001476695165312204,
+      "loss": 1.304,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3421937929400616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014765642439011116,
+      "loss": 1.162,
+      "step": 7222
+    },
+    {
+      "epoch": 0.342241175076996,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001476433311920332,
+      "loss": 1.0679,
+      "step": 7223
+    },
+    {
+      "epoch": 0.34228855721393037,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00014763023693727695,
+      "loss": 0.1273,
+      "step": 7224
+    },
+    {
+      "epoch": 0.3423359393508647,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014761714162613273,
+      "loss": 1.2374,
+      "step": 7225
+    },
+    {
+      "epoch": 0.3423833214877991,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014760404525889105,
+      "loss": 0.7444,
+      "step": 7226
+    },
+    {
+      "epoch": 0.3424307036247335,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014759094783584233,
+      "loss": 0.9537,
+      "step": 7227
+    },
+    {
+      "epoch": 0.3424780857616678,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001475778493572771,
+      "loss": 0.3954,
+      "step": 7228
+    },
+    {
+      "epoch": 0.3425254678986022,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014756474982348584,
+      "loss": 0.3159,
+      "step": 7229
+    },
+    {
+      "epoch": 0.3425728500355366,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014755164923475908,
+      "loss": 1.084,
+      "step": 7230
+    },
+    {
+      "epoch": 0.34262023217247095,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014753854759138742,
+      "loss": 0.9867,
+      "step": 7231
+    },
+    {
+      "epoch": 0.34266761430940534,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001475254448936614,
+      "loss": 0.089,
+      "step": 7232
+    },
+    {
+      "epoch": 0.34271499644633974,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001475123411418716,
+      "loss": 0.9966,
+      "step": 7233
+    },
+    {
+      "epoch": 0.34276237858327413,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014749923633630872,
+      "loss": 0.0149,
+      "step": 7234
+    },
+    {
+      "epoch": 0.34280976072020847,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014748613047726336,
+      "loss": 1.4929,
+      "step": 7235
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00014747302356502622,
+      "loss": 0.3428,
+      "step": 7236
+    },
+    {
+      "epoch": 0.34290452499407725,
+      "grad_norm": 0.0027923583984375,
+      "learning_rate": 0.00014745991559988794,
+      "loss": 0.0002,
+      "step": 7237
+    },
+    {
+      "epoch": 0.3429519071310116,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014744680658213932,
+      "loss": 1.5044,
+      "step": 7238
+    },
+    {
+      "epoch": 0.342999289267946,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014743369651207106,
+      "loss": 1.1535,
+      "step": 7239
+    },
+    {
+      "epoch": 0.3430466714048804,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014742058538997393,
+      "loss": 1.421,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3430940535418147,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014740747321613875,
+      "loss": 1.2337,
+      "step": 7241
+    },
+    {
+      "epoch": 0.3431414356787491,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001473943599908563,
+      "loss": 1.1612,
+      "step": 7242
+    },
+    {
+      "epoch": 0.3431888178156835,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00014738124571441743,
+      "loss": 0.0057,
+      "step": 7243
+    },
+    {
+      "epoch": 0.34323619995261784,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00014736813038711296,
+      "loss": 0.3686,
+      "step": 7244
+    },
+    {
+      "epoch": 0.34328358208955223,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00014735501400923385,
+      "loss": 0.0485,
+      "step": 7245
+    },
+    {
+      "epoch": 0.3433309642264866,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00014734189658107094,
+      "loss": 1.0313,
+      "step": 7246
+    },
+    {
+      "epoch": 0.343378346363421,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014732877810291523,
+      "loss": 0.9282,
+      "step": 7247
+    },
+    {
+      "epoch": 0.34342572850035535,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014731565857505763,
+      "loss": 0.6842,
+      "step": 7248
+    },
+    {
+      "epoch": 0.34347311063728975,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001473025379977891,
+      "loss": 0.2441,
+      "step": 7249
+    },
+    {
+      "epoch": 0.34352049277422414,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014728941637140065,
+      "loss": 0.7064,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3435678749111585,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014727629369618334,
+      "loss": 0.5426,
+      "step": 7251
+    },
+    {
+      "epoch": 0.34361525704809287,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014726316997242818,
+      "loss": 1.2039,
+      "step": 7252
+    },
+    {
+      "epoch": 0.34366263918502726,
+      "grad_norm": 0.01708984375,
+      "learning_rate": 0.00014725004520042628,
+      "loss": 0.0011,
+      "step": 7253
+    },
+    {
+      "epoch": 0.3437100213219616,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014723691938046867,
+      "loss": 1.3481,
+      "step": 7254
+    },
+    {
+      "epoch": 0.343757403458896,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014722379251284654,
+      "loss": 1.3633,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3438047855958304,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014721066459785095,
+      "loss": 0.1503,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3438521677327647,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014719753563577313,
+      "loss": 1.1006,
+      "step": 7257
+    },
+    {
+      "epoch": 0.3438995498696991,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014718440562690424,
+      "loss": 0.5793,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3439469320066335,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001471712745715355,
+      "loss": 1.1158,
+      "step": 7259
+    },
+    {
+      "epoch": 0.34399431414356785,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001471581424699581,
+      "loss": 0.9373,
+      "step": 7260
+    },
+    {
+      "epoch": 0.34404169628050224,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014714500932246332,
+      "loss": 0.1639,
+      "step": 7261
+    },
+    {
+      "epoch": 0.34408907841743663,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00014713187512934253,
+      "loss": 0.1281,
+      "step": 7262
+    },
+    {
+      "epoch": 0.344136460554371,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001471187398908869,
+      "loss": 1.3557,
+      "step": 7263
+    },
+    {
+      "epoch": 0.34418384269130536,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001471056036073878,
+      "loss": 1.4961,
+      "step": 7264
+    },
+    {
+      "epoch": 0.34423122482823976,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001470924662791366,
+      "loss": 0.7998,
+      "step": 7265
+    },
+    {
+      "epoch": 0.34427860696517415,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001470793279064247,
+      "loss": 0.7616,
+      "step": 7266
+    },
+    {
+      "epoch": 0.3443259891021085,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001470661884895434,
+      "loss": 0.5415,
+      "step": 7267
+    },
+    {
+      "epoch": 0.3443733712390429,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014705304802878417,
+      "loss": 0.7378,
+      "step": 7268
+    },
+    {
+      "epoch": 0.3444207533759773,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014703990652443845,
+      "loss": 0.9451,
+      "step": 7269
+    },
+    {
+      "epoch": 0.3444681355129116,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014702676397679776,
+      "loss": 0.3388,
+      "step": 7270
+    },
+    {
+      "epoch": 0.344515517649846,
+      "grad_norm": 0.0196533203125,
+      "learning_rate": 0.00014701362038615348,
+      "loss": 0.0011,
+      "step": 7271
+    },
+    {
+      "epoch": 0.3445628997867804,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001470004757527972,
+      "loss": 1.2691,
+      "step": 7272
+    },
+    {
+      "epoch": 0.34461028192371473,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014698733007702044,
+      "loss": 0.8237,
+      "step": 7273
+    },
+    {
+      "epoch": 0.3446576640606491,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00014697418335911472,
+      "loss": 0.065,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3447050461975835,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014696103559937165,
+      "loss": 1.1218,
+      "step": 7275
+    },
+    {
+      "epoch": 0.3447524283345179,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00014694788679808286,
+      "loss": 0.1589,
+      "step": 7276
+    },
+    {
+      "epoch": 0.34479981047145225,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00014693473695553995,
+      "loss": 0.1621,
+      "step": 7277
+    },
+    {
+      "epoch": 0.34484719260838664,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014692158607203454,
+      "loss": 0.9023,
+      "step": 7278
+    },
+    {
+      "epoch": 0.34489457474532104,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014690843414785835,
+      "loss": 0.8748,
+      "step": 7279
+    },
+    {
+      "epoch": 0.3449419568822554,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00014689528118330304,
+      "loss": 0.1103,
+      "step": 7280
+    },
+    {
+      "epoch": 0.34498933901918977,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00014688212717866038,
+      "loss": 0.0525,
+      "step": 7281
+    },
+    {
+      "epoch": 0.34503672115612416,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.000146868972134222,
+      "loss": 0.008,
+      "step": 7282
+    },
+    {
+      "epoch": 0.3450841032930585,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014685581605027978,
+      "loss": 0.9089,
+      "step": 7283
+    },
+    {
+      "epoch": 0.3451314854299929,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014684265892712548,
+      "loss": 0.2723,
+      "step": 7284
+    },
+    {
+      "epoch": 0.3451788675669273,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0001468295007650509,
+      "loss": 1.4206,
+      "step": 7285
+    },
+    {
+      "epoch": 0.3452262497038616,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014681634156434785,
+      "loss": 1.3776,
+      "step": 7286
+    },
+    {
+      "epoch": 0.345273631840796,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014680318132530827,
+      "loss": 1.148,
+      "step": 7287
+    },
+    {
+      "epoch": 0.3453210139777304,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001467900200482239,
+      "loss": 0.6705,
+      "step": 7288
+    },
+    {
+      "epoch": 0.34536839611466474,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014677685773338678,
+      "loss": 0.72,
+      "step": 7289
+    },
+    {
+      "epoch": 0.34541577825159914,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014676369438108874,
+      "loss": 0.0097,
+      "step": 7290
+    },
+    {
+      "epoch": 0.34546316038853353,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014675052999162179,
+      "loss": 1.5282,
+      "step": 7291
+    },
+    {
+      "epoch": 0.3455105425254679,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00014673736456527787,
+      "loss": 1.3205,
+      "step": 7292
+    },
+    {
+      "epoch": 0.34555792466240226,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014672419810234902,
+      "loss": 1.3132,
+      "step": 7293
+    },
+    {
+      "epoch": 0.34560530679933665,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00014671103060312718,
+      "loss": 0.0965,
+      "step": 7294
+    },
+    {
+      "epoch": 0.34565268893627105,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00014669786206790447,
+      "loss": 0.1393,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3457000710732054,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001466846924969729,
+      "loss": 1.0978,
+      "step": 7296
+    },
+    {
+      "epoch": 0.3457474532101398,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00014667152189062462,
+      "loss": 1.045,
+      "step": 7297
+    },
+    {
+      "epoch": 0.34579483534707417,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014665835024915165,
+      "loss": 0.7312,
+      "step": 7298
+    },
+    {
+      "epoch": 0.3458422174840085,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014664517757284617,
+      "loss": 0.732,
+      "step": 7299
+    },
+    {
+      "epoch": 0.3458895996209429,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00014663200386200035,
+      "loss": 0.8799,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3459369817578773,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014661882911690634,
+      "loss": 0.5094,
+      "step": 7301
+    },
+    {
+      "epoch": 0.34598436389481163,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00014660565333785637,
+      "loss": 0.969,
+      "step": 7302
+    },
+    {
+      "epoch": 0.346031746031746,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00014659247652514266,
+      "loss": 0.02,
+      "step": 7303
+    },
+    {
+      "epoch": 0.3460791281686804,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014657929867905746,
+      "loss": 0.5696,
+      "step": 7304
+    },
+    {
+      "epoch": 0.3461265103056148,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014656611979989298,
+      "loss": 0.1315,
+      "step": 7305
+    },
+    {
+      "epoch": 0.34617389244254915,
+      "grad_norm": 0.0283203125,
+      "learning_rate": 0.00014655293988794158,
+      "loss": 0.0009,
+      "step": 7306
+    },
+    {
+      "epoch": 0.34622127457948354,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001465397589434956,
+      "loss": 0.9577,
+      "step": 7307
+    },
+    {
+      "epoch": 0.34626865671641793,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001465265769668473,
+      "loss": 1.1871,
+      "step": 7308
+    },
+    {
+      "epoch": 0.34631603885335227,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014651339395828906,
+      "loss": 0.1278,
+      "step": 7309
+    },
+    {
+      "epoch": 0.34636342099028666,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014650020991811334,
+      "loss": 1.3712,
+      "step": 7310
+    },
+    {
+      "epoch": 0.34641080312722106,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014648702484661245,
+      "loss": 1.0471,
+      "step": 7311
+    },
+    {
+      "epoch": 0.3464581852641554,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001464738387440789,
+      "loss": 0.0317,
+      "step": 7312
+    },
+    {
+      "epoch": 0.3465055674010898,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00014646065161080509,
+      "loss": 0.14,
+      "step": 7313
+    },
+    {
+      "epoch": 0.3465529495380242,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00014644746344708351,
+      "loss": 0.913,
+      "step": 7314
+    },
+    {
+      "epoch": 0.3466003316749585,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014643427425320665,
+      "loss": 0.3591,
+      "step": 7315
+    },
+    {
+      "epoch": 0.3466477138118929,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00014642108402946707,
+      "loss": 1.342,
+      "step": 7316
+    },
+    {
+      "epoch": 0.3466950959488273,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0001464078927761573,
+      "loss": 0.0029,
+      "step": 7317
+    },
+    {
+      "epoch": 0.34674247808576164,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.00014639470049356988,
+      "loss": 0.0056,
+      "step": 7318
+    },
+    {
+      "epoch": 0.34678986022269603,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014638150718199744,
+      "loss": 0.0129,
+      "step": 7319
+    },
+    {
+      "epoch": 0.3468372423596304,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014636831284173257,
+      "loss": 0.406,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3468846244965648,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001463551174730679,
+      "loss": 0.7431,
+      "step": 7321
+    },
+    {
+      "epoch": 0.34693200663349916,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001463419210762961,
+      "loss": 0.4843,
+      "step": 7322
+    },
+    {
+      "epoch": 0.34697938877043355,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014632872365170986,
+      "loss": 1.1616,
+      "step": 7323
+    },
+    {
+      "epoch": 0.34702677090736794,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00014631552519960185,
+      "loss": 0.9166,
+      "step": 7324
+    },
+    {
+      "epoch": 0.3470741530443023,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00014630232572026484,
+      "loss": 1.215,
+      "step": 7325
+    },
+    {
+      "epoch": 0.3471215351812367,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001462891252139916,
+      "loss": 0.1048,
+      "step": 7326
+    },
+    {
+      "epoch": 0.34716891731817107,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014627592368107484,
+      "loss": 0.485,
+      "step": 7327
+    },
+    {
+      "epoch": 0.3472162994551054,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00014626272112180737,
+      "loss": 1.0416,
+      "step": 7328
+    },
+    {
+      "epoch": 0.3472636815920398,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014624951753648203,
+      "loss": 0.5534,
+      "step": 7329
+    },
+    {
+      "epoch": 0.3473110637289742,
+      "grad_norm": 0.0162353515625,
+      "learning_rate": 0.00014623631292539163,
+      "loss": 0.0008,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3473584458659085,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014622310728882912,
+      "loss": 1.2545,
+      "step": 7331
+    },
+    {
+      "epoch": 0.3474058280028429,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001462099006270873,
+      "loss": 0.1492,
+      "step": 7332
+    },
+    {
+      "epoch": 0.3474532101397773,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00014619669294045905,
+      "loss": 0.9416,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3475005922767117,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014618348422923742,
+      "loss": 1.1713,
+      "step": 7334
+    },
+    {
+      "epoch": 0.34754797441364604,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00014617027449371532,
+      "loss": 0.0498,
+      "step": 7335
+    },
+    {
+      "epoch": 0.34759535655058044,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014615706373418566,
+      "loss": 0.1106,
+      "step": 7336
+    },
+    {
+      "epoch": 0.34764273868751483,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001461438519509415,
+      "loss": 1.1389,
+      "step": 7337
+    },
+    {
+      "epoch": 0.34769012082444917,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014613063914427585,
+      "loss": 0.9827,
+      "step": 7338
+    },
+    {
+      "epoch": 0.34773750296138356,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001461174253144818,
+      "loss": 0.7366,
+      "step": 7339
+    },
+    {
+      "epoch": 0.34778488509831795,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014610421046185233,
+      "loss": 0.0852,
+      "step": 7340
+    },
+    {
+      "epoch": 0.3478322672352523,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001460909945866806,
+      "loss": 0.8307,
+      "step": 7341
+    },
+    {
+      "epoch": 0.3478796493721867,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001460777776892597,
+      "loss": 1.0301,
+      "step": 7342
+    },
+    {
+      "epoch": 0.3479270315091211,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001460645597698828,
+      "loss": 1.4619,
+      "step": 7343
+    },
+    {
+      "epoch": 0.3479744136460554,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014605134082884295,
+      "loss": 1.3222,
+      "step": 7344
+    },
+    {
+      "epoch": 0.3480217957829898,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00014603812086643348,
+      "loss": 1.4361,
+      "step": 7345
+    },
+    {
+      "epoch": 0.3480691779199242,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001460248998829475,
+      "loss": 0.4261,
+      "step": 7346
+    },
+    {
+      "epoch": 0.34811656005685854,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00014601167787867827,
+      "loss": 0.9054,
+      "step": 7347
+    },
+    {
+      "epoch": 0.34816394219379293,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00014599845485391906,
+      "loss": 0.8762,
+      "step": 7348
+    },
+    {
+      "epoch": 0.3482113243307273,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014598523080896307,
+      "loss": 0.1223,
+      "step": 7349
+    },
+    {
+      "epoch": 0.3482587064676617,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001459720057441037,
+      "loss": 0.5596,
+      "step": 7350
+    },
+    {
+      "epoch": 0.34830608860459605,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00014595877965963418,
+      "loss": 1.2321,
+      "step": 7351
+    },
+    {
+      "epoch": 0.34835347074153045,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014594555255584786,
+      "loss": 1.2072,
+      "step": 7352
+    },
+    {
+      "epoch": 0.34840085287846484,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00014593232443303812,
+      "loss": 1.08,
+      "step": 7353
+    },
+    {
+      "epoch": 0.3484482350153992,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014591909529149838,
+      "loss": 0.8205,
+      "step": 7354
+    },
+    {
+      "epoch": 0.34849561715233357,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00014590586513152202,
+      "loss": 1.0821,
+      "step": 7355
+    },
+    {
+      "epoch": 0.34854299928926796,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014589263395340245,
+      "loss": 0.8623,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3485903814262023,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00014587940175743317,
+      "loss": 0.7111,
+      "step": 7357
+    },
+    {
+      "epoch": 0.3486377635631367,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001458661685439076,
+      "loss": 0.1724,
+      "step": 7358
+    },
+    {
+      "epoch": 0.3486851457000711,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014585293431311925,
+      "loss": 0.7656,
+      "step": 7359
+    },
+    {
+      "epoch": 0.3487325278370054,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014583969906536168,
+      "loss": 0.4699,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3487799099739398,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0001458264628009284,
+      "loss": 1.0327,
+      "step": 7361
+    },
+    {
+      "epoch": 0.3488272921108742,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000145813225520113,
+      "loss": 1.2035,
+      "step": 7362
+    },
+    {
+      "epoch": 0.3488746742478086,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014579998722320906,
+      "loss": 0.2131,
+      "step": 7363
+    },
+    {
+      "epoch": 0.34892205638474294,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014578674791051018,
+      "loss": 0.8787,
+      "step": 7364
+    },
+    {
+      "epoch": 0.34896943852167733,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014577350758231,
+      "loss": 0.5334,
+      "step": 7365
+    },
+    {
+      "epoch": 0.3490168206586117,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001457602662389022,
+      "loss": 0.7806,
+      "step": 7366
+    },
+    {
+      "epoch": 0.34906420279554606,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001457470238805804,
+      "loss": 0.957,
+      "step": 7367
+    },
+    {
+      "epoch": 0.34911158493248046,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014573378050763836,
+      "loss": 0.6307,
+      "step": 7368
+    },
+    {
+      "epoch": 0.34915896706941485,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00014572053612036979,
+      "loss": 0.0118,
+      "step": 7369
+    },
+    {
+      "epoch": 0.3492063492063492,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00014570729071906839,
+      "loss": 1.1567,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3492537313432836,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014569404430402798,
+      "loss": 0.8284,
+      "step": 7371
+    },
+    {
+      "epoch": 0.349301113480218,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014568079687554236,
+      "loss": 0.6115,
+      "step": 7372
+    },
+    {
+      "epoch": 0.3493484956171523,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014566754843390535,
+      "loss": 1.1986,
+      "step": 7373
+    },
+    {
+      "epoch": 0.3493958777540867,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001456542989794107,
+      "loss": 0.9807,
+      "step": 7374
+    },
+    {
+      "epoch": 0.3494432598910211,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014564104851235232,
+      "loss": 1.0133,
+      "step": 7375
+    },
+    {
+      "epoch": 0.34949064202795543,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014562779703302411,
+      "loss": 1.1259,
+      "step": 7376
+    },
+    {
+      "epoch": 0.3495380241648898,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00014561454454172,
+      "loss": 1.0618,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3495854063018242,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014560129103873386,
+      "loss": 0.7027,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3496327884387586,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00014558803652435964,
+      "loss": 0.2671,
+      "step": 7379
+    },
+    {
+      "epoch": 0.34968017057569295,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014557478099889136,
+      "loss": 0.7036,
+      "step": 7380
+    },
+    {
+      "epoch": 0.34972755271262734,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00014556152446262296,
+      "loss": 0.6956,
+      "step": 7381
+    },
+    {
+      "epoch": 0.34977493484956174,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00014554826691584846,
+      "loss": 1.3122,
+      "step": 7382
+    },
+    {
+      "epoch": 0.3498223169864961,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014553500835886194,
+      "loss": 1.0056,
+      "step": 7383
+    },
+    {
+      "epoch": 0.34986969912343047,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014552174879195744,
+      "loss": 1.1146,
+      "step": 7384
+    },
+    {
+      "epoch": 0.34991708126036486,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00014550848821542905,
+      "loss": 0.7516,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3499644633972992,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001454952266295708,
+      "loss": 1.2665,
+      "step": 7386
+    },
+    {
+      "epoch": 0.3500118455342336,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014548196403467694,
+      "loss": 0.3745,
+      "step": 7387
+    },
+    {
+      "epoch": 0.350059227671168,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014546870043104156,
+      "loss": 0.5764,
+      "step": 7388
+    },
+    {
+      "epoch": 0.3501066098081023,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014545543581895884,
+      "loss": 0.6194,
+      "step": 7389
+    },
+    {
+      "epoch": 0.3501539919450367,
+      "grad_norm": 0.00154876708984375,
+      "learning_rate": 0.00014544217019872295,
+      "loss": 0.0001,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3502013740819711,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00014542890357062814,
+      "loss": 0.9624,
+      "step": 7391
+    },
+    {
+      "epoch": 0.3502487562189055,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014541563593496864,
+      "loss": 1.2012,
+      "step": 7392
+    },
+    {
+      "epoch": 0.35029613835583984,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014540236729203868,
+      "loss": 0.0248,
+      "step": 7393
+    },
+    {
+      "epoch": 0.35034352049277423,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001453890976421326,
+      "loss": 0.8003,
+      "step": 7394
+    },
+    {
+      "epoch": 0.3503909026297086,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014537582698554466,
+      "loss": 0.9812,
+      "step": 7395
+    },
+    {
+      "epoch": 0.35043828476664296,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014536255532256927,
+      "loss": 0.8983,
+      "step": 7396
+    },
+    {
+      "epoch": 0.35048566690357735,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00014534928265350067,
+      "loss": 0.0213,
+      "step": 7397
+    },
+    {
+      "epoch": 0.35053304904051175,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001453360089786333,
+      "loss": 0.8947,
+      "step": 7398
+    },
+    {
+      "epoch": 0.3505804311774461,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014532273429826152,
+      "loss": 0.8732,
+      "step": 7399
+    },
+    {
+      "epoch": 0.3506278133143805,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001453094586126798,
+      "loss": 0.1235,
+      "step": 7400
+    },
+    {
+      "epoch": 0.35067519545131487,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0001452961819221825,
+      "loss": 0.0053,
+      "step": 7401
+    },
+    {
+      "epoch": 0.3507225775882492,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014528290422706418,
+      "loss": 0.132,
+      "step": 7402
+    },
+    {
+      "epoch": 0.3507699597251836,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014526962552761927,
+      "loss": 0.8855,
+      "step": 7403
+    },
+    {
+      "epoch": 0.350817341862118,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00014525634582414226,
+      "loss": 0.7257,
+      "step": 7404
+    },
+    {
+      "epoch": 0.35086472399905233,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014524306511692772,
+      "loss": 1.1252,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3509121061359867,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014522978340627017,
+      "loss": 1.1197,
+      "step": 7406
+    },
+    {
+      "epoch": 0.3509594882729211,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014521650069246423,
+      "loss": 1.0832,
+      "step": 7407
+    },
+    {
+      "epoch": 0.3510068704098555,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014520321697580446,
+      "loss": 1.3954,
+      "step": 7408
+    },
+    {
+      "epoch": 0.35105425254678985,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00014518993225658548,
+      "loss": 0.1518,
+      "step": 7409
+    },
+    {
+      "epoch": 0.35110163468372424,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001451766465351019,
+      "loss": 0.9196,
+      "step": 7410
+    },
+    {
+      "epoch": 0.35114901682065863,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001451633598116485,
+      "loss": 1.2294,
+      "step": 7411
+    },
+    {
+      "epoch": 0.35119639895759297,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00014515007208651984,
+      "loss": 0.9131,
+      "step": 7412
+    },
+    {
+      "epoch": 0.35124378109452736,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014513678336001068,
+      "loss": 1.2582,
+      "step": 7413
+    },
+    {
+      "epoch": 0.35129116323146176,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014512349363241572,
+      "loss": 0.7336,
+      "step": 7414
+    },
+    {
+      "epoch": 0.3513385453683961,
+      "grad_norm": 0.006500244140625,
+      "learning_rate": 0.00014511020290402976,
+      "loss": 0.0003,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3513859275053305,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014509691117514753,
+      "loss": 0.5614,
+      "step": 7416
+    },
+    {
+      "epoch": 0.3514333096422649,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014508361844606387,
+      "loss": 0.8816,
+      "step": 7417
+    },
+    {
+      "epoch": 0.3514806917791992,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00014507032471707353,
+      "loss": 0.1625,
+      "step": 7418
+    },
+    {
+      "epoch": 0.3515280739161336,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014505702998847145,
+      "loss": 0.1259,
+      "step": 7419
+    },
+    {
+      "epoch": 0.351575456053068,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014504373426055243,
+      "loss": 0.9052,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3516228381900024,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014503043753361135,
+      "loss": 1.4965,
+      "step": 7421
+    },
+    {
+      "epoch": 0.35167022032693673,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014501713980794308,
+      "loss": 1.3937,
+      "step": 7422
+    },
+    {
+      "epoch": 0.3517176024638711,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00014500384108384268,
+      "loss": 0.0065,
+      "step": 7423
+    },
+    {
+      "epoch": 0.3517649846008055,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014499054136160496,
+      "loss": 1.1131,
+      "step": 7424
+    },
+    {
+      "epoch": 0.35181236673773986,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 0.00014497724064152498,
+      "loss": 0.0025,
+      "step": 7425
+    },
+    {
+      "epoch": 0.35185974887467425,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001449639389238977,
+      "loss": 1.2031,
+      "step": 7426
+    },
+    {
+      "epoch": 0.35190713101160864,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001449506362090182,
+      "loss": 0.9041,
+      "step": 7427
+    },
+    {
+      "epoch": 0.351954513148543,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001449373324971814,
+      "loss": 1.0974,
+      "step": 7428
+    },
+    {
+      "epoch": 0.3520018952854774,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00014492402778868246,
+      "loss": 0.0086,
+      "step": 7429
+    },
+    {
+      "epoch": 0.35204927742241177,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00014491072208381643,
+      "loss": 0.0784,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3520966595593461,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001448974153828784,
+      "loss": 1.2892,
+      "step": 7431
+    },
+    {
+      "epoch": 0.3521440416962805,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014488410768616355,
+      "loss": 0.4134,
+      "step": 7432
+    },
+    {
+      "epoch": 0.3521914238332149,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014487079899396698,
+      "loss": 0.5826,
+      "step": 7433
+    },
+    {
+      "epoch": 0.3522388059701492,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001448574893065839,
+      "loss": 0.6569,
+      "step": 7434
+    },
+    {
+      "epoch": 0.3522861881070836,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014484417862430947,
+      "loss": 1.0911,
+      "step": 7435
+    },
+    {
+      "epoch": 0.352333570244018,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014483086694743898,
+      "loss": 0.7696,
+      "step": 7436
+    },
+    {
+      "epoch": 0.3523809523809524,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00014481755427626754,
+      "loss": 1.152,
+      "step": 7437
+    },
+    {
+      "epoch": 0.35242833451788674,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00014480424061109052,
+      "loss": 1.2271,
+      "step": 7438
+    },
+    {
+      "epoch": 0.35247571665482114,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014479092595220315,
+      "loss": 0.4939,
+      "step": 7439
+    },
+    {
+      "epoch": 0.35252309879175553,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00014477761029990074,
+      "loss": 0.8439,
+      "step": 7440
+    },
+    {
+      "epoch": 0.35257048092868987,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00014476429365447866,
+      "loss": 1.2563,
+      "step": 7441
+    },
+    {
+      "epoch": 0.35261786306562426,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001447509760162322,
+      "loss": 1.0431,
+      "step": 7442
+    },
+    {
+      "epoch": 0.35266524520255865,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014473765738545676,
+      "loss": 1.1327,
+      "step": 7443
+    },
+    {
+      "epoch": 0.352712627339493,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014472433776244778,
+      "loss": 1.1597,
+      "step": 7444
+    },
+    {
+      "epoch": 0.3527600094764274,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00014471101714750057,
+      "loss": 1.1022,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3528073916133618,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014469769554091067,
+      "loss": 1.1637,
+      "step": 7446
+    },
+    {
+      "epoch": 0.3528547737502961,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014468437294297345,
+      "loss": 0.52,
+      "step": 7447
+    },
+    {
+      "epoch": 0.3529021558872305,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014467104935398447,
+      "loss": 0.8924,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3529495380241649,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00014465772477423918,
+      "loss": 0.0154,
+      "step": 7449
+    },
+    {
+      "epoch": 0.3529969201610993,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014464439920403312,
+      "loss": 0.5742,
+      "step": 7450
+    },
+    {
+      "epoch": 0.35304430229803363,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014463107264366183,
+      "loss": 0.9505,
+      "step": 7451
+    },
+    {
+      "epoch": 0.353091684434968,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001446177450934209,
+      "loss": 0.8262,
+      "step": 7452
+    },
+    {
+      "epoch": 0.3531390665719024,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00014460441655360587,
+      "loss": 1.1557,
+      "step": 7453
+    },
+    {
+      "epoch": 0.35318644870883675,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00014459108702451245,
+      "loss": 0.0148,
+      "step": 7454
+    },
+    {
+      "epoch": 0.35323383084577115,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001445777565064362,
+      "loss": 1.3203,
+      "step": 7455
+    },
+    {
+      "epoch": 0.35328121298270554,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014456442499967276,
+      "loss": 1.0322,
+      "step": 7456
+    },
+    {
+      "epoch": 0.3533285951196399,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014455109250451789,
+      "loss": 0.0646,
+      "step": 7457
+    },
+    {
+      "epoch": 0.35337597725657427,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014453775902126723,
+      "loss": 0.8735,
+      "step": 7458
+    },
+    {
+      "epoch": 0.35342335939350866,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014452442455021648,
+      "loss": 0.0772,
+      "step": 7459
+    },
+    {
+      "epoch": 0.353470741530443,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014451108909166146,
+      "loss": 1.2499,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3535181236673774,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014449775264589789,
+      "loss": 0.7132,
+      "step": 7461
+    },
+    {
+      "epoch": 0.3535655058043118,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014448441521322153,
+      "loss": 1.0795,
+      "step": 7462
+    },
+    {
+      "epoch": 0.3536128879412461,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014447107679392825,
+      "loss": 0.1977,
+      "step": 7463
+    },
+    {
+      "epoch": 0.3536602700781805,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014445773738831384,
+      "loss": 0.74,
+      "step": 7464
+    },
+    {
+      "epoch": 0.3537076522151149,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00014444439699667417,
+      "loss": 1.2651,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3537550343520493,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00014443105561930513,
+      "loss": 0.7995,
+      "step": 7466
+    },
+    {
+      "epoch": 0.35380241648898364,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00014441771325650256,
+      "loss": 0.1529,
+      "step": 7467
+    },
+    {
+      "epoch": 0.35384979862591803,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001444043699085625,
+      "loss": 0.9945,
+      "step": 7468
+    },
+    {
+      "epoch": 0.3538971807628524,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00014439102557578076,
+      "loss": 0.0181,
+      "step": 7469
+    },
+    {
+      "epoch": 0.35394456289978676,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014437768025845338,
+      "loss": 0.0129,
+      "step": 7470
+    },
+    {
+      "epoch": 0.35399194503672116,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00014436433395687627,
+      "loss": 0.0466,
+      "step": 7471
+    },
+    {
+      "epoch": 0.35403932717365555,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014435098667134555,
+      "loss": 1.275,
+      "step": 7472
+    },
+    {
+      "epoch": 0.3540867093105899,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00014433763840215714,
+      "loss": 0.122,
+      "step": 7473
+    },
+    {
+      "epoch": 0.3541340914475243,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014432428914960715,
+      "loss": 1.2653,
+      "step": 7474
+    },
+    {
+      "epoch": 0.3541814735844587,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014431093891399165,
+      "loss": 1.1488,
+      "step": 7475
+    },
+    {
+      "epoch": 0.354228855721393,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014429758769560672,
+      "loss": 1.43,
+      "step": 7476
+    },
+    {
+      "epoch": 0.3542762378583274,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001442842354947485,
+      "loss": 0.586,
+      "step": 7477
+    },
+    {
+      "epoch": 0.3543236199952618,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014427088231171306,
+      "loss": 1.0131,
+      "step": 7478
+    },
+    {
+      "epoch": 0.3543710021321962,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014425752814679663,
+      "loss": 0.8696,
+      "step": 7479
+    },
+    {
+      "epoch": 0.3544183842691305,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014424417300029537,
+      "loss": 1.1594,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3544657664060649,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00014423081687250545,
+      "loss": 1.1847,
+      "step": 7481
+    },
+    {
+      "epoch": 0.3545131485429993,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014421745976372316,
+      "loss": 0.8928,
+      "step": 7482
+    },
+    {
+      "epoch": 0.35456053067993365,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014420410167424472,
+      "loss": 0.9059,
+      "step": 7483
+    },
+    {
+      "epoch": 0.35460791281686804,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00014419074260436636,
+      "loss": 0.1536,
+      "step": 7484
+    },
+    {
+      "epoch": 0.35465529495380244,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001441773825543844,
+      "loss": 0.0605,
+      "step": 7485
+    },
+    {
+      "epoch": 0.3547026770907368,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014416402152459516,
+      "loss": 0.0113,
+      "step": 7486
+    },
+    {
+      "epoch": 0.35475005922767117,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014415065951529495,
+      "loss": 1.0058,
+      "step": 7487
+    },
+    {
+      "epoch": 0.35479744136460556,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014413729652678017,
+      "loss": 0.7031,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3548448235015399,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014412393255934714,
+      "loss": 0.1247,
+      "step": 7489
+    },
+    {
+      "epoch": 0.3548922056384743,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014411056761329228,
+      "loss": 0.1318,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3549395877754087,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014409720168891203,
+      "loss": 1.499,
+      "step": 7491
+    },
+    {
+      "epoch": 0.354986969912343,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00014408383478650282,
+      "loss": 0.6738,
+      "step": 7492
+    },
+    {
+      "epoch": 0.3550343520492774,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001440704669063611,
+      "loss": 0.9576,
+      "step": 7493
+    },
+    {
+      "epoch": 0.3550817341862118,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00014405709804878335,
+      "loss": 1.0804,
+      "step": 7494
+    },
+    {
+      "epoch": 0.3551291163231462,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00014404372821406612,
+      "loss": 0.0015,
+      "step": 7495
+    },
+    {
+      "epoch": 0.35517649846008054,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014403035740250593,
+      "loss": 0.7805,
+      "step": 7496
+    },
+    {
+      "epoch": 0.35522388059701493,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014401698561439927,
+      "loss": 0.833,
+      "step": 7497
+    },
+    {
+      "epoch": 0.3552712627339493,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00014400361285004276,
+      "loss": 1.1491,
+      "step": 7498
+    },
+    {
+      "epoch": 0.35531864487088366,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000143990239109733,
+      "loss": 0.1052,
+      "step": 7499
+    },
+    {
+      "epoch": 0.35536602700781805,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00014397686439376658,
+      "loss": 1.2958,
+      "step": 7500
+    },
+    {
+      "epoch": 0.35541340914475245,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00014396348870244016,
+      "loss": 0.0168,
+      "step": 7501
+    },
+    {
+      "epoch": 0.3554607912816868,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0001439501120360504,
+      "loss": 0.2103,
+      "step": 7502
+    },
+    {
+      "epoch": 0.3555081734186212,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014393673439489394,
+      "loss": 0.854,
+      "step": 7503
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014392335577926755,
+      "loss": 1.0615,
+      "step": 7504
+    },
+    {
+      "epoch": 0.3556029376924899,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00014390997618946786,
+      "loss": 0.0217,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3556503198294243,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001438965956257917,
+      "loss": 0.0417,
+      "step": 7506
+    },
+    {
+      "epoch": 0.3556977019663587,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014388321408853586,
+      "loss": 0.8922,
+      "step": 7507
+    },
+    {
+      "epoch": 0.3557450841032931,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014386983157799697,
+      "loss": 0.872,
+      "step": 7508
+    },
+    {
+      "epoch": 0.3557924662402274,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000143856448094472,
+      "loss": 1.1858,
+      "step": 7509
+    },
+    {
+      "epoch": 0.3558398483771618,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014384306363825772,
+      "loss": 0.6,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3558872305140962,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014382967820965098,
+      "loss": 0.9507,
+      "step": 7511
+    },
+    {
+      "epoch": 0.35593461265103055,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014381629180894867,
+      "loss": 0.6646,
+      "step": 7512
+    },
+    {
+      "epoch": 0.35598199478796494,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00014380290443644767,
+      "loss": 1.2201,
+      "step": 7513
+    },
+    {
+      "epoch": 0.35602937692489933,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00014378951609244492,
+      "loss": 0.9526,
+      "step": 7514
+    },
+    {
+      "epoch": 0.35607675906183367,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00014377612677723736,
+      "loss": 0.1276,
+      "step": 7515
+    },
+    {
+      "epoch": 0.35612414119876806,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001437627364911219,
+      "loss": 0.7398,
+      "step": 7516
+    },
+    {
+      "epoch": 0.35617152333570246,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014374934523439555,
+      "loss": 0.8415,
+      "step": 7517
+    },
+    {
+      "epoch": 0.3562189054726368,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00014373595300735538,
+      "loss": 0.7519,
+      "step": 7518
+    },
+    {
+      "epoch": 0.3562662876095712,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001437225598102983,
+      "loss": 0.139,
+      "step": 7519
+    },
+    {
+      "epoch": 0.3563136697465056,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014370916564352144,
+      "loss": 1.1972,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3563610518834399,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00014369577050732184,
+      "loss": 1.1084,
+      "step": 7521
+    },
+    {
+      "epoch": 0.3564084340203743,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001436823744019966,
+      "loss": 0.8126,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3564558161573087,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014366897732784285,
+      "loss": 1.1367,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3565031982942431,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00014365557928515762,
+      "loss": 0.1515,
+      "step": 7524
+    },
+    {
+      "epoch": 0.35655058043117743,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001436421802742382,
+      "loss": 0.6548,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3565979625681118,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001436287802953817,
+      "loss": 0.6433,
+      "step": 7526
+    },
+    {
+      "epoch": 0.3566453447050462,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00014361537934888533,
+      "loss": 0.1723,
+      "step": 7527
+    },
+    {
+      "epoch": 0.35669272684198056,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014360197743504627,
+      "loss": 0.9152,
+      "step": 7528
+    },
+    {
+      "epoch": 0.35674010897891495,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014358857455416178,
+      "loss": 1.0276,
+      "step": 7529
+    },
+    {
+      "epoch": 0.35678749111584934,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014357517070652921,
+      "loss": 0.7897,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3568348732527837,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001435617658924457,
+      "loss": 0.995,
+      "step": 7531
+    },
+    {
+      "epoch": 0.3568822553897181,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014354836011220868,
+      "loss": 1.0152,
+      "step": 7532
+    },
+    {
+      "epoch": 0.35692963752665247,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014353495336611537,
+      "loss": 0.6246,
+      "step": 7533
+    },
+    {
+      "epoch": 0.3569770196635868,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.00014352154565446318,
+      "loss": 0.0018,
+      "step": 7534
+    },
+    {
+      "epoch": 0.3570244018005212,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00014350813697754948,
+      "loss": 0.0143,
+      "step": 7535
+    },
+    {
+      "epoch": 0.3570717839374556,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 0.00014349472733567162,
+      "loss": 0.0042,
+      "step": 7536
+    },
+    {
+      "epoch": 0.35711916607439,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014348131672912705,
+      "loss": 0.9628,
+      "step": 7537
+    },
+    {
+      "epoch": 0.3571665482113243,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001434679051582132,
+      "loss": 0.9001,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3572139303482587,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001434544926232275,
+      "loss": 0.2132,
+      "step": 7539
+    },
+    {
+      "epoch": 0.3572613124851931,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014344107912446743,
+      "loss": 0.7011,
+      "step": 7540
+    },
+    {
+      "epoch": 0.35730869462212744,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001434276646622305,
+      "loss": 1.2016,
+      "step": 7541
+    },
+    {
+      "epoch": 0.35735607675906184,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00014341424923681423,
+      "loss": 0.2444,
+      "step": 7542
+    },
+    {
+      "epoch": 0.35740345889599623,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001434008328485162,
+      "loss": 1.0004,
+      "step": 7543
+    },
+    {
+      "epoch": 0.35745084103293057,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00014338741549763383,
+      "loss": 0.9323,
+      "step": 7544
+    },
+    {
+      "epoch": 0.35749822316986496,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014337399718446488,
+      "loss": 0.9443,
+      "step": 7545
+    },
+    {
+      "epoch": 0.35754560530679935,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014336057790930684,
+      "loss": 1.1554,
+      "step": 7546
+    },
+    {
+      "epoch": 0.3575929874437337,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014334715767245736,
+      "loss": 0.1453,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3576403695806681,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001433337364742141,
+      "loss": 1.6827,
+      "step": 7548
+    },
+    {
+      "epoch": 0.3576877517176025,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014332031431487475,
+      "loss": 0.4558,
+      "step": 7549
+    },
+    {
+      "epoch": 0.3577351338545368,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00014330689119473693,
+      "loss": 0.1129,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3577825159914712,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00014329346711409842,
+      "loss": 1.1667,
+      "step": 7551
+    },
+    {
+      "epoch": 0.3578298981284056,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0001432800420732569,
+      "loss": 0.0043,
+      "step": 7552
+    },
+    {
+      "epoch": 0.35787728026534,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014326661607251014,
+      "loss": 1.2576,
+      "step": 7553
+    },
+    {
+      "epoch": 0.35792466240227433,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014325318911215598,
+      "loss": 0.7217,
+      "step": 7554
+    },
+    {
+      "epoch": 0.3579720445392087,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001432397611924921,
+      "loss": 0.5296,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3580194266761431,
+      "grad_norm": 0.004425048828125,
+      "learning_rate": 0.0001432263323138164,
+      "loss": 0.0003,
+      "step": 7556
+    },
+    {
+      "epoch": 0.35806680881307745,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001432129024764267,
+      "loss": 0.1384,
+      "step": 7557
+    },
+    {
+      "epoch": 0.35811419095001185,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014319947168062083,
+      "loss": 0.9144,
+      "step": 7558
+    },
+    {
+      "epoch": 0.35816157308694624,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014318603992669667,
+      "loss": 0.0499,
+      "step": 7559
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014317260721495218,
+      "loss": 0.0449,
+      "step": 7560
+    },
+    {
+      "epoch": 0.35825633736081497,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00014315917354568524,
+      "loss": 0.0961,
+      "step": 7561
+    },
+    {
+      "epoch": 0.35830371949774936,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001431457389191938,
+      "loss": 0.6478,
+      "step": 7562
+    },
+    {
+      "epoch": 0.3583511016346837,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014313230333577582,
+      "loss": 1.1945,
+      "step": 7563
+    },
+    {
+      "epoch": 0.3583984837716181,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001431188667957293,
+      "loss": 1.1632,
+      "step": 7564
+    },
+    {
+      "epoch": 0.3584458659085525,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00014310542929935226,
+      "loss": 1.02,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3584932480454869,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001430919908469427,
+      "loss": 1.3646,
+      "step": 7566
+    },
+    {
+      "epoch": 0.3585406301824212,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00014307855143879866,
+      "loss": 0.8545,
+      "step": 7567
+    },
+    {
+      "epoch": 0.3585880123193556,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00014306511107521828,
+      "loss": 0.9253,
+      "step": 7568
+    },
+    {
+      "epoch": 0.35863539445629,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00014305166975649955,
+      "loss": 0.972,
+      "step": 7569
+    },
+    {
+      "epoch": 0.35868277659322434,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001430382274829407,
+      "loss": 0.142,
+      "step": 7570
+    },
+    {
+      "epoch": 0.35873015873015873,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014302478425483976,
+      "loss": 0.3351,
+      "step": 7571
+    },
+    {
+      "epoch": 0.3587775408670931,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014301134007249498,
+      "loss": 1.3903,
+      "step": 7572
+    },
+    {
+      "epoch": 0.35882492300402746,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00014299789493620445,
+      "loss": 0.014,
+      "step": 7573
+    },
+    {
+      "epoch": 0.35887230514096186,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014298444884626641,
+      "loss": 0.7912,
+      "step": 7574
+    },
+    {
+      "epoch": 0.35891968727789625,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014297100180297905,
+      "loss": 0.9977,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3589670694148306,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014295755380664074,
+      "loss": 1.0591,
+      "step": 7576
+    },
+    {
+      "epoch": 0.359014451551765,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00014294410485754956,
+      "loss": 1.1827,
+      "step": 7577
+    },
+    {
+      "epoch": 0.35906183368869937,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014293065495600392,
+      "loss": 0.8982,
+      "step": 7578
+    },
+    {
+      "epoch": 0.3591092158256337,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00014291720410230204,
+      "loss": 1.2666,
+      "step": 7579
+    },
+    {
+      "epoch": 0.3591565979625681,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00014290375229674234,
+      "loss": 0.9742,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3592039800995025,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014289029953962307,
+      "loss": 1.266,
+      "step": 7581
+    },
+    {
+      "epoch": 0.3592513622364369,
+      "grad_norm": 0.009765625,
+      "learning_rate": 0.00014287684583124264,
+      "loss": 0.0004,
+      "step": 7582
+    },
+    {
+      "epoch": 0.3592987443733712,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014286339117189947,
+      "loss": 0.0734,
+      "step": 7583
+    },
+    {
+      "epoch": 0.3593461265103056,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014284993556189191,
+      "loss": 0.1343,
+      "step": 7584
+    },
+    {
+      "epoch": 0.35939350864724,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014283647900151843,
+      "loss": 0.0427,
+      "step": 7585
+    },
+    {
+      "epoch": 0.35944089078417435,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 0.00014282302149107748,
+      "loss": 0.0012,
+      "step": 7586
+    },
+    {
+      "epoch": 0.35948827292110874,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014280956303086751,
+      "loss": 0.5595,
+      "step": 7587
+    },
+    {
+      "epoch": 0.35953565505804314,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014279610362118706,
+      "loss": 0.6958,
+      "step": 7588
+    },
+    {
+      "epoch": 0.3595830371949775,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00014278264326233458,
+      "loss": 0.9134,
+      "step": 7589
+    },
+    {
+      "epoch": 0.35963041933191187,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014276918195460866,
+      "loss": 0.7247,
+      "step": 7590
+    },
+    {
+      "epoch": 0.35967780146884626,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014275571969830786,
+      "loss": 0.7464,
+      "step": 7591
+    },
+    {
+      "epoch": 0.3597251836057806,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014274225649373072,
+      "loss": 0.4297,
+      "step": 7592
+    },
+    {
+      "epoch": 0.359772565742715,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014272879234117586,
+      "loss": 0.0135,
+      "step": 7593
+    },
+    {
+      "epoch": 0.3598199478796494,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001427153272409419,
+      "loss": 0.9053,
+      "step": 7594
+    },
+    {
+      "epoch": 0.3598673300165838,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001427018611933275,
+      "loss": 0.8336,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3599147121535181,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00014268839419863126,
+      "loss": 0.0803,
+      "step": 7596
+    },
+    {
+      "epoch": 0.3599620942904525,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00014267492625715192,
+      "loss": 0.0176,
+      "step": 7597
+    },
+    {
+      "epoch": 0.3600094764273869,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014266145736918816,
+      "loss": 0.7939,
+      "step": 7598
+    },
+    {
+      "epoch": 0.36005685856432124,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00014264798753503875,
+      "loss": 1.0958,
+      "step": 7599
+    },
+    {
+      "epoch": 0.36010424070125563,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014263451675500236,
+      "loss": 0.0532,
+      "step": 7600
+    },
+    {
+      "epoch": 0.36015162283819,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014262104502937785,
+      "loss": 0.9797,
+      "step": 7601
+    },
+    {
+      "epoch": 0.36019900497512436,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014260757235846393,
+      "loss": 0.1309,
+      "step": 7602
+    },
+    {
+      "epoch": 0.36024638711205875,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00014259409874255947,
+      "loss": 1.236,
+      "step": 7603
+    },
+    {
+      "epoch": 0.36029376924899315,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014258062418196323,
+      "loss": 0.8716,
+      "step": 7604
+    },
+    {
+      "epoch": 0.3603411513859275,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014256714867697413,
+      "loss": 1.2937,
+      "step": 7605
+    },
+    {
+      "epoch": 0.3603885335228619,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000142553672227891,
+      "loss": 1.005,
+      "step": 7606
+    },
+    {
+      "epoch": 0.36043591565979627,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014254019483501273,
+      "loss": 0.7123,
+      "step": 7607
+    },
+    {
+      "epoch": 0.3604832977967306,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00014252671649863825,
+      "loss": 0.059,
+      "step": 7608
+    },
+    {
+      "epoch": 0.360530679933665,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001425132372190665,
+      "loss": 1.1571,
+      "step": 7609
+    },
+    {
+      "epoch": 0.3605780620705994,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00014249975699659646,
+      "loss": 0.0491,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3606254442075338,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014248627583152707,
+      "loss": 1.0308,
+      "step": 7611
+    },
+    {
+      "epoch": 0.3606728263444681,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00014247279372415732,
+      "loss": 0.5094,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3607202084814025,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014245931067478624,
+      "loss": 0.179,
+      "step": 7613
+    },
+    {
+      "epoch": 0.3607675906183369,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014244582668371288,
+      "loss": 1.4096,
+      "step": 7614
+    },
+    {
+      "epoch": 0.36081497275527125,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001424323417512363,
+      "loss": 1.5485,
+      "step": 7615
+    },
+    {
+      "epoch": 0.36086235489220564,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001424188558776556,
+      "loss": 1.1223,
+      "step": 7616
+    },
+    {
+      "epoch": 0.36090973702914003,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014240536906326982,
+      "loss": 0.6456,
+      "step": 7617
+    },
+    {
+      "epoch": 0.36095711916607437,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014239188130837818,
+      "loss": 0.3114,
+      "step": 7618
+    },
+    {
+      "epoch": 0.36100450130300876,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001423783926132797,
+      "loss": 0.0378,
+      "step": 7619
+    },
+    {
+      "epoch": 0.36105188343994316,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00014236490297827364,
+      "loss": 1.0091,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3610992655768775,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00014235141240365913,
+      "loss": 0.1251,
+      "step": 7621
+    },
+    {
+      "epoch": 0.3611466477138119,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00014233792088973543,
+      "loss": 0.9898,
+      "step": 7622
+    },
+    {
+      "epoch": 0.3611940298507463,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00014232442843680176,
+      "loss": 0.0914,
+      "step": 7623
+    },
+    {
+      "epoch": 0.36124141198768067,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001423109350451573,
+      "loss": 0.0264,
+      "step": 7624
+    },
+    {
+      "epoch": 0.361288794124615,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014229744071510143,
+      "loss": 0.3978,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3613361762615494,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00014228394544693335,
+      "loss": 0.9037,
+      "step": 7626
+    },
+    {
+      "epoch": 0.3613835583984838,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014227044924095239,
+      "loss": 0.744,
+      "step": 7627
+    },
+    {
+      "epoch": 0.36143094053541813,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001422569520974579,
+      "loss": 0.1844,
+      "step": 7628
+    },
+    {
+      "epoch": 0.3614783226723525,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001422434540167492,
+      "loss": 0.5736,
+      "step": 7629
+    },
+    {
+      "epoch": 0.3615257048092869,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00014222995499912572,
+      "loss": 1.3629,
+      "step": 7630
+    },
+    {
+      "epoch": 0.36157308694622126,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014221645504488678,
+      "loss": 0.1722,
+      "step": 7631
+    },
+    {
+      "epoch": 0.36162046908315565,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014220295415433184,
+      "loss": 0.7315,
+      "step": 7632
+    },
+    {
+      "epoch": 0.36166785122009004,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014218945232776035,
+      "loss": 0.507,
+      "step": 7633
+    },
+    {
+      "epoch": 0.3617152333570244,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014217594956547174,
+      "loss": 0.3093,
+      "step": 7634
+    },
+    {
+      "epoch": 0.3617626154939588,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001421624458677655,
+      "loss": 1.0157,
+      "step": 7635
+    },
+    {
+      "epoch": 0.36180999763089317,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014214894123494112,
+      "loss": 0.9068,
+      "step": 7636
+    },
+    {
+      "epoch": 0.3618573797678275,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001421354356672981,
+      "loss": 1.2128,
+      "step": 7637
+    },
+    {
+      "epoch": 0.3619047619047619,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.000142121929165136,
+      "loss": 0.0029,
+      "step": 7638
+    },
+    {
+      "epoch": 0.3619521440416963,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014210842172875438,
+      "loss": 1.0106,
+      "step": 7639
+    },
+    {
+      "epoch": 0.3619995261786307,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00014209491335845283,
+      "loss": 1.3891,
+      "step": 7640
+    },
+    {
+      "epoch": 0.362046908315565,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014208140405453097,
+      "loss": 1.4021,
+      "step": 7641
+    },
+    {
+      "epoch": 0.3620942904524994,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014206789381728834,
+      "loss": 0.6981,
+      "step": 7642
+    },
+    {
+      "epoch": 0.3621416725894338,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014205438264702465,
+      "loss": 0.121,
+      "step": 7643
+    },
+    {
+      "epoch": 0.36218905472636814,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001420408705440395,
+      "loss": 1.0905,
+      "step": 7644
+    },
+    {
+      "epoch": 0.36223643686330254,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0001420273575086327,
+      "loss": 0.9875,
+      "step": 7645
+    },
+    {
+      "epoch": 0.36228381900023693,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00014201384354110385,
+      "loss": 1.352,
+      "step": 7646
+    },
+    {
+      "epoch": 0.36233120113717127,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014200032864175268,
+      "loss": 0.1415,
+      "step": 7647
+    },
+    {
+      "epoch": 0.36237858327410566,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014198681281087897,
+      "loss": 0.8686,
+      "step": 7648
+    },
+    {
+      "epoch": 0.36242596541104005,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014197329604878248,
+      "loss": 1.5066,
+      "step": 7649
+    },
+    {
+      "epoch": 0.3624733475479744,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.000141959778355763,
+      "loss": 0.0331,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3625207296849088,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001419462597321203,
+      "loss": 1.1088,
+      "step": 7651
+    },
+    {
+      "epoch": 0.3625681118218432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014193274017815428,
+      "loss": 0.6578,
+      "step": 7652
+    },
+    {
+      "epoch": 0.36261549395877757,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001419192196941647,
+      "loss": 0.7129,
+      "step": 7653
+    },
+    {
+      "epoch": 0.3626628760957119,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001419056982804515,
+      "loss": 0.5904,
+      "step": 7654
+    },
+    {
+      "epoch": 0.3627102582326463,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00014189217593731454,
+      "loss": 0.1664,
+      "step": 7655
+    },
+    {
+      "epoch": 0.3627576403695807,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00014187865266505377,
+      "loss": 0.0093,
+      "step": 7656
+    },
+    {
+      "epoch": 0.36280502250651503,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014186512846396906,
+      "loss": 0.5595,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3628524046434494,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001418516033343604,
+      "loss": 1.0324,
+      "step": 7658
+    },
+    {
+      "epoch": 0.3628997867803838,
+      "grad_norm": 0.0016326904296875,
+      "learning_rate": 0.00014183807727652777,
+      "loss": 0.0001,
+      "step": 7659
+    },
+    {
+      "epoch": 0.36294716891731815,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014182455029077112,
+      "loss": 1.1369,
+      "step": 7660
+    },
+    {
+      "epoch": 0.36299455105425255,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001418110223773905,
+      "loss": 0.402,
+      "step": 7661
+    },
+    {
+      "epoch": 0.36304193319118694,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014179749353668598,
+      "loss": 0.7577,
+      "step": 7662
+    },
+    {
+      "epoch": 0.3630893153281213,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014178396376895755,
+      "loss": 0.7793,
+      "step": 7663
+    },
+    {
+      "epoch": 0.36313669746505567,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001417704330745053,
+      "loss": 0.7439,
+      "step": 7664
+    },
+    {
+      "epoch": 0.36318407960199006,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014175690145362934,
+      "loss": 1.0099,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3632314617389244,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014174336890662977,
+      "loss": 1.1489,
+      "step": 7666
+    },
+    {
+      "epoch": 0.3632788438758588,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014172983543380677,
+      "loss": 0.1427,
+      "step": 7667
+    },
+    {
+      "epoch": 0.3633262260127932,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00014171630103546046,
+      "loss": 1.0486,
+      "step": 7668
+    },
+    {
+      "epoch": 0.3633736081497276,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000141702765711891,
+      "loss": 0.1382,
+      "step": 7669
+    },
+    {
+      "epoch": 0.3634209902866619,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00014168922946339863,
+      "loss": 0.0904,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3634683724235963,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001416756922902836,
+      "loss": 0.0271,
+      "step": 7671
+    },
+    {
+      "epoch": 0.3635157545605307,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014166215419284605,
+      "loss": 0.584,
+      "step": 7672
+    },
+    {
+      "epoch": 0.36356313669746504,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014164861517138632,
+      "loss": 0.5146,
+      "step": 7673
+    },
+    {
+      "epoch": 0.36361051883439943,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00014163507522620464,
+      "loss": 0.0787,
+      "step": 7674
+    },
+    {
+      "epoch": 0.3636579009713338,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00014162153435760139,
+      "loss": 0.0122,
+      "step": 7675
+    },
+    {
+      "epoch": 0.36370528310826816,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00014160799256587682,
+      "loss": 0.0143,
+      "step": 7676
+    },
+    {
+      "epoch": 0.36375266524520256,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.00014159444985133128,
+      "loss": 0.002,
+      "step": 7677
+    },
+    {
+      "epoch": 0.36380004738213695,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014158090621426515,
+      "loss": 0.3689,
+      "step": 7678
+    },
+    {
+      "epoch": 0.3638474295190713,
+      "grad_norm": 0.0027923583984375,
+      "learning_rate": 0.00014156736165497883,
+      "loss": 0.0002,
+      "step": 7679
+    },
+    {
+      "epoch": 0.3638948116560057,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014155381617377268,
+      "loss": 0.8971,
+      "step": 7680
+    },
+    {
+      "epoch": 0.36394219379294007,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.00014154026977094715,
+      "loss": 0.0047,
+      "step": 7681
+    },
+    {
+      "epoch": 0.36398957592987446,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014152672244680268,
+      "loss": 1.345,
+      "step": 7682
+    },
+    {
+      "epoch": 0.3640369580668088,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00014151317420163975,
+      "loss": 1.2625,
+      "step": 7683
+    },
+    {
+      "epoch": 0.3640843402037432,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00014149962503575884,
+      "loss": 0.01,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3641317223406776,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014148607494946045,
+      "loss": 0.8278,
+      "step": 7685
+    },
+    {
+      "epoch": 0.3641791044776119,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00014147252394304508,
+      "loss": 0.2542,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3642264866145463,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014145897201681335,
+      "loss": 0.1399,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3642738687514807,
+      "grad_norm": 0.0135498046875,
+      "learning_rate": 0.00014144541917106575,
+      "loss": 0.0009,
+      "step": 7688
+    },
+    {
+      "epoch": 0.36432125088841505,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001414318654061029,
+      "loss": 1.0653,
+      "step": 7689
+    },
+    {
+      "epoch": 0.36436863302534944,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001414183107222254,
+      "loss": 1.1532,
+      "step": 7690
+    },
+    {
+      "epoch": 0.36441601516228384,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001414047551197339,
+      "loss": 0.0235,
+      "step": 7691
+    },
+    {
+      "epoch": 0.3644633972992182,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.000141391198598929,
+      "loss": 1.3046,
+      "step": 7692
+    },
+    {
+      "epoch": 0.36451077943615257,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014137764116011145,
+      "loss": 0.0336,
+      "step": 7693
+    },
+    {
+      "epoch": 0.36455816157308696,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014136408280358186,
+      "loss": 0.0187,
+      "step": 7694
+    },
+    {
+      "epoch": 0.3646055437100213,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014135052352964102,
+      "loss": 0.8419,
+      "step": 7695
+    },
+    {
+      "epoch": 0.3646529258469557,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.00014133696333858958,
+      "loss": 0.0062,
+      "step": 7696
+    },
+    {
+      "epoch": 0.3647003079838901,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001413234022307283,
+      "loss": 1.0701,
+      "step": 7697
+    },
+    {
+      "epoch": 0.3647476901208245,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.000141309840206358,
+      "loss": 0.8737,
+      "step": 7698
+    },
+    {
+      "epoch": 0.3647950722577588,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014129627726577947,
+      "loss": 0.1132,
+      "step": 7699
+    },
+    {
+      "epoch": 0.3648424543946932,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014128271340929343,
+      "loss": 0.6267,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3648898365316276,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00014126914863720082,
+      "loss": 0.9094,
+      "step": 7701
+    },
+    {
+      "epoch": 0.36493721866856194,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00014125558294980246,
+      "loss": 1.1937,
+      "step": 7702
+    },
+    {
+      "epoch": 0.36498460080549633,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001412420163473992,
+      "loss": 0.98,
+      "step": 7703
+    },
+    {
+      "epoch": 0.3650319829424307,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00014122844883029197,
+      "loss": 1.0339,
+      "step": 7704
+    },
+    {
+      "epoch": 0.36507936507936506,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014121488039878162,
+      "loss": 1.2399,
+      "step": 7705
+    },
+    {
+      "epoch": 0.36512674721629945,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014120131105316915,
+      "loss": 0.5477,
+      "step": 7706
+    },
+    {
+      "epoch": 0.36517412935323385,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00014118774079375544,
+      "loss": 0.1137,
+      "step": 7707
+    },
+    {
+      "epoch": 0.3652215114901682,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014117416962084156,
+      "loss": 1.4176,
+      "step": 7708
+    },
+    {
+      "epoch": 0.3652688936271026,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014116059753472844,
+      "loss": 0.9718,
+      "step": 7709
+    },
+    {
+      "epoch": 0.36531627576403697,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00014114702453571712,
+      "loss": 0.6634,
+      "step": 7710
+    },
+    {
+      "epoch": 0.36536365790097136,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014113345062410863,
+      "loss": 0.8321,
+      "step": 7711
+    },
+    {
+      "epoch": 0.3654110400379057,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000141119875800204,
+      "loss": 0.0025,
+      "step": 7712
+    },
+    {
+      "epoch": 0.3654584221748401,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001411063000643043,
+      "loss": 0.0578,
+      "step": 7713
+    },
+    {
+      "epoch": 0.3655058043117745,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001410927234167107,
+      "loss": 0.6011,
+      "step": 7714
+    },
+    {
+      "epoch": 0.3655531864487088,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014107914585772424,
+      "loss": 0.8508,
+      "step": 7715
+    },
+    {
+      "epoch": 0.3656005685856432,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014106556738764607,
+      "loss": 0.0356,
+      "step": 7716
+    },
+    {
+      "epoch": 0.3656479507225776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014105198800677736,
+      "loss": 1.1596,
+      "step": 7717
+    },
+    {
+      "epoch": 0.36569533285951195,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001410384077154193,
+      "loss": 0.111,
+      "step": 7718
+    },
+    {
+      "epoch": 0.36574271499644634,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014102482651387309,
+      "loss": 0.9378,
+      "step": 7719
+    },
+    {
+      "epoch": 0.36579009713338073,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00014101124440243988,
+      "loss": 1.1642,
+      "step": 7720
+    },
+    {
+      "epoch": 0.36583747927031507,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014099766138142098,
+      "loss": 0.5318,
+      "step": 7721
+    },
+    {
+      "epoch": 0.36588486140724946,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014098407745111757,
+      "loss": 0.8979,
+      "step": 7722
+    },
+    {
+      "epoch": 0.36593224354418386,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014097049261183102,
+      "loss": 0.9741,
+      "step": 7723
+    },
+    {
+      "epoch": 0.3659796256811182,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014095690686386257,
+      "loss": 0.9232,
+      "step": 7724
+    },
+    {
+      "epoch": 0.3660270078180526,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014094332020751356,
+      "loss": 0.8781,
+      "step": 7725
+    },
+    {
+      "epoch": 0.366074389954987,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001409297326430853,
+      "loss": 0.6715,
+      "step": 7726
+    },
+    {
+      "epoch": 0.36612177209192137,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014091614417087922,
+      "loss": 1.0083,
+      "step": 7727
+    },
+    {
+      "epoch": 0.3661691542288557,
+      "grad_norm": 0.024658203125,
+      "learning_rate": 0.00014090255479119659,
+      "loss": 0.0023,
+      "step": 7728
+    },
+    {
+      "epoch": 0.3662165363657901,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00014088896450433887,
+      "loss": 0.0067,
+      "step": 7729
+    },
+    {
+      "epoch": 0.3662639185027245,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001408753733106075,
+      "loss": 1.0664,
+      "step": 7730
+    },
+    {
+      "epoch": 0.36631130063965883,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014086178121030385,
+      "loss": 0.7424,
+      "step": 7731
+    },
+    {
+      "epoch": 0.3663586827765932,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014084818820372943,
+      "loss": 0.5998,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3664060649135276,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014083459429118573,
+      "loss": 1.0081,
+      "step": 7733
+    },
+    {
+      "epoch": 0.36645344705046196,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001408209994729742,
+      "loss": 0.2045,
+      "step": 7734
+    },
+    {
+      "epoch": 0.36650082918739635,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001408074037493964,
+      "loss": 0.0289,
+      "step": 7735
+    },
+    {
+      "epoch": 0.36654821132433074,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001407938071207538,
+      "loss": 0.5849,
+      "step": 7736
+    },
+    {
+      "epoch": 0.3665955934612651,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001407802095873481,
+      "loss": 0.0738,
+      "step": 7737
+    },
+    {
+      "epoch": 0.36664297559819947,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001407666111494807,
+      "loss": 1.0406,
+      "step": 7738
+    },
+    {
+      "epoch": 0.36669035773513387,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00014075301180745332,
+      "loss": 1.2881,
+      "step": 7739
+    },
+    {
+      "epoch": 0.36673773987206826,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.00014073941156156756,
+      "loss": 0.0032,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3667851220090026,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.000140725810412125,
+      "loss": 1.5008,
+      "step": 7741
+    },
+    {
+      "epoch": 0.366832504145937,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001407122083594274,
+      "loss": 0.498,
+      "step": 7742
+    },
+    {
+      "epoch": 0.3668798862828714,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014069860540377635,
+      "loss": 0.7711,
+      "step": 7743
+    },
+    {
+      "epoch": 0.3669272684198057,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014068500154547361,
+      "loss": 1.1254,
+      "step": 7744
+    },
+    {
+      "epoch": 0.3669746505567401,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014067139678482086,
+      "loss": 0.6638,
+      "step": 7745
+    },
+    {
+      "epoch": 0.3670220326936745,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014065779112211983,
+      "loss": 0.7345,
+      "step": 7746
+    },
+    {
+      "epoch": 0.36706941483060884,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001406441845576723,
+      "loss": 0.6039,
+      "step": 7747
+    },
+    {
+      "epoch": 0.36711679696754324,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00014063057709178008,
+      "loss": 0.9442,
+      "step": 7748
+    },
+    {
+      "epoch": 0.36716417910447763,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014061696872474494,
+      "loss": 0.4191,
+      "step": 7749
+    },
+    {
+      "epoch": 0.36721156124141197,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014060335945686867,
+      "loss": 0.5002,
+      "step": 7750
+    },
+    {
+      "epoch": 0.36725894337834636,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014058974928845313,
+      "loss": 0.6164,
+      "step": 7751
+    },
+    {
+      "epoch": 0.36730632551528075,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00014057613821980022,
+      "loss": 0.1459,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3673537076522151,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014056252625121177,
+      "loss": 1.2057,
+      "step": 7753
+    },
+    {
+      "epoch": 0.3674010897891495,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014054891338298968,
+      "loss": 0.0651,
+      "step": 7754
+    },
+    {
+      "epoch": 0.3674484719260839,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014053529961543588,
+      "loss": 1.1087,
+      "step": 7755
+    },
+    {
+      "epoch": 0.36749585406301827,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001405216849488523,
+      "loss": 0.1513,
+      "step": 7756
+    },
+    {
+      "epoch": 0.3675432361999526,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00014050806938354092,
+      "loss": 1.0219,
+      "step": 7757
+    },
+    {
+      "epoch": 0.367590618336887,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014049445291980375,
+      "loss": 0.9232,
+      "step": 7758
+    },
+    {
+      "epoch": 0.3676380004738214,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0001404808355579427,
+      "loss": 0.0035,
+      "step": 7759
+    },
+    {
+      "epoch": 0.36768538261075573,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014046721729825987,
+      "loss": 0.513,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3677327647476901,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00014045359814105724,
+      "loss": 0.2758,
+      "step": 7761
+    },
+    {
+      "epoch": 0.3677801468846245,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014043997808663687,
+      "loss": 1.0791,
+      "step": 7762
+    },
+    {
+      "epoch": 0.36782752902155885,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00014042635713530093,
+      "loss": 0.9252,
+      "step": 7763
+    },
+    {
+      "epoch": 0.36787491115849325,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014041273528735137,
+      "loss": 1.0741,
+      "step": 7764
+    },
+    {
+      "epoch": 0.36792229329542764,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014039911254309044,
+      "loss": 0.3521,
+      "step": 7765
+    },
+    {
+      "epoch": 0.367969675432362,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014038548890282022,
+      "loss": 0.8648,
+      "step": 7766
+    },
+    {
+      "epoch": 0.36801705756929637,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00014037186436684286,
+      "loss": 0.751,
+      "step": 7767
+    },
+    {
+      "epoch": 0.36806443970623076,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00014035823893546054,
+      "loss": 0.0032,
+      "step": 7768
+    },
+    {
+      "epoch": 0.3681118218431651,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00014034461260897547,
+      "loss": 0.5844,
+      "step": 7769
+    },
+    {
+      "epoch": 0.3681592039800995,
+      "grad_norm": 0.0205078125,
+      "learning_rate": 0.00014033098538768988,
+      "loss": 0.0019,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3682065861170339,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000140317357271906,
+      "loss": 0.5582,
+      "step": 7771
+    },
+    {
+      "epoch": 0.3682539682539683,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00014030372826192607,
+      "loss": 0.0237,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3683013503909026,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014029009835805236,
+      "loss": 1.1245,
+      "step": 7773
+    },
+    {
+      "epoch": 0.368348732527837,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014027646756058722,
+      "loss": 0.0236,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3683961146647714,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001402628358698329,
+      "loss": 0.8308,
+      "step": 7775
+    },
+    {
+      "epoch": 0.36844349680170574,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014024920328609177,
+      "loss": 0.5748,
+      "step": 7776
+    },
+    {
+      "epoch": 0.36849087893864013,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00014023556980966617,
+      "loss": 1.1517,
+      "step": 7777
+    },
+    {
+      "epoch": 0.3685382610755745,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001402219354408585,
+      "loss": 0.2758,
+      "step": 7778
+    },
+    {
+      "epoch": 0.36858564321250886,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014020830017997117,
+      "loss": 1.1275,
+      "step": 7779
+    },
+    {
+      "epoch": 0.36863302534944326,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014019466402730654,
+      "loss": 1.3826,
+      "step": 7780
+    },
+    {
+      "epoch": 0.36868040748637765,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014018102698316708,
+      "loss": 1.2337,
+      "step": 7781
+    },
+    {
+      "epoch": 0.368727789623312,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00014016738904785525,
+      "loss": 1.1935,
+      "step": 7782
+    },
+    {
+      "epoch": 0.3687751717602464,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001401537502216735,
+      "loss": 1.1739,
+      "step": 7783
+    },
+    {
+      "epoch": 0.36882255389718077,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014014011050492438,
+      "loss": 0.1418,
+      "step": 7784
+    },
+    {
+      "epoch": 0.36886993603411516,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00014012646989791032,
+      "loss": 1.3438,
+      "step": 7785
+    },
+    {
+      "epoch": 0.3689173181710495,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001401128284009339,
+      "loss": 1.1215,
+      "step": 7786
+    },
+    {
+      "epoch": 0.3689647003079839,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001400991860142977,
+      "loss": 0.2389,
+      "step": 7787
+    },
+    {
+      "epoch": 0.3690120824449183,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014008554273830422,
+      "loss": 1.1212,
+      "step": 7788
+    },
+    {
+      "epoch": 0.3690594645818526,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00014007189857325616,
+      "loss": 1.3939,
+      "step": 7789
+    },
+    {
+      "epoch": 0.369106846718787,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014005825351945609,
+      "loss": 1.1574,
+      "step": 7790
+    },
+    {
+      "epoch": 0.3691542288557214,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014004460757720654,
+      "loss": 0.633,
+      "step": 7791
+    },
+    {
+      "epoch": 0.36920161099265575,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001400309607468103,
+      "loss": 1.1132,
+      "step": 7792
+    },
+    {
+      "epoch": 0.36924899312959014,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014001731302856995,
+      "loss": 0.0308,
+      "step": 7793
+    },
+    {
+      "epoch": 0.36929637526652453,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014000366442278828,
+      "loss": 0.5809,
+      "step": 7794
+    },
+    {
+      "epoch": 0.36934375740345887,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001399900149297679,
+      "loss": 0.8216,
+      "step": 7795
+    },
+    {
+      "epoch": 0.36939113954039327,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00013997636454981158,
+      "loss": 0.9698,
+      "step": 7796
+    },
+    {
+      "epoch": 0.36943852167732766,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001399627132832221,
+      "loss": 1.0882,
+      "step": 7797
+    },
+    {
+      "epoch": 0.369485903814262,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001399490611303022,
+      "loss": 0.7544,
+      "step": 7798
+    },
+    {
+      "epoch": 0.3695332859511964,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013993540809135468,
+      "loss": 0.711,
+      "step": 7799
+    },
+    {
+      "epoch": 0.3695806680881308,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013992175416668233,
+      "loss": 0.3963,
+      "step": 7800
+    },
+    {
+      "epoch": 0.3696280502250652,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013990809935658798,
+      "loss": 0.0312,
+      "step": 7801
+    },
+    {
+      "epoch": 0.3696754323619995,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013989444366137454,
+      "loss": 0.0184,
+      "step": 7802
+    },
+    {
+      "epoch": 0.3697228144989339,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001398807870813448,
+      "loss": 0.1373,
+      "step": 7803
+    },
+    {
+      "epoch": 0.3697701966358683,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013986712961680167,
+      "loss": 1.1736,
+      "step": 7804
+    },
+    {
+      "epoch": 0.36981757877280264,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001398534712680481,
+      "loss": 0.973,
+      "step": 7805
+    },
+    {
+      "epoch": 0.36986496090973703,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000139839812035387,
+      "loss": 0.8525,
+      "step": 7806
+    },
+    {
+      "epoch": 0.3699123430466714,
+      "grad_norm": 0.00148773193359375,
+      "learning_rate": 0.00013982615191912128,
+      "loss": 0.0001,
+      "step": 7807
+    },
+    {
+      "epoch": 0.36995972518360576,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013981249091955393,
+      "loss": 0.2925,
+      "step": 7808
+    },
+    {
+      "epoch": 0.37000710732054015,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013979882903698796,
+      "loss": 0.021,
+      "step": 7809
+    },
+    {
+      "epoch": 0.37005448945747454,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001397851662717263,
+      "loss": 0.3092,
+      "step": 7810
+    },
+    {
+      "epoch": 0.3701018715944089,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013977150262407207,
+      "loss": 1.2958,
+      "step": 7811
+    },
+    {
+      "epoch": 0.3701492537313433,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00013975783809432827,
+      "loss": 0.0816,
+      "step": 7812
+    },
+    {
+      "epoch": 0.37019663586827767,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013974417268279802,
+      "loss": 0.9872,
+      "step": 7813
+    },
+    {
+      "epoch": 0.37024401800521206,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001397305063897843,
+      "loss": 0.4005,
+      "step": 7814
+    },
+    {
+      "epoch": 0.3702914001421464,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013971683921559024,
+      "loss": 1.3248,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3703387822790808,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000139703171160519,
+      "loss": 0.5001,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3703861644160152,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 0.00013968950222487375,
+      "loss": 0.0041,
+      "step": 7817
+    },
+    {
+      "epoch": 0.3704335465529495,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013967583240895762,
+      "loss": 0.1899,
+      "step": 7818
+    },
+    {
+      "epoch": 0.3704809286898839,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013966216171307374,
+      "loss": 0.7107,
+      "step": 7819
+    },
+    {
+      "epoch": 0.3705283108268183,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001396484901375254,
+      "loss": 1.0735,
+      "step": 7820
+    },
+    {
+      "epoch": 0.37057569296375265,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001396348176826158,
+      "loss": 0.914,
+      "step": 7821
+    },
+    {
+      "epoch": 0.37062307510068704,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001396211443486481,
+      "loss": 0.687,
+      "step": 7822
+    },
+    {
+      "epoch": 0.37067045723762143,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013960747013592567,
+      "loss": 0.3023,
+      "step": 7823
+    },
+    {
+      "epoch": 0.37071783937455577,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001395937950447517,
+      "loss": 0.8713,
+      "step": 7824
+    },
+    {
+      "epoch": 0.37076522151149016,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013958011907542957,
+      "loss": 0.7141,
+      "step": 7825
+    },
+    {
+      "epoch": 0.37081260364842455,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001395664422282625,
+      "loss": 0.8826,
+      "step": 7826
+    },
+    {
+      "epoch": 0.3708599857853589,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00013955276450355392,
+      "loss": 1.0949,
+      "step": 7827
+    },
+    {
+      "epoch": 0.3709073679222933,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001395390859016072,
+      "loss": 1.1836,
+      "step": 7828
+    },
+    {
+      "epoch": 0.3709547500592277,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00013952540642272562,
+      "loss": 0.7109,
+      "step": 7829
+    },
+    {
+      "epoch": 0.37100213219616207,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00013951172606721263,
+      "loss": 0.065,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3710495143330964,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013949804483537163,
+      "loss": 0.1274,
+      "step": 7831
+    },
+    {
+      "epoch": 0.3710968964700308,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001394843627275061,
+      "loss": 0.7424,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3711442786069652,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013947067974391943,
+      "loss": 0.9609,
+      "step": 7833
+    },
+    {
+      "epoch": 0.37119166074389953,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013945699588491513,
+      "loss": 0.026,
+      "step": 7834
+    },
+    {
+      "epoch": 0.3712390428808339,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001394433111507967,
+      "loss": 0.1169,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3712864250177683,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013942962554186765,
+      "loss": 1.5699,
+      "step": 7836
+    },
+    {
+      "epoch": 0.37133380715470266,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00013941593905843148,
+      "loss": 1.2091,
+      "step": 7837
+    },
+    {
+      "epoch": 0.37138118929163705,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001394022517007918,
+      "loss": 0.9868,
+      "step": 7838
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001393885634692521,
+      "loss": 0.0908,
+      "step": 7839
+    },
+    {
+      "epoch": 0.3714759535655058,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013937487436411607,
+      "loss": 0.877,
+      "step": 7840
+    },
+    {
+      "epoch": 0.37152333570244017,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001393611843856872,
+      "loss": 0.9356,
+      "step": 7841
+    },
+    {
+      "epoch": 0.37157071783937456,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013934749353426923,
+      "loss": 1.2068,
+      "step": 7842
+    },
+    {
+      "epoch": 0.37161809997630896,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00013933380181016576,
+      "loss": 1.2205,
+      "step": 7843
+    },
+    {
+      "epoch": 0.3716654821132433,
+      "grad_norm": 0.00173187255859375,
+      "learning_rate": 0.00013932010921368049,
+      "loss": 0.0001,
+      "step": 7844
+    },
+    {
+      "epoch": 0.3717128642501777,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00013930641574511705,
+      "loss": 1.0164,
+      "step": 7845
+    },
+    {
+      "epoch": 0.3717602463871121,
+      "grad_norm": 0.033203125,
+      "learning_rate": 0.00013929272140477918,
+      "loss": 0.0033,
+      "step": 7846
+    },
+    {
+      "epoch": 0.3718076285240464,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013927902619297058,
+      "loss": 0.6085,
+      "step": 7847
+    },
+    {
+      "epoch": 0.3718550106609808,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013926533010999505,
+      "loss": 1.1653,
+      "step": 7848
+    },
+    {
+      "epoch": 0.3719023927979152,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013925163315615631,
+      "loss": 1.0358,
+      "step": 7849
+    },
+    {
+      "epoch": 0.37194977493484954,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013923793533175815,
+      "loss": 1.3688,
+      "step": 7850
+    },
+    {
+      "epoch": 0.37199715707178393,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013922423663710438,
+      "loss": 0.641,
+      "step": 7851
+    },
+    {
+      "epoch": 0.37204453920871833,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013921053707249885,
+      "loss": 0.7746,
+      "step": 7852
+    },
+    {
+      "epoch": 0.37209192134565267,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013919683663824532,
+      "loss": 0.0762,
+      "step": 7853
+    },
+    {
+      "epoch": 0.37213930348258706,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013918313533464773,
+      "loss": 0.0893,
+      "step": 7854
+    },
+    {
+      "epoch": 0.37218668561952145,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013916943316200995,
+      "loss": 0.939,
+      "step": 7855
+    },
+    {
+      "epoch": 0.3722340677564558,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013915573012063582,
+      "loss": 1.1455,
+      "step": 7856
+    },
+    {
+      "epoch": 0.3722814498933902,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013914202621082935,
+      "loss": 0.502,
+      "step": 7857
+    },
+    {
+      "epoch": 0.3723288320303246,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001391283214328944,
+      "loss": 1.1999,
+      "step": 7858
+    },
+    {
+      "epoch": 0.37237621416725897,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013911461578713498,
+      "loss": 0.9362,
+      "step": 7859
+    },
+    {
+      "epoch": 0.3724235963041933,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00013910090927385507,
+      "loss": 1.2651,
+      "step": 7860
+    },
+    {
+      "epoch": 0.3724709784411277,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00013908720189335856,
+      "loss": 0.0403,
+      "step": 7861
+    },
+    {
+      "epoch": 0.3725183605780621,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013907349364594959,
+      "loss": 1.3027,
+      "step": 7862
+    },
+    {
+      "epoch": 0.37256574271499643,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00013905978453193217,
+      "loss": 0.1854,
+      "step": 7863
+    },
+    {
+      "epoch": 0.3726131248519308,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001390460745516103,
+      "loss": 1.1814,
+      "step": 7864
+    },
+    {
+      "epoch": 0.3726605069888652,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00013903236370528812,
+      "loss": 1.1386,
+      "step": 7865
+    },
+    {
+      "epoch": 0.37270788912579955,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013901865199326968,
+      "loss": 1.2676,
+      "step": 7866
+    },
+    {
+      "epoch": 0.37275527126273394,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001390049394158591,
+      "loss": 1.6905,
+      "step": 7867
+    },
+    {
+      "epoch": 0.37280265339966834,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001389912259733605,
+      "loss": 1.3225,
+      "step": 7868
+    },
+    {
+      "epoch": 0.3728500355366027,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00013897751166607803,
+      "loss": 0.8471,
+      "step": 7869
+    },
+    {
+      "epoch": 0.37289741767353707,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013896379649431587,
+      "loss": 1.0288,
+      "step": 7870
+    },
+    {
+      "epoch": 0.37294479981047146,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013895008045837823,
+      "loss": 0.7012,
+      "step": 7871
+    },
+    {
+      "epoch": 0.37299218194740585,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013893636355856925,
+      "loss": 0.4255,
+      "step": 7872
+    },
+    {
+      "epoch": 0.3730395640843402,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00013892264579519324,
+      "loss": 0.1423,
+      "step": 7873
+    },
+    {
+      "epoch": 0.3730869462212746,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0001389089271685544,
+      "loss": 0.1286,
+      "step": 7874
+    },
+    {
+      "epoch": 0.373134328358209,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013889520767895698,
+      "loss": 0.8213,
+      "step": 7875
+    },
+    {
+      "epoch": 0.3731817104951433,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001388814873267053,
+      "loss": 0.0258,
+      "step": 7876
+    },
+    {
+      "epoch": 0.3732290926320777,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013886776611210364,
+      "loss": 0.0293,
+      "step": 7877
+    },
+    {
+      "epoch": 0.3732764747690121,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013885404403545635,
+      "loss": 0.4202,
+      "step": 7878
+    },
+    {
+      "epoch": 0.37332385690594644,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001388403210970677,
+      "loss": 0.8329,
+      "step": 7879
+    },
+    {
+      "epoch": 0.37337123904288083,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013882659729724212,
+      "loss": 0.7784,
+      "step": 7880
+    },
+    {
+      "epoch": 0.3734186211798152,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00013881287263628396,
+      "loss": 0.1253,
+      "step": 7881
+    },
+    {
+      "epoch": 0.37346600331674956,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.00013879914711449766,
+      "loss": 0.0111,
+      "step": 7882
+    },
+    {
+      "epoch": 0.37351338545368395,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00013878542073218755,
+      "loss": 0.0033,
+      "step": 7883
+    },
+    {
+      "epoch": 0.37356076759061835,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013877169348965819,
+      "loss": 1.2487,
+      "step": 7884
+    },
+    {
+      "epoch": 0.3736081497275527,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013875796538721392,
+      "loss": 0.2762,
+      "step": 7885
+    },
+    {
+      "epoch": 0.3736555318644871,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013874423642515932,
+      "loss": 1.1925,
+      "step": 7886
+    },
+    {
+      "epoch": 0.37370291400142147,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013873050660379873,
+      "loss": 1.1001,
+      "step": 7887
+    },
+    {
+      "epoch": 0.37375029613835586,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00013871677592343682,
+      "loss": 0.7461,
+      "step": 7888
+    },
+    {
+      "epoch": 0.3737976782752902,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013870304438437804,
+      "loss": 0.6083,
+      "step": 7889
+    },
+    {
+      "epoch": 0.3738450604122246,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013868931198692696,
+      "loss": 0.517,
+      "step": 7890
+    },
+    {
+      "epoch": 0.373892442549159,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013867557873138814,
+      "loss": 0.9896,
+      "step": 7891
+    },
+    {
+      "epoch": 0.3739398246860933,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001386618446180662,
+      "loss": 0.0736,
+      "step": 7892
+    },
+    {
+      "epoch": 0.3739872068230277,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013864810964726572,
+      "loss": 0.8012,
+      "step": 7893
+    },
+    {
+      "epoch": 0.3740345889599621,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013863437381929133,
+      "loss": 0.2274,
+      "step": 7894
+    },
+    {
+      "epoch": 0.37408197109689645,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00013862063713444768,
+      "loss": 0.2137,
+      "step": 7895
+    },
+    {
+      "epoch": 0.37412935323383084,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013860689959303946,
+      "loss": 1.0898,
+      "step": 7896
+    },
+    {
+      "epoch": 0.37417673537076523,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001385931611953713,
+      "loss": 0.0916,
+      "step": 7897
+    },
+    {
+      "epoch": 0.37422411750769957,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013857942194174793,
+      "loss": 0.2023,
+      "step": 7898
+    },
+    {
+      "epoch": 0.37427149964463396,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013856568183247408,
+      "loss": 0.1192,
+      "step": 7899
+    },
+    {
+      "epoch": 0.37431888178156836,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00013855194086785451,
+      "loss": 0.411,
+      "step": 7900
+    },
+    {
+      "epoch": 0.37436626391850275,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013853819904819395,
+      "loss": 0.5893,
+      "step": 7901
+    },
+    {
+      "epoch": 0.3744136460554371,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00013852445637379716,
+      "loss": 0.1442,
+      "step": 7902
+    },
+    {
+      "epoch": 0.3744610281923715,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00013851071284496898,
+      "loss": 0.8285,
+      "step": 7903
+    },
+    {
+      "epoch": 0.3745084103293059,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00013849696846201417,
+      "loss": 0.0176,
+      "step": 7904
+    },
+    {
+      "epoch": 0.3745557924662402,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00013848322322523765,
+      "loss": 0.112,
+      "step": 7905
+    },
+    {
+      "epoch": 0.3746031746031746,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001384694771349442,
+      "loss": 0.1176,
+      "step": 7906
+    },
+    {
+      "epoch": 0.374650556740109,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00013845573019143876,
+      "loss": 0.793,
+      "step": 7907
+    },
+    {
+      "epoch": 0.37469793887704334,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013844198239502614,
+      "loss": 1.3484,
+      "step": 7908
+    },
+    {
+      "epoch": 0.37474532101397773,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00013842823374601135,
+      "loss": 1.2427,
+      "step": 7909
+    },
+    {
+      "epoch": 0.3747927031509121,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013841448424469922,
+      "loss": 1.1234,
+      "step": 7910
+    },
+    {
+      "epoch": 0.37484008528784646,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00013840073389139476,
+      "loss": 0.0352,
+      "step": 7911
+    },
+    {
+      "epoch": 0.37488746742478085,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013838698268640293,
+      "loss": 0.8207,
+      "step": 7912
+    },
+    {
+      "epoch": 0.37493484956171524,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001383732306300287,
+      "loss": 0.5687,
+      "step": 7913
+    },
+    {
+      "epoch": 0.3749822316986496,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00013835947772257708,
+      "loss": 0.0053,
+      "step": 7914
+    },
+    {
+      "epoch": 0.375029613835584,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00013834572396435318,
+      "loss": 0.1277,
+      "step": 7915
+    },
+    {
+      "epoch": 0.37507699597251837,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00013833196935566188,
+      "loss": 1.2075,
+      "step": 7916
+    },
+    {
+      "epoch": 0.37512437810945276,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013831821389680837,
+      "loss": 0.0653,
+      "step": 7917
+    },
+    {
+      "epoch": 0.3751717602463871,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013830445758809766,
+      "loss": 1.2123,
+      "step": 7918
+    },
+    {
+      "epoch": 0.3752191423833215,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001382907004298349,
+      "loss": 1.2025,
+      "step": 7919
+    },
+    {
+      "epoch": 0.3752665245202559,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013827694242232519,
+      "loss": 0.7279,
+      "step": 7920
+    },
+    {
+      "epoch": 0.3753139066571902,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00013826318356587367,
+      "loss": 0.2879,
+      "step": 7921
+    },
+    {
+      "epoch": 0.3753612887941246,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013824942386078548,
+      "loss": 0.8827,
+      "step": 7922
+    },
+    {
+      "epoch": 0.375408670931059,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013823566330736583,
+      "loss": 0.9701,
+      "step": 7923
+    },
+    {
+      "epoch": 0.37545605306799334,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001382219019059199,
+      "loss": 0.8685,
+      "step": 7924
+    },
+    {
+      "epoch": 0.37550343520492774,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00013820813965675287,
+      "loss": 0.6137,
+      "step": 7925
+    },
+    {
+      "epoch": 0.37555081734186213,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013819437656017,
+      "loss": 0.6669,
+      "step": 7926
+    },
+    {
+      "epoch": 0.37559819947879647,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013818061261647654,
+      "loss": 1.0922,
+      "step": 7927
+    },
+    {
+      "epoch": 0.37564558161573086,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001381668478259778,
+      "loss": 0.8572,
+      "step": 7928
+    },
+    {
+      "epoch": 0.37569296375266525,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.000138153082188979,
+      "loss": 0.2273,
+      "step": 7929
+    },
+    {
+      "epoch": 0.37574034588959965,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00013813931570578548,
+      "loss": 1.2525,
+      "step": 7930
+    },
+    {
+      "epoch": 0.375787728026534,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00013812554837670256,
+      "loss": 0.0903,
+      "step": 7931
+    },
+    {
+      "epoch": 0.3758351101634684,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001381117802020356,
+      "loss": 0.0948,
+      "step": 7932
+    },
+    {
+      "epoch": 0.37588249230040277,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00013809801118208994,
+      "loss": 0.852,
+      "step": 7933
+    },
+    {
+      "epoch": 0.3759298744373371,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013808424131717096,
+      "loss": 0.9516,
+      "step": 7934
+    },
+    {
+      "epoch": 0.3759772565742715,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013807047060758408,
+      "loss": 0.7198,
+      "step": 7935
+    },
+    {
+      "epoch": 0.3760246387112059,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00013805669905363473,
+      "loss": 0.0043,
+      "step": 7936
+    },
+    {
+      "epoch": 0.37607202084814023,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001380429266556283,
+      "loss": 0.0083,
+      "step": 7937
+    },
+    {
+      "epoch": 0.3761194029850746,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013802915341387027,
+      "loss": 0.4724,
+      "step": 7938
+    },
+    {
+      "epoch": 0.376166785122009,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013801537932866617,
+      "loss": 1.1248,
+      "step": 7939
+    },
+    {
+      "epoch": 0.37621416725894335,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013800160440032144,
+      "loss": 0.0161,
+      "step": 7940
+    },
+    {
+      "epoch": 0.37626154939587775,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001379878286291416,
+      "loss": 0.6025,
+      "step": 7941
+    },
+    {
+      "epoch": 0.37630893153281214,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013797405201543215,
+      "loss": 0.0961,
+      "step": 7942
+    },
+    {
+      "epoch": 0.3763563136697465,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0001379602745594987,
+      "loss": 0.0171,
+      "step": 7943
+    },
+    {
+      "epoch": 0.37640369580668087,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001379464962616468,
+      "loss": 0.1276,
+      "step": 7944
+    },
+    {
+      "epoch": 0.37645107794361526,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.000137932717122182,
+      "loss": 1.0654,
+      "step": 7945
+    },
+    {
+      "epoch": 0.37649846008054966,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013791893714140997,
+      "loss": 1.248,
+      "step": 7946
+    },
+    {
+      "epoch": 0.376545842217484,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00013790515631963631,
+      "loss": 0.012,
+      "step": 7947
+    },
+    {
+      "epoch": 0.3765932243544184,
+      "grad_norm": 0.0189208984375,
+      "learning_rate": 0.00013789137465716666,
+      "loss": 0.0009,
+      "step": 7948
+    },
+    {
+      "epoch": 0.3766406064913528,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013787759215430664,
+      "loss": 0.8535,
+      "step": 7949
+    },
+    {
+      "epoch": 0.3766879886282871,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000137863808811362,
+      "loss": 0.1482,
+      "step": 7950
+    },
+    {
+      "epoch": 0.3767353707652215,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00013785002462863842,
+      "loss": 0.0033,
+      "step": 7951
+    },
+    {
+      "epoch": 0.3767827529021559,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013783623960644155,
+      "loss": 1.3051,
+      "step": 7952
+    },
+    {
+      "epoch": 0.37683013503909024,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00013782245374507725,
+      "loss": 0.4477,
+      "step": 7953
+    },
+    {
+      "epoch": 0.37687751717602463,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001378086670448512,
+      "loss": 0.0992,
+      "step": 7954
+    },
+    {
+      "epoch": 0.376924899312959,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013779487950606919,
+      "loss": 0.7131,
+      "step": 7955
+    },
+    {
+      "epoch": 0.37697228144989336,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.000137781091129037,
+      "loss": 1.0882,
+      "step": 7956
+    },
+    {
+      "epoch": 0.37701966358682776,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013776730191406044,
+      "loss": 0.9286,
+      "step": 7957
+    },
+    {
+      "epoch": 0.37706704572376215,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013775351186144538,
+      "loss": 1.1329,
+      "step": 7958
+    },
+    {
+      "epoch": 0.37711442786069654,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013773972097149762,
+      "loss": 0.8817,
+      "step": 7959
+    },
+    {
+      "epoch": 0.3771618099976309,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013772592924452304,
+      "loss": 1.0865,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3772091921345653,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00013771213668082753,
+      "loss": 0.0069,
+      "step": 7961
+    },
+    {
+      "epoch": 0.37725657427149967,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00013769834328071704,
+      "loss": 0.8624,
+      "step": 7962
+    },
+    {
+      "epoch": 0.377303956408434,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013768454904449743,
+      "loss": 0.7787,
+      "step": 7963
+    },
+    {
+      "epoch": 0.3773513385453684,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00013767075397247465,
+      "loss": 0.0185,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3773987206823028,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00013765695806495468,
+      "loss": 0.9429,
+      "step": 7965
+    },
+    {
+      "epoch": 0.37744610281923713,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001376431613222435,
+      "loss": 1.2558,
+      "step": 7966
+    },
+    {
+      "epoch": 0.3774934849561715,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00013762936374464707,
+      "loss": 0.9674,
+      "step": 7967
+    },
+    {
+      "epoch": 0.3775408670931059,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00013761556533247143,
+      "loss": 0.7816,
+      "step": 7968
+    },
+    {
+      "epoch": 0.37758824923004025,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013760176608602263,
+      "loss": 0.4469,
+      "step": 7969
+    },
+    {
+      "epoch": 0.37763563136697464,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013758796600560675,
+      "loss": 0.1748,
+      "step": 7970
+    },
+    {
+      "epoch": 0.37768301350390904,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013757416509152978,
+      "loss": 0.0476,
+      "step": 7971
+    },
+    {
+      "epoch": 0.3777303956408434,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00013756036334409784,
+      "loss": 0.1644,
+      "step": 7972
+    },
+    {
+      "epoch": 0.37777777777777777,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013754656076361707,
+      "loss": 0.8887,
+      "step": 7973
+    },
+    {
+      "epoch": 0.37782515991471216,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001375327573503936,
+      "loss": 0.0274,
+      "step": 7974
+    },
+    {
+      "epoch": 0.37787254205164655,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001375189531047335,
+      "loss": 1.3528,
+      "step": 7975
+    },
+    {
+      "epoch": 0.3779199241885809,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000137505148026943,
+      "loss": 0.0035,
+      "step": 7976
+    },
+    {
+      "epoch": 0.3779673063255153,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013749134211732828,
+      "loss": 0.9132,
+      "step": 7977
+    },
+    {
+      "epoch": 0.3780146884624497,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00013747753537619552,
+      "loss": 0.7628,
+      "step": 7978
+    },
+    {
+      "epoch": 0.378062070599384,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013746372780385095,
+      "loss": 0.0163,
+      "step": 7979
+    },
+    {
+      "epoch": 0.3781094527363184,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013744991940060078,
+      "loss": 1.0008,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3781568348732528,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013743611016675132,
+      "loss": 0.8128,
+      "step": 7981
+    },
+    {
+      "epoch": 0.37820421701018714,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013742230010260878,
+      "loss": 0.9581,
+      "step": 7982
+    },
+    {
+      "epoch": 0.37825159914712153,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001374084892084795,
+      "loss": 0.9201,
+      "step": 7983
+    },
+    {
+      "epoch": 0.3782989812840559,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013739467748466973,
+      "loss": 1.4496,
+      "step": 7984
+    },
+    {
+      "epoch": 0.37834636342099026,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001373808649314859,
+      "loss": 1.2382,
+      "step": 7985
+    },
+    {
+      "epoch": 0.37839374555792465,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001373670515492343,
+      "loss": 0.1384,
+      "step": 7986
+    },
+    {
+      "epoch": 0.37844112769485905,
+      "grad_norm": 0.00262451171875,
+      "learning_rate": 0.00013735323733822129,
+      "loss": 0.0002,
+      "step": 7987
+    },
+    {
+      "epoch": 0.37848850983179344,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00013733942229875323,
+      "loss": 1.2091,
+      "step": 7988
+    },
+    {
+      "epoch": 0.3785358919687278,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001373256064311366,
+      "loss": 0.8466,
+      "step": 7989
+    },
+    {
+      "epoch": 0.37858327410566217,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013731178973567775,
+      "loss": 0.9977,
+      "step": 7990
+    },
+    {
+      "epoch": 0.37863065624259656,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013729797221268317,
+      "loss": 0.1717,
+      "step": 7991
+    },
+    {
+      "epoch": 0.3786780383795309,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013728415386245928,
+      "loss": 0.5609,
+      "step": 7992
+    },
+    {
+      "epoch": 0.3787254205164653,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00013727033468531255,
+      "loss": 0.844,
+      "step": 7993
+    },
+    {
+      "epoch": 0.3787728026533997,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00013725651468154954,
+      "loss": 0.3045,
+      "step": 7994
+    },
+    {
+      "epoch": 0.378820184790334,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001372426938514767,
+      "loss": 0.8474,
+      "step": 7995
+    },
+    {
+      "epoch": 0.3788675669272684,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013722887219540057,
+      "loss": 0.6619,
+      "step": 7996
+    },
+    {
+      "epoch": 0.3789149490642028,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001372150497136277,
+      "loss": 0.0171,
+      "step": 7997
+    },
+    {
+      "epoch": 0.37896233120113715,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001372012264064647,
+      "loss": 0.0394,
+      "step": 7998
+    },
+    {
+      "epoch": 0.37900971333807154,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001371874022742181,
+      "loss": 0.7911,
+      "step": 7999
+    },
+    {
+      "epoch": 0.37905709547500593,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00013717357731719455,
+      "loss": 1.3489,
+      "step": 8000
+    },
+    {
+      "epoch": 0.37910447761194027,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001371597515357006,
+      "loss": 0.2617,
+      "step": 8001
+    },
+    {
+      "epoch": 0.37915185974887466,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013714592493004299,
+      "loss": 0.9439,
+      "step": 8002
+    },
+    {
+      "epoch": 0.37919924188580906,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001371320975005283,
+      "loss": 1.1759,
+      "step": 8003
+    },
+    {
+      "epoch": 0.37924662402274345,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013711826924746328,
+      "loss": 0.4806,
+      "step": 8004
+    },
+    {
+      "epoch": 0.3792940061596778,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00013710444017115452,
+      "loss": 0.7866,
+      "step": 8005
+    },
+    {
+      "epoch": 0.3793413882966122,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013709061027190886,
+      "loss": 0.8067,
+      "step": 8006
+    },
+    {
+      "epoch": 0.3793887704335466,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00013707677955003292,
+      "loss": 1.244,
+      "step": 8007
+    },
+    {
+      "epoch": 0.3794361525704809,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013706294800583355,
+      "loss": 1.1158,
+      "step": 8008
+    },
+    {
+      "epoch": 0.3794835347074153,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013704911563961747,
+      "loss": 1.2878,
+      "step": 8009
+    },
+    {
+      "epoch": 0.3795309168443497,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013703528245169144,
+      "loss": 0.9559,
+      "step": 8010
+    },
+    {
+      "epoch": 0.37957829898128403,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001370214484423623,
+      "loss": 0.9671,
+      "step": 8011
+    },
+    {
+      "epoch": 0.3796256811182184,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013700761361193687,
+      "loss": 0.1168,
+      "step": 8012
+    },
+    {
+      "epoch": 0.3796730632551528,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.000136993777960722,
+      "loss": 0.0051,
+      "step": 8013
+    },
+    {
+      "epoch": 0.37972044539208716,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001369799414890245,
+      "loss": 0.9547,
+      "step": 8014
+    },
+    {
+      "epoch": 0.37976782752902155,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013696610419715132,
+      "loss": 0.0117,
+      "step": 8015
+    },
+    {
+      "epoch": 0.37981520966595594,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00013695226608540932,
+      "loss": 1.1277,
+      "step": 8016
+    },
+    {
+      "epoch": 0.37986259180289034,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013693842715410543,
+      "loss": 1.2176,
+      "step": 8017
+    },
+    {
+      "epoch": 0.3799099739398247,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00013692458740354656,
+      "loss": 1.3435,
+      "step": 8018
+    },
+    {
+      "epoch": 0.37995735607675907,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013691074683403967,
+      "loss": 0.7874,
+      "step": 8019
+    },
+    {
+      "epoch": 0.38000473821369346,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00013689690544589172,
+      "loss": 0.0587,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3800521203506278,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00013688306323940972,
+      "loss": 0.8233,
+      "step": 8021
+    },
+    {
+      "epoch": 0.3800995024875622,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013686922021490064,
+      "loss": 1.1287,
+      "step": 8022
+    },
+    {
+      "epoch": 0.3801468846244966,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013685537637267157,
+      "loss": 1.0793,
+      "step": 8023
+    },
+    {
+      "epoch": 0.3801942667614309,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001368415317130295,
+      "loss": 0.697,
+      "step": 8024
+    },
+    {
+      "epoch": 0.3802416488983653,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001368276862362815,
+      "loss": 0.0584,
+      "step": 8025
+    },
+    {
+      "epoch": 0.3802890310352997,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001368138399427346,
+      "loss": 0.984,
+      "step": 8026
+    },
+    {
+      "epoch": 0.38033641317223404,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.000136799992832696,
+      "loss": 0.554,
+      "step": 8027
+    },
+    {
+      "epoch": 0.38038379530916844,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013678614490647271,
+      "loss": 1.06,
+      "step": 8028
+    },
+    {
+      "epoch": 0.38043117744610283,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00013677229616437193,
+      "loss": 0.3189,
+      "step": 8029
+    },
+    {
+      "epoch": 0.38047855958303717,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001367584466067008,
+      "loss": 0.7537,
+      "step": 8030
+    },
+    {
+      "epoch": 0.38052594171997156,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013674459623376646,
+      "loss": 0.9905,
+      "step": 8031
+    },
+    {
+      "epoch": 0.38057332385690595,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001367307450458761,
+      "loss": 0.979,
+      "step": 8032
+    },
+    {
+      "epoch": 0.38062070599384035,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000136716893043337,
+      "loss": 0.0686,
+      "step": 8033
+    },
+    {
+      "epoch": 0.3806680881307747,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013670304022645626,
+      "loss": 0.0448,
+      "step": 8034
+    },
+    {
+      "epoch": 0.3807154702677091,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00013668918659554122,
+      "loss": 0.9802,
+      "step": 8035
+    },
+    {
+      "epoch": 0.38076285240464347,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00013667533215089907,
+      "loss": 0.9204,
+      "step": 8036
+    },
+    {
+      "epoch": 0.3808102345415778,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00013666147689283712,
+      "loss": 1.2561,
+      "step": 8037
+    },
+    {
+      "epoch": 0.3808576166785122,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013664762082166268,
+      "loss": 0.0745,
+      "step": 8038
+    },
+    {
+      "epoch": 0.3809049988154466,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013663376393768304,
+      "loss": 1.1572,
+      "step": 8039
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013661990624120552,
+      "loss": 0.8301,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3809997630893153,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0001366060477325375,
+      "loss": 1.0107,
+      "step": 8041
+    },
+    {
+      "epoch": 0.3810471452262497,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013659218841198634,
+      "loss": 0.5433,
+      "step": 8042
+    },
+    {
+      "epoch": 0.38109452736318405,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013657832827985942,
+      "loss": 1.0698,
+      "step": 8043
+    },
+    {
+      "epoch": 0.38114190950011845,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00013656446733646414,
+      "loss": 0.0785,
+      "step": 8044
+    },
+    {
+      "epoch": 0.38118929163705284,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013655060558210788,
+      "loss": 1.1157,
+      "step": 8045
+    },
+    {
+      "epoch": 0.38123667377398723,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001365367430170982,
+      "loss": 1.1293,
+      "step": 8046
+    },
+    {
+      "epoch": 0.38128405591092157,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0001365228796417424,
+      "loss": 0.0096,
+      "step": 8047
+    },
+    {
+      "epoch": 0.38133143804785596,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013650901545634805,
+      "loss": 0.27,
+      "step": 8048
+    },
+    {
+      "epoch": 0.38137882018479036,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013649515046122265,
+      "loss": 0.0726,
+      "step": 8049
+    },
+    {
+      "epoch": 0.3814262023217247,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0001364812846566737,
+      "loss": 0.0058,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3814735844586591,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013646741804300866,
+      "loss": 1.2114,
+      "step": 8051
+    },
+    {
+      "epoch": 0.3815209665955935,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00013645355062053515,
+      "loss": 1.4249,
+      "step": 8052
+    },
+    {
+      "epoch": 0.3815683487325278,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001364396823895607,
+      "loss": 0.911,
+      "step": 8053
+    },
+    {
+      "epoch": 0.3816157308694622,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013642581335039295,
+      "loss": 0.7898,
+      "step": 8054
+    },
+    {
+      "epoch": 0.3816631130063966,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00013641194350333944,
+      "loss": 0.0948,
+      "step": 8055
+    },
+    {
+      "epoch": 0.38171049514333094,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001363980728487078,
+      "loss": 1.5226,
+      "step": 8056
+    },
+    {
+      "epoch": 0.38175787728026533,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001363842013868057,
+      "loss": 0.9883,
+      "step": 8057
+    },
+    {
+      "epoch": 0.3818052594171997,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013637032911794074,
+      "loss": 0.7916,
+      "step": 8058
+    },
+    {
+      "epoch": 0.38185264155413406,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013635645604242064,
+      "loss": 1.3916,
+      "step": 8059
+    },
+    {
+      "epoch": 0.38190002369106846,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00013634258216055304,
+      "loss": 0.1157,
+      "step": 8060
+    },
+    {
+      "epoch": 0.38194740582800285,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001363287074726457,
+      "loss": 0.939,
+      "step": 8061
+    },
+    {
+      "epoch": 0.38199478796493724,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013631483197900631,
+      "loss": 0.6919,
+      "step": 8062
+    },
+    {
+      "epoch": 0.3820421701018716,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00013630095567994266,
+      "loss": 0.7033,
+      "step": 8063
+    },
+    {
+      "epoch": 0.382089552238806,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013628707857576245,
+      "loss": 0.672,
+      "step": 8064
+    },
+    {
+      "epoch": 0.38213693437574037,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001362732006667735,
+      "loss": 1.0353,
+      "step": 8065
+    },
+    {
+      "epoch": 0.3821843165126747,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001362593219532836,
+      "loss": 0.8868,
+      "step": 8066
+    },
+    {
+      "epoch": 0.3822316986496091,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013624544243560056,
+      "loss": 0.1084,
+      "step": 8067
+    },
+    {
+      "epoch": 0.3822790807865435,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001362315621140322,
+      "loss": 0.9786,
+      "step": 8068
+    },
+    {
+      "epoch": 0.38232646292347783,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013621768098888638,
+      "loss": 0.8478,
+      "step": 8069
+    },
+    {
+      "epoch": 0.3823738450604122,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00013620379906047097,
+      "loss": 1.2938,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3824212271973466,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00013618991632909387,
+      "loss": 0.1151,
+      "step": 8071
+    },
+    {
+      "epoch": 0.38246860933428095,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.000136176032795063,
+      "loss": 0.6844,
+      "step": 8072
+    },
+    {
+      "epoch": 0.38251599147121534,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013616214845868624,
+      "loss": 0.9966,
+      "step": 8073
+    },
+    {
+      "epoch": 0.38256337360814974,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00013614826332027154,
+      "loss": 1.0518,
+      "step": 8074
+    },
+    {
+      "epoch": 0.38261075574508413,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013613437738012684,
+      "loss": 1.0863,
+      "step": 8075
+    },
+    {
+      "epoch": 0.38265813788201847,
+      "grad_norm": 0.0037689208984375,
+      "learning_rate": 0.00013612049063856015,
+      "loss": 0.0003,
+      "step": 8076
+    },
+    {
+      "epoch": 0.38270552001895286,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013610660309587948,
+      "loss": 1.4448,
+      "step": 8077
+    },
+    {
+      "epoch": 0.38275290215588725,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001360927147523928,
+      "loss": 1.0429,
+      "step": 8078
+    },
+    {
+      "epoch": 0.3828002842928216,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013607882560840812,
+      "loss": 0.922,
+      "step": 8079
+    },
+    {
+      "epoch": 0.382847666429756,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013606493566423357,
+      "loss": 0.0116,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3828950485666904,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013605104492017713,
+      "loss": 1.015,
+      "step": 8081
+    },
+    {
+      "epoch": 0.3829424307036247,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013603715337654694,
+      "loss": 0.0668,
+      "step": 8082
+    },
+    {
+      "epoch": 0.3829898128405591,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013602326103365103,
+      "loss": 0.4468,
+      "step": 8083
+    },
+    {
+      "epoch": 0.3830371949774935,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013600936789179757,
+      "loss": 1.3271,
+      "step": 8084
+    },
+    {
+      "epoch": 0.38308457711442784,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00013599547395129472,
+      "loss": 1.0323,
+      "step": 8085
+    },
+    {
+      "epoch": 0.38313195925136223,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00013598157921245054,
+      "loss": 0.4701,
+      "step": 8086
+    },
+    {
+      "epoch": 0.3831793413882966,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001359676836755733,
+      "loss": 0.132,
+      "step": 8087
+    },
+    {
+      "epoch": 0.38322672352523096,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00013595378734097114,
+      "loss": 1.1943,
+      "step": 8088
+    },
+    {
+      "epoch": 0.38327410566216535,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00013593989020895228,
+      "loss": 0.0516,
+      "step": 8089
+    },
+    {
+      "epoch": 0.38332148779909975,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013592599227982491,
+      "loss": 0.1347,
+      "step": 8090
+    },
+    {
+      "epoch": 0.38336886993603414,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013591209355389734,
+      "loss": 1.0963,
+      "step": 8091
+    },
+    {
+      "epoch": 0.3834162520729685,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00013589819403147775,
+      "loss": 0.8227,
+      "step": 8092
+    },
+    {
+      "epoch": 0.38346363420990287,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013588429371287449,
+      "loss": 1.1687,
+      "step": 8093
+    },
+    {
+      "epoch": 0.38351101634683726,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00013587039259839578,
+      "loss": 1.2914,
+      "step": 8094
+    },
+    {
+      "epoch": 0.3835583984837716,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013585649068835,
+      "loss": 0.8823,
+      "step": 8095
+    },
+    {
+      "epoch": 0.383605780620706,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013584258798304542,
+      "loss": 0.7378,
+      "step": 8096
+    },
+    {
+      "epoch": 0.3836531627576404,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00013582868448279044,
+      "loss": 1.3109,
+      "step": 8097
+    },
+    {
+      "epoch": 0.3837005448945747,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013581478018789337,
+      "loss": 0.4641,
+      "step": 8098
+    },
+    {
+      "epoch": 0.3837479270315091,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013580087509866266,
+      "loss": 0.8215,
+      "step": 8099
+    },
+    {
+      "epoch": 0.3837953091684435,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 0.00013578696921540665,
+      "loss": 0.0013,
+      "step": 8100
+    },
+    {
+      "epoch": 0.38384269130537785,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001357730625384338,
+      "loss": 0.022,
+      "step": 8101
+    },
+    {
+      "epoch": 0.38389007344231224,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001357591550680525,
+      "loss": 0.0354,
+      "step": 8102
+    },
+    {
+      "epoch": 0.38393745557924663,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013574524680457127,
+      "loss": 0.6458,
+      "step": 8103
+    },
+    {
+      "epoch": 0.383984837716181,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013573133774829853,
+      "loss": 1.2551,
+      "step": 8104
+    },
+    {
+      "epoch": 0.38403221985311536,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00013571742789954277,
+      "loss": 1.129,
+      "step": 8105
+    },
+    {
+      "epoch": 0.38407960199004976,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013570351725861247,
+      "loss": 1.4974,
+      "step": 8106
+    },
+    {
+      "epoch": 0.38412698412698415,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013568960582581623,
+      "loss": 0.1634,
+      "step": 8107
+    },
+    {
+      "epoch": 0.3841743662639185,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013567569360146254,
+      "loss": 0.0321,
+      "step": 8108
+    },
+    {
+      "epoch": 0.3842217484008529,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00013566178058585995,
+      "loss": 0.16,
+      "step": 8109
+    },
+    {
+      "epoch": 0.3842691305377873,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00013564786677931705,
+      "loss": 0.9219,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3843165126747216,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013563395218214245,
+      "loss": 0.9605,
+      "step": 8111
+    },
+    {
+      "epoch": 0.384363894811656,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00013562003679464476,
+      "loss": 0.0044,
+      "step": 8112
+    },
+    {
+      "epoch": 0.3844112769485904,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00013560612061713255,
+      "loss": 0.0093,
+      "step": 8113
+    },
+    {
+      "epoch": 0.38445865908552473,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00013559220364991453,
+      "loss": 0.1572,
+      "step": 8114
+    },
+    {
+      "epoch": 0.3845060412224591,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013557828589329937,
+      "loss": 0.1326,
+      "step": 8115
+    },
+    {
+      "epoch": 0.3845534233593935,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001355643673475957,
+      "loss": 0.883,
+      "step": 8116
+    },
+    {
+      "epoch": 0.38460080549632786,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013555044801311225,
+      "loss": 1.2444,
+      "step": 8117
+    },
+    {
+      "epoch": 0.38464818763326225,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001355365278901577,
+      "loss": 0.9022,
+      "step": 8118
+    },
+    {
+      "epoch": 0.38469556977019664,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00013552260697904088,
+      "loss": 0.7201,
+      "step": 8119
+    },
+    {
+      "epoch": 0.38474295190713104,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013550868528007043,
+      "loss": 1.5274,
+      "step": 8120
+    },
+    {
+      "epoch": 0.3847903340440654,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013549476279355518,
+      "loss": 0.8623,
+      "step": 8121
+    },
+    {
+      "epoch": 0.38483771618099977,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013548083951980388,
+      "loss": 1.3328,
+      "step": 8122
+    },
+    {
+      "epoch": 0.38488509831793416,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00013546691545912538,
+      "loss": 0.1407,
+      "step": 8123
+    },
+    {
+      "epoch": 0.3849324804548685,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013545299061182842,
+      "loss": 1.1061,
+      "step": 8124
+    },
+    {
+      "epoch": 0.3849798625918029,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00013543906497822194,
+      "loss": 1.2707,
+      "step": 8125
+    },
+    {
+      "epoch": 0.3850272447287373,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00013542513855861475,
+      "loss": 0.1502,
+      "step": 8126
+    },
+    {
+      "epoch": 0.3850746268656716,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001354112113533157,
+      "loss": 0.1422,
+      "step": 8127
+    },
+    {
+      "epoch": 0.385122009002606,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001353972833626337,
+      "loss": 0.8849,
+      "step": 8128
+    },
+    {
+      "epoch": 0.3851693911395404,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013538335458687764,
+      "loss": 0.0291,
+      "step": 8129
+    },
+    {
+      "epoch": 0.38521677327647474,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001353694250263565,
+      "loss": 0.7615,
+      "step": 8130
+    },
+    {
+      "epoch": 0.38526415541340914,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013535549468137917,
+      "loss": 0.8799,
+      "step": 8131
+    },
+    {
+      "epoch": 0.38531153755034353,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013534156355225462,
+      "loss": 1.2742,
+      "step": 8132
+    },
+    {
+      "epoch": 0.3853589196872779,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00013532763163929184,
+      "loss": 0.1134,
+      "step": 8133
+    },
+    {
+      "epoch": 0.38540630182421226,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013531369894279985,
+      "loss": 0.9775,
+      "step": 8134
+    },
+    {
+      "epoch": 0.38545368396114665,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001352997654630876,
+      "loss": 1.1179,
+      "step": 8135
+    },
+    {
+      "epoch": 0.38550106609808105,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013528583120046413,
+      "loss": 0.415,
+      "step": 8136
+    },
+    {
+      "epoch": 0.3855484482350154,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00013527189615523854,
+      "loss": 1.18,
+      "step": 8137
+    },
+    {
+      "epoch": 0.3855958303719498,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013525796032771986,
+      "loss": 0.4005,
+      "step": 8138
+    },
+    {
+      "epoch": 0.38564321250888417,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001352440237182172,
+      "loss": 0.387,
+      "step": 8139
+    },
+    {
+      "epoch": 0.3856905946458185,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013523008632703958,
+      "loss": 0.237,
+      "step": 8140
+    },
+    {
+      "epoch": 0.3857379767827529,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001352161481544962,
+      "loss": 0.0283,
+      "step": 8141
+    },
+    {
+      "epoch": 0.3857853589196873,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00013520220920089617,
+      "loss": 0.0253,
+      "step": 8142
+    },
+    {
+      "epoch": 0.38583274105662163,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013518826946654864,
+      "loss": 1.2714,
+      "step": 8143
+    },
+    {
+      "epoch": 0.385880123193556,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00013517432895176275,
+      "loss": 0.0174,
+      "step": 8144
+    },
+    {
+      "epoch": 0.3859275053304904,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001351603876568477,
+      "loss": 1.0646,
+      "step": 8145
+    },
+    {
+      "epoch": 0.38597488746742475,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00013514644558211276,
+      "loss": 0.5291,
+      "step": 8146
+    },
+    {
+      "epoch": 0.38602226960435915,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013513250272786703,
+      "loss": 0.7843,
+      "step": 8147
+    },
+    {
+      "epoch": 0.38606965174129354,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00013511855909441984,
+      "loss": 0.804,
+      "step": 8148
+    },
+    {
+      "epoch": 0.38611703387822793,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00013510461468208042,
+      "loss": 1.0318,
+      "step": 8149
+    },
+    {
+      "epoch": 0.38616441601516227,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00013509066949115802,
+      "loss": 1.0913,
+      "step": 8150
+    },
+    {
+      "epoch": 0.38621179815209666,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00013507672352196197,
+      "loss": 0.0499,
+      "step": 8151
+    },
+    {
+      "epoch": 0.38625918028903106,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013506277677480155,
+      "loss": 0.948,
+      "step": 8152
+    },
+    {
+      "epoch": 0.3863065624259654,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0001350488292499861,
+      "loss": 0.6322,
+      "step": 8153
+    },
+    {
+      "epoch": 0.3863539445628998,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00013503488094782494,
+      "loss": 0.7711,
+      "step": 8154
+    },
+    {
+      "epoch": 0.3864013266998342,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001350209318686274,
+      "loss": 0.8854,
+      "step": 8155
+    },
+    {
+      "epoch": 0.3864487088367685,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001350069820127029,
+      "loss": 1.0564,
+      "step": 8156
+    },
+    {
+      "epoch": 0.3864960909737029,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013499303138036087,
+      "loss": 0.8249,
+      "step": 8157
+    },
+    {
+      "epoch": 0.3865434731106373,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013497907997191065,
+      "loss": 0.7084,
+      "step": 8158
+    },
+    {
+      "epoch": 0.38659085524757164,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00013496512778766174,
+      "loss": 0.0291,
+      "step": 8159
+    },
+    {
+      "epoch": 0.38663823738450603,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013495117482792348,
+      "loss": 0.7449,
+      "step": 8160
+    },
+    {
+      "epoch": 0.3866856195214404,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001349372210930054,
+      "loss": 1.0236,
+      "step": 8161
+    },
+    {
+      "epoch": 0.3867330016583748,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.000134923266583217,
+      "loss": 0.2611,
+      "step": 8162
+    },
+    {
+      "epoch": 0.38678038379530916,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001349093112988677,
+      "loss": 0.0238,
+      "step": 8163
+    },
+    {
+      "epoch": 0.38682776593224355,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001348953552402671,
+      "loss": 0.9971,
+      "step": 8164
+    },
+    {
+      "epoch": 0.38687514806917794,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001348813984077247,
+      "loss": 1.1173,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3869225302061123,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00013486744080155,
+      "loss": 1.029,
+      "step": 8166
+    },
+    {
+      "epoch": 0.3869699123430467,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013485348242205263,
+      "loss": 1.1005,
+      "step": 8167
+    },
+    {
+      "epoch": 0.38701729447998107,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001348395232695421,
+      "loss": 0.6973,
+      "step": 8168
+    },
+    {
+      "epoch": 0.3870646766169154,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001348255633443281,
+      "loss": 1.0075,
+      "step": 8169
+    },
+    {
+      "epoch": 0.3871120587538498,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013481160264672016,
+      "loss": 0.0349,
+      "step": 8170
+    },
+    {
+      "epoch": 0.3871594408907842,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013479764117702798,
+      "loss": 1.034,
+      "step": 8171
+    },
+    {
+      "epoch": 0.3872068230277185,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001347836789355612,
+      "loss": 0.5948,
+      "step": 8172
+    },
+    {
+      "epoch": 0.3872542051646529,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00013476971592262944,
+      "loss": 0.006,
+      "step": 8173
+    },
+    {
+      "epoch": 0.3873015873015873,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013475575213854244,
+      "loss": 0.8148,
+      "step": 8174
+    },
+    {
+      "epoch": 0.38734896943852165,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013474178758360984,
+      "loss": 0.9542,
+      "step": 8175
+    },
+    {
+      "epoch": 0.38739635157545604,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013472782225814142,
+      "loss": 0.7074,
+      "step": 8176
+    },
+    {
+      "epoch": 0.38744373371239044,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001347138561624469,
+      "loss": 1.0238,
+      "step": 8177
+    },
+    {
+      "epoch": 0.38749111584932483,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013469988929683602,
+      "loss": 0.2586,
+      "step": 8178
+    },
+    {
+      "epoch": 0.38753849798625917,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013468592166161855,
+      "loss": 0.8672,
+      "step": 8179
+    },
+    {
+      "epoch": 0.38758588012319356,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001346719532571043,
+      "loss": 1.4888,
+      "step": 8180
+    },
+    {
+      "epoch": 0.38763326226012795,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00013465798408360304,
+      "loss": 0.8679,
+      "step": 8181
+    },
+    {
+      "epoch": 0.3876806443970623,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001346440141414246,
+      "loss": 1.1073,
+      "step": 8182
+    },
+    {
+      "epoch": 0.3877280265339967,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013463004343087885,
+      "loss": 0.7844,
+      "step": 8183
+    },
+    {
+      "epoch": 0.3877754086709311,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013461607195227563,
+      "loss": 0.7868,
+      "step": 8184
+    },
+    {
+      "epoch": 0.3878227908078654,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001346020997059248,
+      "loss": 0.0389,
+      "step": 8185
+    },
+    {
+      "epoch": 0.3878701729447998,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00013458812669213624,
+      "loss": 0.1952,
+      "step": 8186
+    },
+    {
+      "epoch": 0.3879175550817342,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013457415291121984,
+      "loss": 1.0734,
+      "step": 8187
+    },
+    {
+      "epoch": 0.38796493721866854,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013456017836348561,
+      "loss": 0.6853,
+      "step": 8188
+    },
+    {
+      "epoch": 0.38801231935560293,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00013454620304924342,
+      "loss": 0.1069,
+      "step": 8189
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013453222696880324,
+      "loss": 0.9593,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3881070836294717,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013451825012247506,
+      "loss": 1.3341,
+      "step": 8191
+    },
+    {
+      "epoch": 0.38815446576640605,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013450427251056885,
+      "loss": 0.0085,
+      "step": 8192
+    },
+    {
+      "epoch": 0.38820184790334045,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001344902941333946,
+      "loss": 0.3653,
+      "step": 8193
+    },
+    {
+      "epoch": 0.38824923004027484,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013447631499126242,
+      "loss": 1.182,
+      "step": 8194
+    },
+    {
+      "epoch": 0.3882966121772092,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013446233508448228,
+      "loss": 0.2139,
+      "step": 8195
+    },
+    {
+      "epoch": 0.38834399431414357,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00013444835441336423,
+      "loss": 0.9673,
+      "step": 8196
+    },
+    {
+      "epoch": 0.38839137645107796,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013443437297821836,
+      "loss": 1.0825,
+      "step": 8197
+    },
+    {
+      "epoch": 0.3884387585880123,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00013442039077935482,
+      "loss": 1.0597,
+      "step": 8198
+    },
+    {
+      "epoch": 0.3884861407249467,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013440640781708365,
+      "loss": 0.9482,
+      "step": 8199
+    },
+    {
+      "epoch": 0.3885335228618811,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00013439242409171503,
+      "loss": 0.6872,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3885809049988154,
+      "grad_norm": 0.021484375,
+      "learning_rate": 0.00013437843960355903,
+      "loss": 0.0011,
+      "step": 8201
+    },
+    {
+      "epoch": 0.3886282871357498,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00013436445435292588,
+      "loss": 0.1094,
+      "step": 8202
+    },
+    {
+      "epoch": 0.3886756692726842,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013435046834012575,
+      "loss": 0.1463,
+      "step": 8203
+    },
+    {
+      "epoch": 0.38872305140961855,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001343364815654688,
+      "loss": 0.9555,
+      "step": 8204
+    },
+    {
+      "epoch": 0.38877043354655294,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00013432249402926526,
+      "loss": 0.7445,
+      "step": 8205
+    },
+    {
+      "epoch": 0.38881781568348733,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001343085057318254,
+      "loss": 0.3253,
+      "step": 8206
+    },
+    {
+      "epoch": 0.3888651978204217,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0001342945166734594,
+      "loss": 0.0077,
+      "step": 8207
+    },
+    {
+      "epoch": 0.38891257995735606,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00013428052685447755,
+      "loss": 0.0575,
+      "step": 8208
+    },
+    {
+      "epoch": 0.38895996209429046,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001342665362751901,
+      "loss": 0.2605,
+      "step": 8209
+    },
+    {
+      "epoch": 0.38900734423122485,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001342525449359074,
+      "loss": 0.997,
+      "step": 8210
+    },
+    {
+      "epoch": 0.3890547263681592,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00013423855283693973,
+      "loss": 1.1376,
+      "step": 8211
+    },
+    {
+      "epoch": 0.3891021085050936,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013422455997859744,
+      "loss": 0.828,
+      "step": 8212
+    },
+    {
+      "epoch": 0.389149490642028,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00013421056636119086,
+      "loss": 1.1193,
+      "step": 8213
+    },
+    {
+      "epoch": 0.3891968727789623,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00013419657198503039,
+      "loss": 0.1448,
+      "step": 8214
+    },
+    {
+      "epoch": 0.3892442549158967,
+      "grad_norm": 0.0059814453125,
+      "learning_rate": 0.00013418257685042634,
+      "loss": 0.0004,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3892916370528311,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013416858095768915,
+      "loss": 0.1043,
+      "step": 8216
+    },
+    {
+      "epoch": 0.38933901918976543,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001341545843071292,
+      "loss": 1.0105,
+      "step": 8217
+    },
+    {
+      "epoch": 0.3893864013266998,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.000134140586899057,
+      "loss": 1.4491,
+      "step": 8218
+    },
+    {
+      "epoch": 0.3894337834636342,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013412658873378293,
+      "loss": 1.2056,
+      "step": 8219
+    },
+    {
+      "epoch": 0.3894811656005686,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013411258981161744,
+      "loss": 0.9109,
+      "step": 8220
+    },
+    {
+      "epoch": 0.38952854773750295,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013409859013287107,
+      "loss": 0.8034,
+      "step": 8221
+    },
+    {
+      "epoch": 0.38957592987443734,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001340845896978543,
+      "loss": 1.1814,
+      "step": 8222
+    },
+    {
+      "epoch": 0.38962331201137174,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00013407058850687764,
+      "loss": 1.0716,
+      "step": 8223
+    },
+    {
+      "epoch": 0.3896706941483061,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001340565865602516,
+      "loss": 0.7023,
+      "step": 8224
+    },
+    {
+      "epoch": 0.38971807628524047,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00013404258385828674,
+      "loss": 0.1406,
+      "step": 8225
+    },
+    {
+      "epoch": 0.38976545842217486,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00013402858040129363,
+      "loss": 1.024,
+      "step": 8226
+    },
+    {
+      "epoch": 0.3898128405591092,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001340145761895829,
+      "loss": 0.55,
+      "step": 8227
+    },
+    {
+      "epoch": 0.3898602226960436,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013400057122346505,
+      "loss": 0.8254,
+      "step": 8228
+    },
+    {
+      "epoch": 0.389907604832978,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013398656550325078,
+      "loss": 0.1839,
+      "step": 8229
+    },
+    {
+      "epoch": 0.3899549869699123,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013397255902925065,
+      "loss": 0.7558,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3900023691068467,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013395855180177535,
+      "loss": 0.9242,
+      "step": 8231
+    },
+    {
+      "epoch": 0.3900497512437811,
+      "grad_norm": 0.0245361328125,
+      "learning_rate": 0.00013394454382113557,
+      "loss": 0.0006,
+      "step": 8232
+    },
+    {
+      "epoch": 0.39009713338071544,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013393053508764196,
+      "loss": 0.9738,
+      "step": 8233
+    },
+    {
+      "epoch": 0.39014451551764984,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.00013391652560160522,
+      "loss": 0.0041,
+      "step": 8234
+    },
+    {
+      "epoch": 0.39019189765458423,
+      "grad_norm": 0.0034027099609375,
+      "learning_rate": 0.00013390251536333605,
+      "loss": 0.0002,
+      "step": 8235
+    },
+    {
+      "epoch": 0.3902392797915186,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00013388850437314522,
+      "loss": 0.9698,
+      "step": 8236
+    },
+    {
+      "epoch": 0.39028666192845296,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00013387449263134346,
+      "loss": 0.0449,
+      "step": 8237
+    },
+    {
+      "epoch": 0.39033404406538735,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013386048013824153,
+      "loss": 1.0048,
+      "step": 8238
+    },
+    {
+      "epoch": 0.39038142620232175,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001338464668941502,
+      "loss": 1.4565,
+      "step": 8239
+    },
+    {
+      "epoch": 0.3904288083392561,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013383245289938031,
+      "loss": 0.6494,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3904761904761905,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00013381843815424264,
+      "loss": 0.0027,
+      "step": 8241
+    },
+    {
+      "epoch": 0.39052357261312487,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013380442265904801,
+      "loss": 0.8223,
+      "step": 8242
+    },
+    {
+      "epoch": 0.3905709547500592,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013379040641410734,
+      "loss": 0.9571,
+      "step": 8243
+    },
+    {
+      "epoch": 0.3906183368869936,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001337763894197314,
+      "loss": 1.1093,
+      "step": 8244
+    },
+    {
+      "epoch": 0.390665719023928,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013376237167623116,
+      "loss": 1.0675,
+      "step": 8245
+    },
+    {
+      "epoch": 0.39071310116086233,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00013374835318391746,
+      "loss": 0.8834,
+      "step": 8246
+    },
+    {
+      "epoch": 0.3907604832977967,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013373433394310124,
+      "loss": 0.9922,
+      "step": 8247
+    },
+    {
+      "epoch": 0.3908078654347311,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013372031395409342,
+      "loss": 0.3326,
+      "step": 8248
+    },
+    {
+      "epoch": 0.3908552475716655,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013370629321720498,
+      "loss": 0.9127,
+      "step": 8249
+    },
+    {
+      "epoch": 0.39090262970859985,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00013369227173274682,
+      "loss": 0.0458,
+      "step": 8250
+    },
+    {
+      "epoch": 0.39095001184553424,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00013367824950102997,
+      "loss": 0.0229,
+      "step": 8251
+    },
+    {
+      "epoch": 0.39099739398246863,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001336642265223654,
+      "loss": 0.1301,
+      "step": 8252
+    },
+    {
+      "epoch": 0.39104477611940297,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001336502027970642,
+      "loss": 0.134,
+      "step": 8253
+    },
+    {
+      "epoch": 0.39109215825633736,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001336361783254373,
+      "loss": 0.9979,
+      "step": 8254
+    },
+    {
+      "epoch": 0.39113954039327176,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013362215310779583,
+      "loss": 1.2613,
+      "step": 8255
+    },
+    {
+      "epoch": 0.3911869225302061,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00013360812714445077,
+      "loss": 0.0034,
+      "step": 8256
+    },
+    {
+      "epoch": 0.3912343046671405,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00013359410043571328,
+      "loss": 0.1307,
+      "step": 8257
+    },
+    {
+      "epoch": 0.3912816868040749,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001335800729818944,
+      "loss": 0.5249,
+      "step": 8258
+    },
+    {
+      "epoch": 0.3913290689410092,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00013356604478330527,
+      "loss": 0.0075,
+      "step": 8259
+    },
+    {
+      "epoch": 0.3913764510779436,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013355201584025706,
+      "loss": 1.3444,
+      "step": 8260
+    },
+    {
+      "epoch": 0.391423833214878,
+      "grad_norm": 0.023193359375,
+      "learning_rate": 0.00013353798615306086,
+      "loss": 0.0017,
+      "step": 8261
+    },
+    {
+      "epoch": 0.39147121535181234,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013352395572202783,
+      "loss": 1.1571,
+      "step": 8262
+    },
+    {
+      "epoch": 0.39151859748874673,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013350992454746918,
+      "loss": 0.082,
+      "step": 8263
+    },
+    {
+      "epoch": 0.3915659796256811,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001334958926296961,
+      "loss": 1.4404,
+      "step": 8264
+    },
+    {
+      "epoch": 0.3916133617626155,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001334818599690198,
+      "loss": 0.0723,
+      "step": 8265
+    },
+    {
+      "epoch": 0.39166074389954986,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001334678265657515,
+      "loss": 0.2663,
+      "step": 8266
+    },
+    {
+      "epoch": 0.39170812603648425,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00013345379242020244,
+      "loss": 0.1553,
+      "step": 8267
+    },
+    {
+      "epoch": 0.39175550817341864,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013343975753268394,
+      "loss": 0.0766,
+      "step": 8268
+    },
+    {
+      "epoch": 0.391802890310353,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00013342572190350722,
+      "loss": 0.6747,
+      "step": 8269
+    },
+    {
+      "epoch": 0.3918502724472874,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013341168553298357,
+      "loss": 0.9877,
+      "step": 8270
+    },
+    {
+      "epoch": 0.39189765458422177,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013339764842142433,
+      "loss": 1.0344,
+      "step": 8271
+    },
+    {
+      "epoch": 0.3919450367211561,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.00013338361056914084,
+      "loss": 0.0048,
+      "step": 8272
+    },
+    {
+      "epoch": 0.3919924188580905,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013336957197644441,
+      "loss": 1.2426,
+      "step": 8273
+    },
+    {
+      "epoch": 0.3920398009950249,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001333555326436464,
+      "loss": 0.9186,
+      "step": 8274
+    },
+    {
+      "epoch": 0.3920871831319592,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013334149257105822,
+      "loss": 0.8823,
+      "step": 8275
+    },
+    {
+      "epoch": 0.3921345652688936,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001333274517589913,
+      "loss": 1.0725,
+      "step": 8276
+    },
+    {
+      "epoch": 0.392181947405828,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013331341020775695,
+      "loss": 0.9247,
+      "step": 8277
+    },
+    {
+      "epoch": 0.3922293295427624,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013329936791766665,
+      "loss": 0.9583,
+      "step": 8278
+    },
+    {
+      "epoch": 0.39227671167969674,
+      "grad_norm": 0.0037994384765625,
+      "learning_rate": 0.00013328532488903185,
+      "loss": 0.0002,
+      "step": 8279
+    },
+    {
+      "epoch": 0.39232409381663114,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000133271281122164,
+      "loss": 0.6199,
+      "step": 8280
+    },
+    {
+      "epoch": 0.39237147595356553,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00013325723661737455,
+      "loss": 0.0224,
+      "step": 8281
+    },
+    {
+      "epoch": 0.39241885809049987,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013324319137497504,
+      "loss": 1.1767,
+      "step": 8282
+    },
+    {
+      "epoch": 0.39246624022743426,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013322914539527694,
+      "loss": 0.775,
+      "step": 8283
+    },
+    {
+      "epoch": 0.39251362236436865,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013321509867859178,
+      "loss": 0.0219,
+      "step": 8284
+    },
+    {
+      "epoch": 0.392561004501303,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013320105122523112,
+      "loss": 0.9422,
+      "step": 8285
+    },
+    {
+      "epoch": 0.3926083866382374,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001331870030355065,
+      "loss": 1.0467,
+      "step": 8286
+    },
+    {
+      "epoch": 0.3926557687751718,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013317295410972951,
+      "loss": 0.3509,
+      "step": 8287
+    },
+    {
+      "epoch": 0.3927031509121061,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00013315890444821175,
+      "loss": 1.0666,
+      "step": 8288
+    },
+    {
+      "epoch": 0.3927505330490405,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013314485405126477,
+      "loss": 0.1586,
+      "step": 8289
+    },
+    {
+      "epoch": 0.3927979151859749,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013313080291920025,
+      "loss": 0.3858,
+      "step": 8290
+    },
+    {
+      "epoch": 0.39284529732290924,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00013311675105232983,
+      "loss": 0.2896,
+      "step": 8291
+    },
+    {
+      "epoch": 0.39289267945984363,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013310269845096516,
+      "loss": 0.6466,
+      "step": 8292
+    },
+    {
+      "epoch": 0.392940061596778,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00013308864511541786,
+      "loss": 0.6098,
+      "step": 8293
+    },
+    {
+      "epoch": 0.3929874437337124,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013307459104599964,
+      "loss": 1.1771,
+      "step": 8294
+    },
+    {
+      "epoch": 0.39303482587064675,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013306053624302227,
+      "loss": 0.7478,
+      "step": 8295
+    },
+    {
+      "epoch": 0.39308220800758115,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013304648070679737,
+      "loss": 0.0336,
+      "step": 8296
+    },
+    {
+      "epoch": 0.39312959014451554,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013303242443763675,
+      "loss": 0.872,
+      "step": 8297
+    },
+    {
+      "epoch": 0.3931769722814499,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013301836743585214,
+      "loss": 1.2149,
+      "step": 8298
+    },
+    {
+      "epoch": 0.39322435441838427,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00013300430970175533,
+      "loss": 1.2844,
+      "step": 8299
+    },
+    {
+      "epoch": 0.39327173655531866,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00013299025123565808,
+      "loss": 1.2886,
+      "step": 8300
+    },
+    {
+      "epoch": 0.393319118692253,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00013297619203787216,
+      "loss": 0.0843,
+      "step": 8301
+    },
+    {
+      "epoch": 0.3933665008291874,
+      "grad_norm": 0.0289306640625,
+      "learning_rate": 0.00013296213210870945,
+      "loss": 0.0025,
+      "step": 8302
+    },
+    {
+      "epoch": 0.3934138829661218,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001329480714484818,
+      "loss": 1.1182,
+      "step": 8303
+    },
+    {
+      "epoch": 0.3934612651030561,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00013293401005750095,
+      "loss": 1.0311,
+      "step": 8304
+    },
+    {
+      "epoch": 0.3935086472399905,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00013291994793607884,
+      "loss": 0.0156,
+      "step": 8305
+    },
+    {
+      "epoch": 0.3935560293769249,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00013290588508452743,
+      "loss": 0.1274,
+      "step": 8306
+    },
+    {
+      "epoch": 0.3936034115138593,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013289182150315844,
+      "loss": 0.3793,
+      "step": 8307
+    },
+    {
+      "epoch": 0.39365079365079364,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013287775719228394,
+      "loss": 0.9794,
+      "step": 8308
+    },
+    {
+      "epoch": 0.39369817578772803,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001328636921522158,
+      "loss": 0.1713,
+      "step": 8309
+    },
+    {
+      "epoch": 0.3937455579246624,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00013284962638326596,
+      "loss": 1.7904,
+      "step": 8310
+    },
+    {
+      "epoch": 0.39379294006159676,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013283555988574642,
+      "loss": 0.7938,
+      "step": 8311
+    },
+    {
+      "epoch": 0.39384032219853116,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00013282149265996912,
+      "loss": 0.818,
+      "step": 8312
+    },
+    {
+      "epoch": 0.39388770433546555,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013280742470624606,
+      "loss": 0.1211,
+      "step": 8313
+    },
+    {
+      "epoch": 0.3939350864723999,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00013279335602488926,
+      "loss": 0.1617,
+      "step": 8314
+    },
+    {
+      "epoch": 0.3939824686093343,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00013277928661621077,
+      "loss": 0.0098,
+      "step": 8315
+    },
+    {
+      "epoch": 0.3940298507462687,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013276521648052266,
+      "loss": 0.802,
+      "step": 8316
+    },
+    {
+      "epoch": 0.394077232883203,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001327511456181369,
+      "loss": 0.1753,
+      "step": 8317
+    },
+    {
+      "epoch": 0.3941246150201374,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001327370740293656,
+      "loss": 1.001,
+      "step": 8318
+    },
+    {
+      "epoch": 0.3941719971570718,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001327230017145209,
+      "loss": 0.0278,
+      "step": 8319
+    },
+    {
+      "epoch": 0.39421937929400613,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013270892867391486,
+      "loss": 1.0463,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3942667614309405,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00013269485490785963,
+      "loss": 0.177,
+      "step": 8321
+    },
+    {
+      "epoch": 0.3943141435678749,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013268078041666734,
+      "loss": 0.2885,
+      "step": 8322
+    },
+    {
+      "epoch": 0.3943615257048093,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013266670520065013,
+      "loss": 1.1419,
+      "step": 8323
+    },
+    {
+      "epoch": 0.39440890784174365,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013265262926012025,
+      "loss": 1.1623,
+      "step": 8324
+    },
+    {
+      "epoch": 0.39445628997867804,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001326385525953898,
+      "loss": 0.0795,
+      "step": 8325
+    },
+    {
+      "epoch": 0.39450367211561244,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000132624475206771,
+      "loss": 0.8408,
+      "step": 8326
+    },
+    {
+      "epoch": 0.3945510542525468,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013261039709457613,
+      "loss": 0.3733,
+      "step": 8327
+    },
+    {
+      "epoch": 0.39459843638948117,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00013259631825911735,
+      "loss": 0.1416,
+      "step": 8328
+    },
+    {
+      "epoch": 0.39464581852641556,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013258223870070697,
+      "loss": 0.8125,
+      "step": 8329
+    },
+    {
+      "epoch": 0.3946932006633499,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013256815841965723,
+      "loss": 0.4684,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3947405828002843,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013255407741628045,
+      "loss": 1.0509,
+      "step": 8331
+    },
+    {
+      "epoch": 0.3947879649372187,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00013253999569088888,
+      "loss": 0.8827,
+      "step": 8332
+    },
+    {
+      "epoch": 0.394835347074153,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00013252591324379486,
+      "loss": 0.098,
+      "step": 8333
+    },
+    {
+      "epoch": 0.3948827292110874,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013251183007531077,
+      "loss": 0.0248,
+      "step": 8334
+    },
+    {
+      "epoch": 0.3949301113480218,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00013249774618574888,
+      "loss": 0.1278,
+      "step": 8335
+    },
+    {
+      "epoch": 0.3949774934849562,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013248366157542161,
+      "loss": 0.1062,
+      "step": 8336
+    },
+    {
+      "epoch": 0.39502487562189054,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013246957624464133,
+      "loss": 0.9185,
+      "step": 8337
+    },
+    {
+      "epoch": 0.39507225775882493,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013245549019372043,
+      "loss": 1.1236,
+      "step": 8338
+    },
+    {
+      "epoch": 0.3951196398957593,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00013244140342297134,
+      "loss": 0.0376,
+      "step": 8339
+    },
+    {
+      "epoch": 0.39516702203269366,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013242731593270644,
+      "loss": 0.994,
+      "step": 8340
+    },
+    {
+      "epoch": 0.39521440416962805,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013241322772323825,
+      "loss": 0.9374,
+      "step": 8341
+    },
+    {
+      "epoch": 0.39526178630656245,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013239913879487919,
+      "loss": 1.3045,
+      "step": 8342
+    },
+    {
+      "epoch": 0.3953091684434968,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001323850491479417,
+      "loss": 1.4344,
+      "step": 8343
+    },
+    {
+      "epoch": 0.3953565505804312,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013237095878273835,
+      "loss": 0.5707,
+      "step": 8344
+    },
+    {
+      "epoch": 0.39540393271736557,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.00013235686769958162,
+      "loss": 0.0062,
+      "step": 8345
+    },
+    {
+      "epoch": 0.3954513148542999,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00013234277589878403,
+      "loss": 0.0154,
+      "step": 8346
+    },
+    {
+      "epoch": 0.3954986969912343,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001323286833806581,
+      "loss": 1.0657,
+      "step": 8347
+    },
+    {
+      "epoch": 0.3955460791281687,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013231459014551643,
+      "loss": 0.4864,
+      "step": 8348
+    },
+    {
+      "epoch": 0.39559346126510303,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013230049619367156,
+      "loss": 0.7659,
+      "step": 8349
+    },
+    {
+      "epoch": 0.3956408434020374,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001322864015254361,
+      "loss": 1.1576,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3956882255389718,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013227230614112264,
+      "loss": 0.923,
+      "step": 8351
+    },
+    {
+      "epoch": 0.3957356076759062,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00013225821004104378,
+      "loss": 1.0582,
+      "step": 8352
+    },
+    {
+      "epoch": 0.39578298981284055,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001322441132255122,
+      "loss": 0.5032,
+      "step": 8353
+    },
+    {
+      "epoch": 0.39583037194977494,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013223001569484053,
+      "loss": 0.884,
+      "step": 8354
+    },
+    {
+      "epoch": 0.39587775408670933,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00013221591744934144,
+      "loss": 0.9107,
+      "step": 8355
+    },
+    {
+      "epoch": 0.39592513622364367,
+      "grad_norm": 0.017822265625,
+      "learning_rate": 0.00013220181848932761,
+      "loss": 0.0011,
+      "step": 8356
+    },
+    {
+      "epoch": 0.39597251836057806,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001321877188151118,
+      "loss": 0.0106,
+      "step": 8357
+    },
+    {
+      "epoch": 0.39601990049751246,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001321736184270066,
+      "loss": 0.7856,
+      "step": 8358
+    },
+    {
+      "epoch": 0.3960672826344468,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00013215951732532482,
+      "loss": 0.0062,
+      "step": 8359
+    },
+    {
+      "epoch": 0.3961146647713812,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013214541551037927,
+      "loss": 0.9263,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3961620469083156,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013213131298248255,
+      "loss": 0.8682,
+      "step": 8361
+    },
+    {
+      "epoch": 0.3962094290452499,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001321172097419476,
+      "loss": 0.1484,
+      "step": 8362
+    },
+    {
+      "epoch": 0.3962568111821843,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00013210310578908713,
+      "loss": 1.3438,
+      "step": 8363
+    },
+    {
+      "epoch": 0.3963041933191187,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013208900112421395,
+      "loss": 1.0264,
+      "step": 8364
+    },
+    {
+      "epoch": 0.3963515754560531,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00013207489574764095,
+      "loss": 0.0142,
+      "step": 8365
+    },
+    {
+      "epoch": 0.39639895759298743,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013206078965968089,
+      "loss": 1.2929,
+      "step": 8366
+    },
+    {
+      "epoch": 0.3964463397299218,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00013204668286064664,
+      "loss": 0.9476,
+      "step": 8367
+    },
+    {
+      "epoch": 0.3964937218668562,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00013203257535085114,
+      "loss": 0.2027,
+      "step": 8368
+    },
+    {
+      "epoch": 0.39654110400379056,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013201846713060725,
+      "loss": 0.9504,
+      "step": 8369
+    },
+    {
+      "epoch": 0.39658848614072495,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001320043582002278,
+      "loss": 0.1219,
+      "step": 8370
+    },
+    {
+      "epoch": 0.39663586827765934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013199024856002585,
+      "loss": 0.7931,
+      "step": 8371
+    },
+    {
+      "epoch": 0.3966832504145937,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013197613821031422,
+      "loss": 1.1369,
+      "step": 8372
+    },
+    {
+      "epoch": 0.3967306325515281,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013196202715140591,
+      "loss": 0.7452,
+      "step": 8373
+    },
+    {
+      "epoch": 0.39677801468846247,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001319479153836139,
+      "loss": 0.8202,
+      "step": 8374
+    },
+    {
+      "epoch": 0.3968253968253968,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00013193380290725111,
+      "loss": 1.3312,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3968727789623312,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013191968972263063,
+      "loss": 0.9073,
+      "step": 8376
+    },
+    {
+      "epoch": 0.3969201610992656,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013190557583006538,
+      "loss": 1.0335,
+      "step": 8377
+    },
+    {
+      "epoch": 0.3969675432361999,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013189146122986848,
+      "loss": 0.2046,
+      "step": 8378
+    },
+    {
+      "epoch": 0.3970149253731343,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013187734592235296,
+      "loss": 0.1444,
+      "step": 8379
+    },
+    {
+      "epoch": 0.3970623075100687,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013186322990783186,
+      "loss": 1.1527,
+      "step": 8380
+    },
+    {
+      "epoch": 0.3971096896470031,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013184911318661824,
+      "loss": 1.0669,
+      "step": 8381
+    },
+    {
+      "epoch": 0.39715707178393744,
+      "grad_norm": 0.00848388671875,
+      "learning_rate": 0.0001318349957590252,
+      "loss": 0.0004,
+      "step": 8382
+    },
+    {
+      "epoch": 0.39720445392087184,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00013182087762536588,
+      "loss": 0.2528,
+      "step": 8383
+    },
+    {
+      "epoch": 0.39725183605780623,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013180675878595336,
+      "loss": 0.0269,
+      "step": 8384
+    },
+    {
+      "epoch": 0.39729921819474057,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013179263924110085,
+      "loss": 1.1373,
+      "step": 8385
+    },
+    {
+      "epoch": 0.39734660033167496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013177851899112147,
+      "loss": 0.3504,
+      "step": 8386
+    },
+    {
+      "epoch": 0.39739398246860935,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013176439803632835,
+      "loss": 1.5232,
+      "step": 8387
+    },
+    {
+      "epoch": 0.3974413646055437,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013175027637703473,
+      "loss": 0.8524,
+      "step": 8388
+    },
+    {
+      "epoch": 0.3974887467424781,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001317361540135538,
+      "loss": 0.6525,
+      "step": 8389
+    },
+    {
+      "epoch": 0.3975361288794125,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013172203094619878,
+      "loss": 0.2698,
+      "step": 8390
+    },
+    {
+      "epoch": 0.3975835110163468,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013170790717528292,
+      "loss": 1.1265,
+      "step": 8391
+    },
+    {
+      "epoch": 0.3976308931532812,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001316937827011194,
+      "loss": 0.0248,
+      "step": 8392
+    },
+    {
+      "epoch": 0.3976782752902156,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013167965752402158,
+      "loss": 0.6908,
+      "step": 8393
+    },
+    {
+      "epoch": 0.39772565742714994,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001316655316443027,
+      "loss": 1.1508,
+      "step": 8394
+    },
+    {
+      "epoch": 0.39777303956408433,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013165140506227606,
+      "loss": 1.5209,
+      "step": 8395
+    },
+    {
+      "epoch": 0.3978204217010187,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013163727777825497,
+      "loss": 1.3025,
+      "step": 8396
+    },
+    {
+      "epoch": 0.3978678038379531,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013162314979255272,
+      "loss": 0.9712,
+      "step": 8397
+    },
+    {
+      "epoch": 0.39791518597488745,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00013160902110548274,
+      "loss": 0.5788,
+      "step": 8398
+    },
+    {
+      "epoch": 0.39796256811182185,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0001315948917173583,
+      "loss": 0.0088,
+      "step": 8399
+    },
+    {
+      "epoch": 0.39800995024875624,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013158076162849281,
+      "loss": 1.0409,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3980573323856906,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0001315666308391997,
+      "loss": 0.0045,
+      "step": 8401
+    },
+    {
+      "epoch": 0.39810471452262497,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013155249934979234,
+      "loss": 0.9082,
+      "step": 8402
+    },
+    {
+      "epoch": 0.39815209665955936,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013153836716058414,
+      "loss": 1.0735,
+      "step": 8403
+    },
+    {
+      "epoch": 0.3981994787964937,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00013152423427188856,
+      "loss": 0.0275,
+      "step": 8404
+    },
+    {
+      "epoch": 0.3982468609334281,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000131510100684019,
+      "loss": 0.9725,
+      "step": 8405
+    },
+    {
+      "epoch": 0.3982942430703625,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013149596639728904,
+      "loss": 0.4258,
+      "step": 8406
+    },
+    {
+      "epoch": 0.3983416252072968,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00013148183141201204,
+      "loss": 1.2988,
+      "step": 8407
+    },
+    {
+      "epoch": 0.3983890073442312,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00013146769572850158,
+      "loss": 0.0233,
+      "step": 8408
+    },
+    {
+      "epoch": 0.3984363894811656,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013145355934707112,
+      "loss": 0.9486,
+      "step": 8409
+    },
+    {
+      "epoch": 0.3984837716181,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013143942226803427,
+      "loss": 0.1009,
+      "step": 8410
+    },
+    {
+      "epoch": 0.39853115375503434,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013142528449170446,
+      "loss": 0.6194,
+      "step": 8411
+    },
+    {
+      "epoch": 0.39857853589196873,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013141114601839532,
+      "loss": 1.2989,
+      "step": 8412
+    },
+    {
+      "epoch": 0.3986259180289031,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013139700684842043,
+      "loss": 0.5218,
+      "step": 8413
+    },
+    {
+      "epoch": 0.39867330016583746,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001313828669820934,
+      "loss": 1.2017,
+      "step": 8414
+    },
+    {
+      "epoch": 0.39872068230277186,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013136872641972776,
+      "loss": 0.1648,
+      "step": 8415
+    },
+    {
+      "epoch": 0.39876806443970625,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013135458516163722,
+      "loss": 0.2398,
+      "step": 8416
+    },
+    {
+      "epoch": 0.3988154465766406,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013134044320813537,
+      "loss": 1.25,
+      "step": 8417
+    },
+    {
+      "epoch": 0.398862828713575,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001313263005595359,
+      "loss": 0.9613,
+      "step": 8418
+    },
+    {
+      "epoch": 0.3989102108505094,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001313121572161524,
+      "loss": 1.1156,
+      "step": 8419
+    },
+    {
+      "epoch": 0.3989575929874437,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013129801317829858,
+      "loss": 0.5553,
+      "step": 8420
+    },
+    {
+      "epoch": 0.3990049751243781,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013128386844628824,
+      "loss": 1.3128,
+      "step": 8421
+    },
+    {
+      "epoch": 0.3990523572613125,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00013126972302043497,
+      "loss": 1.3793,
+      "step": 8422
+    },
+    {
+      "epoch": 0.39909973939824683,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00013125557690105258,
+      "loss": 1.0913,
+      "step": 8423
+    },
+    {
+      "epoch": 0.3991471215351812,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.00013124143008845474,
+      "loss": 0.0059,
+      "step": 8424
+    },
+    {
+      "epoch": 0.3991945036721156,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00013122728258295528,
+      "loss": 1.249,
+      "step": 8425
+    },
+    {
+      "epoch": 0.39924188580905,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013121313438486799,
+      "loss": 1.2313,
+      "step": 8426
+    },
+    {
+      "epoch": 0.39928926794598435,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001311989854945066,
+      "loss": 0.0372,
+      "step": 8427
+    },
+    {
+      "epoch": 0.39933665008291874,
+      "grad_norm": 0.03076171875,
+      "learning_rate": 0.00013118483591218494,
+      "loss": 0.0015,
+      "step": 8428
+    },
+    {
+      "epoch": 0.39938403221985314,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00013117068563821684,
+      "loss": 1.2636,
+      "step": 8429
+    },
+    {
+      "epoch": 0.3994314143567875,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013115653467291613,
+      "loss": 1.565,
+      "step": 8430
+    },
+    {
+      "epoch": 0.39947879649372187,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00013114238301659668,
+      "loss": 0.0358,
+      "step": 8431
+    },
+    {
+      "epoch": 0.39952617863065626,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001311282306695723,
+      "loss": 0.8851,
+      "step": 8432
+    },
+    {
+      "epoch": 0.3995735607675906,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013111407763215696,
+      "loss": 0.1173,
+      "step": 8433
+    },
+    {
+      "epoch": 0.399620942904525,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013109992390466455,
+      "loss": 0.9858,
+      "step": 8434
+    },
+    {
+      "epoch": 0.3996683250414594,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013108576948740893,
+      "loss": 0.1175,
+      "step": 8435
+    },
+    {
+      "epoch": 0.3997157071783937,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00013107161438070405,
+      "loss": 0.6959,
+      "step": 8436
+    },
+    {
+      "epoch": 0.3997630893153281,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013105745858486384,
+      "loss": 0.694,
+      "step": 8437
+    },
+    {
+      "epoch": 0.3998104714522625,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001310433021002023,
+      "loss": 0.0261,
+      "step": 8438
+    },
+    {
+      "epoch": 0.3998578535891969,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00013102914492703335,
+      "loss": 0.6536,
+      "step": 8439
+    },
+    {
+      "epoch": 0.39990523572613124,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013101498706567105,
+      "loss": 1.1436,
+      "step": 8440
+    },
+    {
+      "epoch": 0.39995261786306563,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00013100082851642939,
+      "loss": 1.562,
+      "step": 8441
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00013098666927962235,
+      "loss": 0.1947,
+      "step": 8442
+    },
+    {
+      "epoch": 0.40004738213693436,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00013097250935556397,
+      "loss": 0.1247,
+      "step": 8443
+    },
+    {
+      "epoch": 0.40009476427386875,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013095834874456835,
+      "loss": 0.6569,
+      "step": 8444
+    },
+    {
+      "epoch": 0.40014214641080315,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001309441874469495,
+      "loss": 0.1322,
+      "step": 8445
+    },
+    {
+      "epoch": 0.4001895285477375,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013093002546302158,
+      "loss": 0.1332,
+      "step": 8446
+    },
+    {
+      "epoch": 0.4002369106846719,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001309158627930986,
+      "loss": 1.1967,
+      "step": 8447
+    },
+    {
+      "epoch": 0.40028429282160627,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.4765,
+      "step": 8448
+    },
+    {
+      "epoch": 0.4003316749585406,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013088753539652412,
+      "loss": 0.6844,
+      "step": 8449
+    },
+    {
+      "epoch": 0.400379057095475,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00013087337067050082,
+      "loss": 0.9418,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4004264392324094,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013085920525973904,
+      "loss": 0.6466,
+      "step": 8451
+    },
+    {
+      "epoch": 0.40047382136934373,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013084503916455301,
+      "loss": 0.9358,
+      "step": 8452
+    },
+    {
+      "epoch": 0.4005212035062781,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00013083087238525685,
+      "loss": 1.5236,
+      "step": 8453
+    },
+    {
+      "epoch": 0.4005685856432125,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00013081670492216474,
+      "loss": 0.8968,
+      "step": 8454
+    },
+    {
+      "epoch": 0.4006159677801469,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013080253677559095,
+      "loss": 0.7483,
+      "step": 8455
+    },
+    {
+      "epoch": 0.40066334991708125,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00013078836794584971,
+      "loss": 0.0287,
+      "step": 8456
+    },
+    {
+      "epoch": 0.40071073205401564,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001307741984332553,
+      "loss": 0.7295,
+      "step": 8457
+    },
+    {
+      "epoch": 0.40075811419095003,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001307600282381219,
+      "loss": 0.9218,
+      "step": 8458
+    },
+    {
+      "epoch": 0.40080549632788437,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00013074585736076386,
+      "loss": 0.6534,
+      "step": 8459
+    },
+    {
+      "epoch": 0.40085287846481876,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00013073168580149546,
+      "loss": 1.2929,
+      "step": 8460
+    },
+    {
+      "epoch": 0.40090026060175316,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00013071751356063098,
+      "loss": 0.0068,
+      "step": 8461
+    },
+    {
+      "epoch": 0.4009476427386875,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00013070334063848478,
+      "loss": 1.1136,
+      "step": 8462
+    },
+    {
+      "epoch": 0.4009950248756219,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001306891670353712,
+      "loss": 0.0706,
+      "step": 8463
+    },
+    {
+      "epoch": 0.4010424070125563,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013067499275160459,
+      "loss": 0.129,
+      "step": 8464
+    },
+    {
+      "epoch": 0.4010897891494906,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00013066081778749928,
+      "loss": 0.1445,
+      "step": 8465
+    },
+    {
+      "epoch": 0.401137171286425,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001306466421433697,
+      "loss": 0.8519,
+      "step": 8466
+    },
+    {
+      "epoch": 0.4011845534233594,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013063246581953027,
+      "loss": 0.8091,
+      "step": 8467
+    },
+    {
+      "epoch": 0.4012319355602938,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00013061828881629533,
+      "loss": 0.0299,
+      "step": 8468
+    },
+    {
+      "epoch": 0.40127931769722813,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013060411113397936,
+      "loss": 0.8453,
+      "step": 8469
+    },
+    {
+      "epoch": 0.4013266998341625,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00013058993277289684,
+      "loss": 1.1901,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4013740819710969,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001305757537333622,
+      "loss": 1.4622,
+      "step": 8471
+    },
+    {
+      "epoch": 0.40142146410803126,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00013056157401568984,
+      "loss": 0.3346,
+      "step": 8472
+    },
+    {
+      "epoch": 0.40146884624496565,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013054739362019442,
+      "loss": 0.9678,
+      "step": 8473
+    },
+    {
+      "epoch": 0.40151622838190004,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013053321254719028,
+      "loss": 0.4233,
+      "step": 8474
+    },
+    {
+      "epoch": 0.4015636105188344,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013051903079699205,
+      "loss": 0.9604,
+      "step": 8475
+    },
+    {
+      "epoch": 0.4016109926557688,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013050484836991418,
+      "loss": 0.8562,
+      "step": 8476
+    },
+    {
+      "epoch": 0.40165837479270317,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013049066526627128,
+      "loss": 1.1919,
+      "step": 8477
+    },
+    {
+      "epoch": 0.4017057569296375,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013047648148637787,
+      "loss": 0.9606,
+      "step": 8478
+    },
+    {
+      "epoch": 0.4017531390665719,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00013046229703054862,
+      "loss": 1.0702,
+      "step": 8479
+    },
+    {
+      "epoch": 0.4018005212035063,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013044811189909803,
+      "loss": 1.1904,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4018479033404406,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013043392609234078,
+      "loss": 0.7271,
+      "step": 8481
+    },
+    {
+      "epoch": 0.401895285477375,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00013041973961059147,
+      "loss": 1.1301,
+      "step": 8482
+    },
+    {
+      "epoch": 0.4019426676143094,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001304055524541647,
+      "loss": 0.0194,
+      "step": 8483
+    },
+    {
+      "epoch": 0.4019900497512438,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001303913646233752,
+      "loss": 0.6856,
+      "step": 8484
+    },
+    {
+      "epoch": 0.40203743188817814,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013037717611853758,
+      "loss": 1.0049,
+      "step": 8485
+    },
+    {
+      "epoch": 0.40208481402511254,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013036298693996657,
+      "loss": 0.7963,
+      "step": 8486
+    },
+    {
+      "epoch": 0.40213219616204693,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013034879708797683,
+      "loss": 0.5421,
+      "step": 8487
+    },
+    {
+      "epoch": 0.40217957829898127,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013033460656288308,
+      "loss": 1.0176,
+      "step": 8488
+    },
+    {
+      "epoch": 0.40222696043591566,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001303204153650001,
+      "loss": 0.8604,
+      "step": 8489
+    },
+    {
+      "epoch": 0.40227434257285005,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00013030622349464262,
+      "loss": 0.9674,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4023217247097844,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013029203095212534,
+      "loss": 0.9742,
+      "step": 8491
+    },
+    {
+      "epoch": 0.4023691068467188,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001302778377377631,
+      "loss": 0.034,
+      "step": 8492
+    },
+    {
+      "epoch": 0.4024164889836532,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013026364385187065,
+      "loss": 1.0015,
+      "step": 8493
+    },
+    {
+      "epoch": 0.4024638711205875,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00013024944929476283,
+      "loss": 1.2935,
+      "step": 8494
+    },
+    {
+      "epoch": 0.4025112532575219,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013023525406675445,
+      "loss": 1.0523,
+      "step": 8495
+    },
+    {
+      "epoch": 0.4025586353944563,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013022105816816034,
+      "loss": 1.1254,
+      "step": 8496
+    },
+    {
+      "epoch": 0.4026060175313907,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013020686159929536,
+      "loss": 0.8375,
+      "step": 8497
+    },
+    {
+      "epoch": 0.40265339966832503,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00013019266436047438,
+      "loss": 1.2827,
+      "step": 8498
+    },
+    {
+      "epoch": 0.4027007818052594,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013017846645201221,
+      "loss": 0.343,
+      "step": 8499
+    },
+    {
+      "epoch": 0.4027481639421938,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00013016426787422383,
+      "loss": 1.2327,
+      "step": 8500
+    },
+    {
+      "epoch": 0.40279554607912815,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001301500686274241,
+      "loss": 0.943,
+      "step": 8501
+    },
+    {
+      "epoch": 0.40284292821606255,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013013586871192797,
+      "loss": 0.0642,
+      "step": 8502
+    },
+    {
+      "epoch": 0.40289031035299694,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013012166812805039,
+      "loss": 0.1377,
+      "step": 8503
+    },
+    {
+      "epoch": 0.4029376924899313,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001301074668761063,
+      "loss": 0.0106,
+      "step": 8504
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00013009326495641061,
+      "loss": 0.0096,
+      "step": 8505
+    },
+    {
+      "epoch": 0.40303245676380006,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001300790623692784,
+      "loss": 1.1677,
+      "step": 8506
+    },
+    {
+      "epoch": 0.4030798389007344,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001300648591150246,
+      "loss": 0.7628,
+      "step": 8507
+    },
+    {
+      "epoch": 0.4031272210376688,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013005065519396429,
+      "loss": 0.0236,
+      "step": 8508
+    },
+    {
+      "epoch": 0.4031746031746032,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013003645060641244,
+      "loss": 1.0059,
+      "step": 8509
+    },
+    {
+      "epoch": 0.4032219853115375,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001300222453526841,
+      "loss": 0.0604,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4032693674484719,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013000803943309433,
+      "loss": 0.9209,
+      "step": 8511
+    },
+    {
+      "epoch": 0.4033167495854063,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001299938328479582,
+      "loss": 1.0169,
+      "step": 8512
+    },
+    {
+      "epoch": 0.4033641317223407,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012997962559759086,
+      "loss": 1.153,
+      "step": 8513
+    },
+    {
+      "epoch": 0.40341151385927504,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001299654176823073,
+      "loss": 0.0071,
+      "step": 8514
+    },
+    {
+      "epoch": 0.40345889599620943,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001299512091024227,
+      "loss": 1.1245,
+      "step": 8515
+    },
+    {
+      "epoch": 0.4035062781331438,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001299369998582522,
+      "loss": 1.1065,
+      "step": 8516
+    },
+    {
+      "epoch": 0.40355366027007816,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012992278995011095,
+      "loss": 1.4431,
+      "step": 8517
+    },
+    {
+      "epoch": 0.40360104240701256,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012990857937831405,
+      "loss": 0.6911,
+      "step": 8518
+    },
+    {
+      "epoch": 0.40364842454394695,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0001298943681431767,
+      "loss": 0.0012,
+      "step": 8519
+    },
+    {
+      "epoch": 0.4036958066808813,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012988015624501418,
+      "loss": 0.0266,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4037431888178157,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012986594368414158,
+      "loss": 1.0844,
+      "step": 8521
+    },
+    {
+      "epoch": 0.40379057095475007,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00012985173046087416,
+      "loss": 0.0193,
+      "step": 8522
+    },
+    {
+      "epoch": 0.4038379530916844,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00012983751657552714,
+      "loss": 0.0063,
+      "step": 8523
+    },
+    {
+      "epoch": 0.4038853352286188,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001298233020284158,
+      "loss": 0.4996,
+      "step": 8524
+    },
+    {
+      "epoch": 0.4039327173655532,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001298090868198554,
+      "loss": 0.1321,
+      "step": 8525
+    },
+    {
+      "epoch": 0.4039800995024876,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012979487095016122,
+      "loss": 0.8057,
+      "step": 8526
+    },
+    {
+      "epoch": 0.4040274816394219,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00012978065441964848,
+      "loss": 0.8067,
+      "step": 8527
+    },
+    {
+      "epoch": 0.4040748637763563,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0001297664372286326,
+      "loss": 0.0177,
+      "step": 8528
+    },
+    {
+      "epoch": 0.4041222459132907,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012975221937742886,
+      "loss": 0.9545,
+      "step": 8529
+    },
+    {
+      "epoch": 0.40416962805022505,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00012973800086635252,
+      "loss": 0.016,
+      "step": 8530
+    },
+    {
+      "epoch": 0.40421701018715944,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00012972378169571905,
+      "loss": 0.0937,
+      "step": 8531
+    },
+    {
+      "epoch": 0.40426439232409384,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00012970956186584374,
+      "loss": 0.0335,
+      "step": 8532
+    },
+    {
+      "epoch": 0.4043117744610282,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.000129695341377042,
+      "loss": 0.0048,
+      "step": 8533
+    },
+    {
+      "epoch": 0.40435915659796257,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012968112022962918,
+      "loss": 0.8922,
+      "step": 8534
+    },
+    {
+      "epoch": 0.40440653873489696,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00012966689842392075,
+      "loss": 0.1109,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4044539208718313,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00012965267596023216,
+      "loss": 0.1531,
+      "step": 8536
+    },
+    {
+      "epoch": 0.4045013030087657,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00012963845283887876,
+      "loss": 1.0037,
+      "step": 8537
+    },
+    {
+      "epoch": 0.4045486851457001,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012962422906017605,
+      "loss": 0.0206,
+      "step": 8538
+    },
+    {
+      "epoch": 0.4045960672826344,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012961000462443948,
+      "loss": 1.3797,
+      "step": 8539
+    },
+    {
+      "epoch": 0.4046434494195688,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012959577953198457,
+      "loss": 0.9688,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4046908315565032,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00012958155378312678,
+      "loss": 1.0048,
+      "step": 8541
+    },
+    {
+      "epoch": 0.4047382136934376,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001295673273781816,
+      "loss": 0.6154,
+      "step": 8542
+    },
+    {
+      "epoch": 0.40478559583037194,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012955310031746467,
+      "loss": 0.7247,
+      "step": 8543
+    },
+    {
+      "epoch": 0.40483297796730633,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012953887260129144,
+      "loss": 0.9636,
+      "step": 8544
+    },
+    {
+      "epoch": 0.4048803601042407,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00012952464422997744,
+      "loss": 0.9512,
+      "step": 8545
+    },
+    {
+      "epoch": 0.40492774224117506,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001295104152038383,
+      "loss": 1.5034,
+      "step": 8546
+    },
+    {
+      "epoch": 0.40497512437810945,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 0.00012949618552318955,
+      "loss": 0.0033,
+      "step": 8547
+    },
+    {
+      "epoch": 0.40502250651504385,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012948195518834688,
+      "loss": 1.1164,
+      "step": 8548
+    },
+    {
+      "epoch": 0.4050698886519782,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001294677241996258,
+      "loss": 0.9889,
+      "step": 8549
+    },
+    {
+      "epoch": 0.4051172707889126,
+      "grad_norm": 1.0,
+      "learning_rate": 0.000129453492557342,
+      "loss": 1.658,
+      "step": 8550
+    },
+    {
+      "epoch": 0.40516465292584697,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00012943926026181112,
+      "loss": 1.1179,
+      "step": 8551
+    },
+    {
+      "epoch": 0.4052120350627813,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00012942502731334884,
+      "loss": 0.413,
+      "step": 8552
+    },
+    {
+      "epoch": 0.4052594171997157,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012941079371227074,
+      "loss": 0.9343,
+      "step": 8553
+    },
+    {
+      "epoch": 0.4053067993366501,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012939655945889256,
+      "loss": 0.7487,
+      "step": 8554
+    },
+    {
+      "epoch": 0.4053541814735845,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00012938232455353004,
+      "loss": 0.0043,
+      "step": 8555
+    },
+    {
+      "epoch": 0.4054015636105188,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00012936808899649882,
+      "loss": 1.2496,
+      "step": 8556
+    },
+    {
+      "epoch": 0.4054489457474532,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00012935385278811467,
+      "loss": 1.3019,
+      "step": 8557
+    },
+    {
+      "epoch": 0.4054963278843876,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012933961592869335,
+      "loss": 0.9618,
+      "step": 8558
+    },
+    {
+      "epoch": 0.40554371002132195,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001293253784185506,
+      "loss": 1.1096,
+      "step": 8559
+    },
+    {
+      "epoch": 0.40559109215825634,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001293111402580022,
+      "loss": 0.8938,
+      "step": 8560
+    },
+    {
+      "epoch": 0.40563847429519073,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001292969014473639,
+      "loss": 1.2005,
+      "step": 8561
+    },
+    {
+      "epoch": 0.40568585643212507,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00012928266198695154,
+      "loss": 1.3525,
+      "step": 8562
+    },
+    {
+      "epoch": 0.40573323856905946,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012926842187708094,
+      "loss": 0.9494,
+      "step": 8563
+    },
+    {
+      "epoch": 0.40578062070599386,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001292541811180679,
+      "loss": 0.7537,
+      "step": 8564
+    },
+    {
+      "epoch": 0.4058280028429282,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00012923993971022832,
+      "loss": 1.1526,
+      "step": 8565
+    },
+    {
+      "epoch": 0.4058753849798626,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000129225697653878,
+      "loss": 0.9039,
+      "step": 8566
+    },
+    {
+      "epoch": 0.405922767116797,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00012921145494933285,
+      "loss": 1.4967,
+      "step": 8567
+    },
+    {
+      "epoch": 0.4059701492537313,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00012919721159690873,
+      "loss": 1.018,
+      "step": 8568
+    },
+    {
+      "epoch": 0.4060175313906657,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012918296759692154,
+      "loss": 0.1771,
+      "step": 8569
+    },
+    {
+      "epoch": 0.4060649135276001,
+      "grad_norm": 0.01397705078125,
+      "learning_rate": 0.00012916872294968724,
+      "loss": 0.0008,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4061122956645345,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.00012915447765552172,
+      "loss": 0.0083,
+      "step": 8571
+    },
+    {
+      "epoch": 0.40615967780146883,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00012914023171474096,
+      "loss": 0.0438,
+      "step": 8572
+    },
+    {
+      "epoch": 0.4062070599384032,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012912598512766085,
+      "loss": 0.7081,
+      "step": 8573
+    },
+    {
+      "epoch": 0.4062544420753376,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00012911173789459745,
+      "loss": 0.3742,
+      "step": 8574
+    },
+    {
+      "epoch": 0.40630182421227196,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012909749001586672,
+      "loss": 1.0059,
+      "step": 8575
+    },
+    {
+      "epoch": 0.40634920634920635,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00012908324149178463,
+      "loss": 0.0195,
+      "step": 8576
+    },
+    {
+      "epoch": 0.40639658848614074,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012906899232266724,
+      "loss": 0.0254,
+      "step": 8577
+    },
+    {
+      "epoch": 0.4064439706230751,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00012905474250883057,
+      "loss": 0.0615,
+      "step": 8578
+    },
+    {
+      "epoch": 0.4064913527600095,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012904049205059065,
+      "loss": 0.5947,
+      "step": 8579
+    },
+    {
+      "epoch": 0.40653873489694387,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012902624094826353,
+      "loss": 0.644,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4065861170338782,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012901198920216528,
+      "loss": 1.1777,
+      "step": 8581
+    },
+    {
+      "epoch": 0.4066334991708126,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012899773681261206,
+      "loss": 0.5417,
+      "step": 8582
+    },
+    {
+      "epoch": 0.406680881307747,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012898348377991993,
+      "loss": 0.8528,
+      "step": 8583
+    },
+    {
+      "epoch": 0.4067282634446814,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00012896923010440498,
+      "loss": 0.0175,
+      "step": 8584
+    },
+    {
+      "epoch": 0.4067756455816157,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001289549757863834,
+      "loss": 0.9764,
+      "step": 8585
+    },
+    {
+      "epoch": 0.4068230277185501,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012894072082617126,
+      "loss": 0.6498,
+      "step": 8586
+    },
+    {
+      "epoch": 0.4068704098554845,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012892646522408475,
+      "loss": 0.3805,
+      "step": 8587
+    },
+    {
+      "epoch": 0.40691779199241884,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012891220898044006,
+      "loss": 0.1385,
+      "step": 8588
+    },
+    {
+      "epoch": 0.40696517412935324,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001288979520955534,
+      "loss": 0.8701,
+      "step": 8589
+    },
+    {
+      "epoch": 0.40701255626628763,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001288836945697409,
+      "loss": 0.2025,
+      "step": 8590
+    },
+    {
+      "epoch": 0.40705993840322197,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012886943640331885,
+      "loss": 0.7638,
+      "step": 8591
+    },
+    {
+      "epoch": 0.40710732054015636,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012885517759660345,
+      "loss": 1.0234,
+      "step": 8592
+    },
+    {
+      "epoch": 0.40715470267709075,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012884091814991094,
+      "loss": 1.3021,
+      "step": 8593
+    },
+    {
+      "epoch": 0.4072020848140251,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012882665806355757,
+      "loss": 0.7696,
+      "step": 8594
+    },
+    {
+      "epoch": 0.4072494669509595,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012881239733785965,
+      "loss": 0.7608,
+      "step": 8595
+    },
+    {
+      "epoch": 0.4072968490878939,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00012879813597313345,
+      "loss": 0.7023,
+      "step": 8596
+    },
+    {
+      "epoch": 0.4073442312248282,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00012878387396969524,
+      "loss": 0.8593,
+      "step": 8597
+    },
+    {
+      "epoch": 0.4073916133617626,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001287696113278614,
+      "loss": 1.1883,
+      "step": 8598
+    },
+    {
+      "epoch": 0.407438995498697,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001287553480479482,
+      "loss": 0.5168,
+      "step": 8599
+    },
+    {
+      "epoch": 0.4074863776356314,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000128741084130272,
+      "loss": 0.5142,
+      "step": 8600
+    },
+    {
+      "epoch": 0.40753375977256573,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012872681957514922,
+      "loss": 0.1935,
+      "step": 8601
+    },
+    {
+      "epoch": 0.4075811419095001,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00012871255438289613,
+      "loss": 0.0267,
+      "step": 8602
+    },
+    {
+      "epoch": 0.4076285240464345,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012869828855382917,
+      "loss": 1.2798,
+      "step": 8603
+    },
+    {
+      "epoch": 0.40767590618336885,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012868402208826473,
+      "loss": 1.0788,
+      "step": 8604
+    },
+    {
+      "epoch": 0.40772328832030325,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00012866975498651928,
+      "loss": 0.1732,
+      "step": 8605
+    },
+    {
+      "epoch": 0.40777067045723764,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001286554872489092,
+      "loss": 1.135,
+      "step": 8606
+    },
+    {
+      "epoch": 0.407818052594172,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001286412188757509,
+      "loss": 0.6741,
+      "step": 8607
+    },
+    {
+      "epoch": 0.40786543473110637,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00012862694986736086,
+      "loss": 0.8214,
+      "step": 8608
+    },
+    {
+      "epoch": 0.40791281686804076,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012861268022405558,
+      "loss": 0.7118,
+      "step": 8609
+    },
+    {
+      "epoch": 0.4079601990049751,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012859840994615155,
+      "loss": 0.7582,
+      "step": 8610
+    },
+    {
+      "epoch": 0.4080075811419095,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012858413903396522,
+      "loss": 0.7089,
+      "step": 8611
+    },
+    {
+      "epoch": 0.4080549632788439,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00012856986748781317,
+      "loss": 0.9185,
+      "step": 8612
+    },
+    {
+      "epoch": 0.4081023454157783,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00012855559530801185,
+      "loss": 0.6825,
+      "step": 8613
+    },
+    {
+      "epoch": 0.4081497275527126,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012854132249487786,
+      "loss": 0.7458,
+      "step": 8614
+    },
+    {
+      "epoch": 0.408197109689647,
+      "grad_norm": 0.0084228515625,
+      "learning_rate": 0.00012852704904872775,
+      "loss": 0.0001,
+      "step": 8615
+    },
+    {
+      "epoch": 0.4082444918265814,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00012851277496987807,
+      "loss": 0.033,
+      "step": 8616
+    },
+    {
+      "epoch": 0.40829187396351574,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001284985002586454,
+      "loss": 0.7717,
+      "step": 8617
+    },
+    {
+      "epoch": 0.40833925610045013,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012848422491534636,
+      "loss": 0.5768,
+      "step": 8618
+    },
+    {
+      "epoch": 0.4083866382373845,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012846994894029755,
+      "loss": 1.0762,
+      "step": 8619
+    },
+    {
+      "epoch": 0.40843402037431886,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001284556723338156,
+      "loss": 1.1826,
+      "step": 8620
+    },
+    {
+      "epoch": 0.40848140251125326,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00012844139509621714,
+      "loss": 0.0071,
+      "step": 8621
+    },
+    {
+      "epoch": 0.40852878464818765,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012842711722781884,
+      "loss": 1.2741,
+      "step": 8622
+    },
+    {
+      "epoch": 0.408576166785122,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012841283872893735,
+      "loss": 0.7631,
+      "step": 8623
+    },
+    {
+      "epoch": 0.4086235489220564,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012839855959988935,
+      "loss": 0.9233,
+      "step": 8624
+    },
+    {
+      "epoch": 0.40867093105899077,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012838427984099158,
+      "loss": 1.0251,
+      "step": 8625
+    },
+    {
+      "epoch": 0.4087183131959251,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012836999945256067,
+      "loss": 0.7825,
+      "step": 8626
+    },
+    {
+      "epoch": 0.4087656953328595,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012835571843491342,
+      "loss": 0.4097,
+      "step": 8627
+    },
+    {
+      "epoch": 0.4088130774697939,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012834143678836654,
+      "loss": 0.5478,
+      "step": 8628
+    },
+    {
+      "epoch": 0.4088604596067283,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00012832715451323678,
+      "loss": 1.1211,
+      "step": 8629
+    },
+    {
+      "epoch": 0.4089078417436626,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012831287160984092,
+      "loss": 0.9444,
+      "step": 8630
+    },
+    {
+      "epoch": 0.408955223880597,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012829858807849567,
+      "loss": 0.1511,
+      "step": 8631
+    },
+    {
+      "epoch": 0.4090026060175314,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00012828430391951794,
+      "loss": 1.0221,
+      "step": 8632
+    },
+    {
+      "epoch": 0.40904998815446575,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012827001913322445,
+      "loss": 1.001,
+      "step": 8633
+    },
+    {
+      "epoch": 0.40909737029140014,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012825573371993206,
+      "loss": 0.9136,
+      "step": 8634
+    },
+    {
+      "epoch": 0.40914475242833453,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012824144767995758,
+      "loss": 0.1502,
+      "step": 8635
+    },
+    {
+      "epoch": 0.4091921345652689,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012822716101361788,
+      "loss": 0.1261,
+      "step": 8636
+    },
+    {
+      "epoch": 0.40923951670220327,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012821287372122978,
+      "loss": 0.3275,
+      "step": 8637
+    },
+    {
+      "epoch": 0.40928689883913766,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012819858580311022,
+      "loss": 1.0782,
+      "step": 8638
+    },
+    {
+      "epoch": 0.409334280976072,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012818429725957605,
+      "loss": 0.9956,
+      "step": 8639
+    },
+    {
+      "epoch": 0.4093816631130064,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012817000809094424,
+      "loss": 0.0998,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4094290452499408,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001281557182975316,
+      "loss": 0.9675,
+      "step": 8641
+    },
+    {
+      "epoch": 0.4094764273868752,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00012814142787965512,
+      "loss": 1.1673,
+      "step": 8642
+    },
+    {
+      "epoch": 0.4095238095238095,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001281271368376318,
+      "loss": 0.1159,
+      "step": 8643
+    },
+    {
+      "epoch": 0.4095711916607439,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00012811284517177848,
+      "loss": 1.1577,
+      "step": 8644
+    },
+    {
+      "epoch": 0.4096185737976783,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001280985528824122,
+      "loss": 0.7769,
+      "step": 8645
+    },
+    {
+      "epoch": 0.40966595593461264,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012808425996985,
+      "loss": 0.0316,
+      "step": 8646
+    },
+    {
+      "epoch": 0.40971333807154703,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012806996643440881,
+      "loss": 0.0188,
+      "step": 8647
+    },
+    {
+      "epoch": 0.4097607202084814,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012805567227640565,
+      "loss": 0.8488,
+      "step": 8648
+    },
+    {
+      "epoch": 0.40980810234541576,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012804137749615753,
+      "loss": 0.0344,
+      "step": 8649
+    },
+    {
+      "epoch": 0.40985548448235015,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012802708209398155,
+      "loss": 0.7397,
+      "step": 8650
+    },
+    {
+      "epoch": 0.40990286661928454,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00012801278607019477,
+      "loss": 0.8155,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4099502487562189,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012799848942511422,
+      "loss": 0.1896,
+      "step": 8652
+    },
+    {
+      "epoch": 0.4099976308931533,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000127984192159057,
+      "loss": 1.0832,
+      "step": 8653
+    },
+    {
+      "epoch": 0.41004501303008767,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012796989427234017,
+      "loss": 1.272,
+      "step": 8654
+    },
+    {
+      "epoch": 0.410092395167022,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012795559576528093,
+      "loss": 0.0558,
+      "step": 8655
+    },
+    {
+      "epoch": 0.4101397773039564,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001279412966381963,
+      "loss": 1.2894,
+      "step": 8656
+    },
+    {
+      "epoch": 0.4101871594408908,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012792699689140351,
+      "loss": 0.0748,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4102345415778252,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012791269652521965,
+      "loss": 0.7179,
+      "step": 8658
+    },
+    {
+      "epoch": 0.4102819237147595,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012789839553996194,
+      "loss": 1.427,
+      "step": 8659
+    },
+    {
+      "epoch": 0.4103293058516939,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012788409393594752,
+      "loss": 0.7396,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4103766879886283,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001278697917134936,
+      "loss": 0.7206,
+      "step": 8661
+    },
+    {
+      "epoch": 0.41042407012556265,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012785548887291737,
+      "loss": 0.8337,
+      "step": 8662
+    },
+    {
+      "epoch": 0.41047145226249704,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.00012784118541453609,
+      "loss": 0.0057,
+      "step": 8663
+    },
+    {
+      "epoch": 0.41051883439943143,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012782688133866697,
+      "loss": 1.4189,
+      "step": 8664
+    },
+    {
+      "epoch": 0.41056621653636577,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00012781257664562726,
+      "loss": 0.252,
+      "step": 8665
+    },
+    {
+      "epoch": 0.41061359867330016,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012779827133573422,
+      "loss": 1.1228,
+      "step": 8666
+    },
+    {
+      "epoch": 0.41066098081023455,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012778396540930512,
+      "loss": 0.5086,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4107083629471689,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012776965886665727,
+      "loss": 0.7849,
+      "step": 8668
+    },
+    {
+      "epoch": 0.4107557450841033,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012775535170810796,
+      "loss": 0.8292,
+      "step": 8669
+    },
+    {
+      "epoch": 0.4108031272210377,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.00012774104393397452,
+      "loss": 0.0033,
+      "step": 8670
+    },
+    {
+      "epoch": 0.41085050935797207,
+      "grad_norm": 0.0120849609375,
+      "learning_rate": 0.0001277267355445743,
+      "loss": 0.0006,
+      "step": 8671
+    },
+    {
+      "epoch": 0.4108978914949064,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012771242654022454,
+      "loss": 1.0628,
+      "step": 8672
+    },
+    {
+      "epoch": 0.4109452736318408,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012769811692124274,
+      "loss": 0.4934,
+      "step": 8673
+    },
+    {
+      "epoch": 0.4109926557687752,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00012768380668794622,
+      "loss": 1.1352,
+      "step": 8674
+    },
+    {
+      "epoch": 0.41104003790570953,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001276694958406523,
+      "loss": 0.6742,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4110874200426439,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012765518437967847,
+      "loss": 1.153,
+      "step": 8676
+    },
+    {
+      "epoch": 0.4111348021795783,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012764087230534213,
+      "loss": 1.3431,
+      "step": 8677
+    },
+    {
+      "epoch": 0.41118218431651266,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00012762655961796066,
+      "loss": 0.0087,
+      "step": 8678
+    },
+    {
+      "epoch": 0.41122956645344705,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012761224631785152,
+      "loss": 0.7768,
+      "step": 8679
+    },
+    {
+      "epoch": 0.41127694859038144,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012759793240533216,
+      "loss": 1.138,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4113243307273158,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012758361788072007,
+      "loss": 1.1168,
+      "step": 8681
+    },
+    {
+      "epoch": 0.41137171286425017,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001275693027443327,
+      "loss": 0.6417,
+      "step": 8682
+    },
+    {
+      "epoch": 0.41141909500118456,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00012755498699648757,
+      "loss": 0.1818,
+      "step": 8683
+    },
+    {
+      "epoch": 0.4114664771381189,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0001275406706375022,
+      "loss": 1.044,
+      "step": 8684
+    },
+    {
+      "epoch": 0.4115138592750533,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012752635366769405,
+      "loss": 1.1786,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4115612414119877,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012751203608738073,
+      "loss": 0.9176,
+      "step": 8686
+    },
+    {
+      "epoch": 0.4116086235489221,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012749771789687974,
+      "loss": 0.9885,
+      "step": 8687
+    },
+    {
+      "epoch": 0.4116560056858564,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012748339909650865,
+      "loss": 0.6357,
+      "step": 8688
+    },
+    {
+      "epoch": 0.4117033878227908,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00012746907968658508,
+      "loss": 0.0454,
+      "step": 8689
+    },
+    {
+      "epoch": 0.4117507699597252,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.00012745475966742653,
+      "loss": 0.0118,
+      "step": 8690
+    },
+    {
+      "epoch": 0.41179815209665954,
+      "grad_norm": 0.003387451171875,
+      "learning_rate": 0.00012744043903935065,
+      "loss": 0.0001,
+      "step": 8691
+    },
+    {
+      "epoch": 0.41184553423359394,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001274261178026751,
+      "loss": 1.1407,
+      "step": 8692
+    },
+    {
+      "epoch": 0.41189291637052833,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012741179595771748,
+      "loss": 1.2763,
+      "step": 8693
+    },
+    {
+      "epoch": 0.41194029850746267,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00012739747350479542,
+      "loss": 1.1881,
+      "step": 8694
+    },
+    {
+      "epoch": 0.41198768064439706,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012738315044422653,
+      "loss": 1.1318,
+      "step": 8695
+    },
+    {
+      "epoch": 0.41203506278133145,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012736882677632854,
+      "loss": 1.301,
+      "step": 8696
+    },
+    {
+      "epoch": 0.4120824449182658,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012735450250141915,
+      "loss": 0.0163,
+      "step": 8697
+    },
+    {
+      "epoch": 0.4121298270552002,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012734017761981603,
+      "loss": 1.1192,
+      "step": 8698
+    },
+    {
+      "epoch": 0.4121772091921346,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012732585213183687,
+      "loss": 0.7565,
+      "step": 8699
+    },
+    {
+      "epoch": 0.41222459132906897,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012731152603779946,
+      "loss": 0.9723,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4122719734660033,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012729719933802147,
+      "loss": 1.3824,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4123193556029377,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00012728287203282065,
+      "loss": 0.8237,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4123667377398721,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001272685441225148,
+      "loss": 0.891,
+      "step": 8703
+    },
+    {
+      "epoch": 0.41241411987680643,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001272542156074217,
+      "loss": 0.9442,
+      "step": 8704
+    },
+    {
+      "epoch": 0.4124615020137408,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012723988648785913,
+      "loss": 0.7798,
+      "step": 8705
+    },
+    {
+      "epoch": 0.4125088841506752,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001272255567641449,
+      "loss": 0.8677,
+      "step": 8706
+    },
+    {
+      "epoch": 0.41255626628760955,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00012721122643659678,
+      "loss": 0.8635,
+      "step": 8707
+    },
+    {
+      "epoch": 0.41260364842454395,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00012719689550553268,
+      "loss": 0.0828,
+      "step": 8708
+    },
+    {
+      "epoch": 0.41265103056147834,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00012718256397127042,
+      "loss": 1.1555,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4126984126984127,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001271682318341278,
+      "loss": 1.2882,
+      "step": 8710
+    },
+    {
+      "epoch": 0.41274579483534707,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012715389909442274,
+      "loss": 1.1442,
+      "step": 8711
+    },
+    {
+      "epoch": 0.41279317697228146,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012713956575247318,
+      "loss": 1.3723,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4128405591092158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00012712523180859688,
+      "loss": 0.5592,
+      "step": 8713
+    },
+    {
+      "epoch": 0.4128879412461502,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012711089726311187,
+      "loss": 0.7329,
+      "step": 8714
+    },
+    {
+      "epoch": 0.4129353233830846,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012709656211633603,
+      "loss": 0.2984,
+      "step": 8715
+    },
+    {
+      "epoch": 0.412982705520019,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012708222636858733,
+      "loss": 1.2764,
+      "step": 8716
+    },
+    {
+      "epoch": 0.4130300876569533,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012706789002018366,
+      "loss": 0.9975,
+      "step": 8717
+    },
+    {
+      "epoch": 0.4130774697938877,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00012705355307144305,
+      "loss": 0.0301,
+      "step": 8718
+    },
+    {
+      "epoch": 0.4131248519308221,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001270392155226834,
+      "loss": 1.0671,
+      "step": 8719
+    },
+    {
+      "epoch": 0.41317223406775644,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001270248773742228,
+      "loss": 0.6159,
+      "step": 8720
+    },
+    {
+      "epoch": 0.41321961620469083,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012701053862637917,
+      "loss": 0.0311,
+      "step": 8721
+    },
+    {
+      "epoch": 0.4132669983416252,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001269961992794706,
+      "loss": 0.9573,
+      "step": 8722
+    },
+    {
+      "epoch": 0.41331438047855956,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012698185933381507,
+      "loss": 1.1541,
+      "step": 8723
+    },
+    {
+      "epoch": 0.41336176261549396,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012696751878973065,
+      "loss": 1.4427,
+      "step": 8724
+    },
+    {
+      "epoch": 0.41340914475242835,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012695317764753537,
+      "loss": 0.9246,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4134565268893627,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001269388359075473,
+      "loss": 0.7327,
+      "step": 8726
+    },
+    {
+      "epoch": 0.4135039090262971,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0001269244935700846,
+      "loss": 0.9051,
+      "step": 8727
+    },
+    {
+      "epoch": 0.41355129116323147,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012691015063546525,
+      "loss": 0.1719,
+      "step": 8728
+    },
+    {
+      "epoch": 0.41359867330016586,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012689580710400746,
+      "loss": 0.9895,
+      "step": 8729
+    },
+    {
+      "epoch": 0.4136460554371002,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00012688146297602928,
+      "loss": 0.0378,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4136934375740346,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012686711825184895,
+      "loss": 0.9236,
+      "step": 8731
+    },
+    {
+      "epoch": 0.413740819710969,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001268527729317845,
+      "loss": 1.1369,
+      "step": 8732
+    },
+    {
+      "epoch": 0.4137882018479033,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00012683842701615417,
+      "loss": 0.0278,
+      "step": 8733
+    },
+    {
+      "epoch": 0.4138355839848377,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001268240805052761,
+      "loss": 0.6815,
+      "step": 8734
+    },
+    {
+      "epoch": 0.4138829661217721,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00012680973339946854,
+      "loss": 0.0055,
+      "step": 8735
+    },
+    {
+      "epoch": 0.41393034825870645,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001267953856990496,
+      "loss": 1.0057,
+      "step": 8736
+    },
+    {
+      "epoch": 0.41397773039564084,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012678103740433754,
+      "loss": 0.7069,
+      "step": 8737
+    },
+    {
+      "epoch": 0.41402511253257523,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001267666885156506,
+      "loss": 0.0898,
+      "step": 8738
+    },
+    {
+      "epoch": 0.41407249466950957,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012675233903330707,
+      "loss": 0.9589,
+      "step": 8739
+    },
+    {
+      "epoch": 0.41411987680644397,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012673798895762513,
+      "loss": 1.1259,
+      "step": 8740
+    },
+    {
+      "epoch": 0.41416725894337836,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012672363828892307,
+      "loss": 0.5192,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4142146410803127,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00012670928702751915,
+      "loss": 0.0111,
+      "step": 8742
+    },
+    {
+      "epoch": 0.4142620232172471,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012669493517373175,
+      "loss": 0.9285,
+      "step": 8743
+    },
+    {
+      "epoch": 0.4143094053541815,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00012668058272787908,
+      "loss": 0.013,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4143567874911159,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012666622969027952,
+      "loss": 0.6148,
+      "step": 8745
+    },
+    {
+      "epoch": 0.4144041696280502,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00012665187606125138,
+      "loss": 0.7689,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4144515517649846,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012663752184111302,
+      "loss": 1.0613,
+      "step": 8747
+    },
+    {
+      "epoch": 0.414498933901919,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012662316703018279,
+      "loss": 1.1997,
+      "step": 8748
+    },
+    {
+      "epoch": 0.41454631603885334,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012660881162877908,
+      "loss": 0.5236,
+      "step": 8749
+    },
+    {
+      "epoch": 0.41459369817578773,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001265944556372203,
+      "loss": 0.0557,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4146410803127221,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001265800990558248,
+      "loss": 0.132,
+      "step": 8751
+    },
+    {
+      "epoch": 0.41468846244965646,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012656574188491099,
+      "loss": 0.4266,
+      "step": 8752
+    },
+    {
+      "epoch": 0.41473584458659085,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012655138412479732,
+      "loss": 1.1053,
+      "step": 8753
+    },
+    {
+      "epoch": 0.41478322672352524,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012653702577580228,
+      "loss": 1.6178,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4148306088604596,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012652266683824424,
+      "loss": 0.2094,
+      "step": 8755
+    },
+    {
+      "epoch": 0.414877990997394,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001265083073124417,
+      "loss": 0.754,
+      "step": 8756
+    },
+    {
+      "epoch": 0.41492537313432837,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.00012649394719871314,
+      "loss": 0.0055,
+      "step": 8757
+    },
+    {
+      "epoch": 0.41497275527126276,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012647958649737707,
+      "loss": 0.97,
+      "step": 8758
+    },
+    {
+      "epoch": 0.4150201374081971,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00012646522520875196,
+      "loss": 0.0633,
+      "step": 8759
+    },
+    {
+      "epoch": 0.4150675195451315,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012645086333315633,
+      "loss": 1.0843,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4151149016820659,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012643650087090875,
+      "loss": 0.9516,
+      "step": 8761
+    },
+    {
+      "epoch": 0.4151622838190002,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012642213782232775,
+      "loss": 1.0484,
+      "step": 8762
+    },
+    {
+      "epoch": 0.4152096659559346,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012640777418773186,
+      "loss": 1.0012,
+      "step": 8763
+    },
+    {
+      "epoch": 0.415257048092869,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012639340996743967,
+      "loss": 0.5978,
+      "step": 8764
+    },
+    {
+      "epoch": 0.41530443022980335,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00012637904516176975,
+      "loss": 0.4532,
+      "step": 8765
+    },
+    {
+      "epoch": 0.41535181236673774,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001263646797710407,
+      "loss": 0.25,
+      "step": 8766
+    },
+    {
+      "epoch": 0.41539919450367213,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012635031379557116,
+      "loss": 1.0736,
+      "step": 8767
+    },
+    {
+      "epoch": 0.41544657664060647,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012633594723567972,
+      "loss": 1.2801,
+      "step": 8768
+    },
+    {
+      "epoch": 0.41549395877754086,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000126321580091685,
+      "loss": 0.3909,
+      "step": 8769
+    },
+    {
+      "epoch": 0.41554134091447525,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001263072123639057,
+      "loss": 0.0057,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4155887230514096,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012629284405266044,
+      "loss": 0.879,
+      "step": 8771
+    },
+    {
+      "epoch": 0.415636105188344,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0001262784751582679,
+      "loss": 0.0036,
+      "step": 8772
+    },
+    {
+      "epoch": 0.4156834873252784,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001262641056810468,
+      "loss": 0.0827,
+      "step": 8773
+    },
+    {
+      "epoch": 0.41573086946221277,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012624973562131578,
+      "loss": 0.5373,
+      "step": 8774
+    },
+    {
+      "epoch": 0.4157782515991471,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00012623536497939356,
+      "loss": 0.0488,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4158256337360815,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012622099375559894,
+      "loss": 0.0914,
+      "step": 8776
+    },
+    {
+      "epoch": 0.4158730158730159,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012620662195025057,
+      "loss": 0.032,
+      "step": 8777
+    },
+    {
+      "epoch": 0.41592039800995023,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012619224956366724,
+      "loss": 1.1273,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4159677801468846,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012617787659616775,
+      "loss": 0.6303,
+      "step": 8779
+    },
+    {
+      "epoch": 0.416015162283819,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00012616350304807083,
+      "loss": 1.3105,
+      "step": 8780
+    },
+    {
+      "epoch": 0.41606254442075336,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00012614912891969525,
+      "loss": 1.2002,
+      "step": 8781
+    },
+    {
+      "epoch": 0.41610992655768775,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012613475421135987,
+      "loss": 1.2593,
+      "step": 8782
+    },
+    {
+      "epoch": 0.41615730869462214,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012612037892338347,
+      "loss": 1.1892,
+      "step": 8783
+    },
+    {
+      "epoch": 0.4162046908315565,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001261060030560849,
+      "loss": 0.271,
+      "step": 8784
+    },
+    {
+      "epoch": 0.41625207296849087,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012609162660978296,
+      "loss": 0.1717,
+      "step": 8785
+    },
+    {
+      "epoch": 0.41629945510542526,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012607724958479656,
+      "loss": 0.7417,
+      "step": 8786
+    },
+    {
+      "epoch": 0.41634683724235966,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012606287198144454,
+      "loss": 0.5609,
+      "step": 8787
+    },
+    {
+      "epoch": 0.416394219379294,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012604849380004577,
+      "loss": 0.7522,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4164416015162284,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012603411504091917,
+      "loss": 0.8829,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4164889836531628,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012601973570438362,
+      "loss": 0.49,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4165363657900971,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012600535579075802,
+      "loss": 1.0319,
+      "step": 8791
+    },
+    {
+      "epoch": 0.4165837479270315,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012599097530036136,
+      "loss": 0.6672,
+      "step": 8792
+    },
+    {
+      "epoch": 0.4166311300639659,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012597659423351252,
+      "loss": 1.1843,
+      "step": 8793
+    },
+    {
+      "epoch": 0.41667851220090024,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012596221259053052,
+      "loss": 0.6169,
+      "step": 8794
+    },
+    {
+      "epoch": 0.41672589433783463,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001259478303717343,
+      "loss": 1.0632,
+      "step": 8795
+    },
+    {
+      "epoch": 0.416773276474769,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012593344757744286,
+      "loss": 0.8213,
+      "step": 8796
+    },
+    {
+      "epoch": 0.41682065861170337,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00012591906420797512,
+      "loss": 1.3638,
+      "step": 8797
+    },
+    {
+      "epoch": 0.41686804074863776,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012590468026365016,
+      "loss": 0.0186,
+      "step": 8798
+    },
+    {
+      "epoch": 0.41691542288557215,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012589029574478698,
+      "loss": 0.0075,
+      "step": 8799
+    },
+    {
+      "epoch": 0.4169628050225065,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00012587591065170466,
+      "loss": 0.0046,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4170101871594409,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012586152498472218,
+      "loss": 1.0377,
+      "step": 8801
+    },
+    {
+      "epoch": 0.4170575692963753,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001258471387441586,
+      "loss": 1.0316,
+      "step": 8802
+    },
+    {
+      "epoch": 0.41710495143330967,
+      "grad_norm": 0.005462646484375,
+      "learning_rate": 0.00012583275193033303,
+      "loss": 0.0002,
+      "step": 8803
+    },
+    {
+      "epoch": 0.417152333570244,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00012581836454356456,
+      "loss": 0.0459,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4171997157071784,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00012580397658417224,
+      "loss": 1.2865,
+      "step": 8805
+    },
+    {
+      "epoch": 0.4172470978441128,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00012578958805247522,
+      "loss": 0.7229,
+      "step": 8806
+    },
+    {
+      "epoch": 0.41729447998104713,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012577519894879256,
+      "loss": 0.0235,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4173418621179815,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00012576080927344352,
+      "loss": 0.0176,
+      "step": 8808
+    },
+    {
+      "epoch": 0.4173892442549159,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001257464190267471,
+      "loss": 0.6251,
+      "step": 8809
+    },
+    {
+      "epoch": 0.41743662639185025,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012573202820902256,
+      "loss": 1.1001,
+      "step": 8810
+    },
+    {
+      "epoch": 0.41748400852878464,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012571763682058904,
+      "loss": 0.5405,
+      "step": 8811
+    },
+    {
+      "epoch": 0.41753139066571904,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00012570324486176575,
+      "loss": 0.988,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4175787728026534,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001256888523328718,
+      "loss": 0.0729,
+      "step": 8813
+    },
+    {
+      "epoch": 0.41762615493958777,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012567445923422654,
+      "loss": 0.8022,
+      "step": 8814
+    },
+    {
+      "epoch": 0.41767353707652216,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00012566006556614912,
+      "loss": 0.0961,
+      "step": 8815
+    },
+    {
+      "epoch": 0.41772091921345655,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012564567132895873,
+      "loss": 0.932,
+      "step": 8816
+    },
+    {
+      "epoch": 0.4177683013503909,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012563127652297467,
+      "loss": 0.0159,
+      "step": 8817
+    },
+    {
+      "epoch": 0.4178156834873253,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012561688114851624,
+      "loss": 0.6744,
+      "step": 8818
+    },
+    {
+      "epoch": 0.4178630656242597,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00012560248520590263,
+      "loss": 0.556,
+      "step": 8819
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001255880886954532,
+      "loss": 1.4377,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4179578298981284,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001255736916174872,
+      "loss": 1.3856,
+      "step": 8821
+    },
+    {
+      "epoch": 0.4180052120350628,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012555929397232396,
+      "loss": 0.9241,
+      "step": 8822
+    },
+    {
+      "epoch": 0.41805259417199714,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012554489576028282,
+      "loss": 0.9361,
+      "step": 8823
+    },
+    {
+      "epoch": 0.41809997630893153,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001255304969816831,
+      "loss": 0.0571,
+      "step": 8824
+    },
+    {
+      "epoch": 0.4181473584458659,
+      "grad_norm": 0.0196533203125,
+      "learning_rate": 0.0001255160976368441,
+      "loss": 0.0014,
+      "step": 8825
+    },
+    {
+      "epoch": 0.41819474058280026,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001255016977260853,
+      "loss": 0.017,
+      "step": 8826
+    },
+    {
+      "epoch": 0.41824212271973465,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000125487297249726,
+      "loss": 0.5335,
+      "step": 8827
+    },
+    {
+      "epoch": 0.41828950485666905,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00012547289620808556,
+      "loss": 0.751,
+      "step": 8828
+    },
+    {
+      "epoch": 0.4183368869936034,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012545849460148342,
+      "loss": 0.5795,
+      "step": 8829
+    },
+    {
+      "epoch": 0.4183842691305378,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000125444092430239,
+      "loss": 0.018,
+      "step": 8830
+    },
+    {
+      "epoch": 0.41843165126747217,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012542968969467172,
+      "loss": 1.1114,
+      "step": 8831
+    },
+    {
+      "epoch": 0.41847903340440656,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00012541528639510097,
+      "loss": 1.3415,
+      "step": 8832
+    },
+    {
+      "epoch": 0.4185264155413409,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012540088253184628,
+      "loss": 0.1147,
+      "step": 8833
+    },
+    {
+      "epoch": 0.4185737976782753,
+      "grad_norm": 0.0196533203125,
+      "learning_rate": 0.00012538647810522708,
+      "loss": 0.0015,
+      "step": 8834
+    },
+    {
+      "epoch": 0.4186211798152097,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00012537207311556282,
+      "loss": 0.0165,
+      "step": 8835
+    },
+    {
+      "epoch": 0.418668561952144,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000125357667563173,
+      "loss": 0.0302,
+      "step": 8836
+    },
+    {
+      "epoch": 0.4187159440890784,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012534326144837712,
+      "loss": 0.0651,
+      "step": 8837
+    },
+    {
+      "epoch": 0.4187633262260128,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001253288547714947,
+      "loss": 1.0661,
+      "step": 8838
+    },
+    {
+      "epoch": 0.41881070836294715,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012531444753284524,
+      "loss": 0.9368,
+      "step": 8839
+    },
+    {
+      "epoch": 0.41885809049988154,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001253000397327483,
+      "loss": 1.0135,
+      "step": 8840
+    },
+    {
+      "epoch": 0.41890547263681593,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012528563137152344,
+      "loss": 0.989,
+      "step": 8841
+    },
+    {
+      "epoch": 0.41895285477375027,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012527122244949024,
+      "loss": 0.571,
+      "step": 8842
+    },
+    {
+      "epoch": 0.41900023691068466,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001252568129669682,
+      "loss": 0.7921,
+      "step": 8843
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012524240292427697,
+      "loss": 0.8786,
+      "step": 8844
+    },
+    {
+      "epoch": 0.41909500118455345,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012522799232173612,
+      "loss": 0.8598,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4191423833214878,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012521358115966528,
+      "loss": 0.6554,
+      "step": 8846
+    },
+    {
+      "epoch": 0.4191897654584222,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012519916943838405,
+      "loss": 0.698,
+      "step": 8847
+    },
+    {
+      "epoch": 0.4192371475953566,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001251847571582121,
+      "loss": 0.1573,
+      "step": 8848
+    },
+    {
+      "epoch": 0.4192845297322909,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001251703443194691,
+      "loss": 1.1251,
+      "step": 8849
+    },
+    {
+      "epoch": 0.4193319118692253,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001251559309224746,
+      "loss": 1.0505,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4193792940061597,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001251415169675484,
+      "loss": 0.6131,
+      "step": 8851
+    },
+    {
+      "epoch": 0.41942667614309403,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001251271024550101,
+      "loss": 0.8077,
+      "step": 8852
+    },
+    {
+      "epoch": 0.41947405828002843,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00012511268738517943,
+      "loss": 1.0084,
+      "step": 8853
+    },
+    {
+      "epoch": 0.4195214404169628,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012509827175837614,
+      "loss": 0.3676,
+      "step": 8854
+    },
+    {
+      "epoch": 0.41956882255389716,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012508385557491987,
+      "loss": 0.7035,
+      "step": 8855
+    },
+    {
+      "epoch": 0.41961620469083155,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012506943883513043,
+      "loss": 1.079,
+      "step": 8856
+    },
+    {
+      "epoch": 0.41966358682776594,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012505502153932753,
+      "loss": 1.1291,
+      "step": 8857
+    },
+    {
+      "epoch": 0.4197109689647003,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012504060368783096,
+      "loss": 1.1054,
+      "step": 8858
+    },
+    {
+      "epoch": 0.4197583511016347,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012502618528096045,
+      "loss": 1.0571,
+      "step": 8859
+    },
+    {
+      "epoch": 0.41980573323856907,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00012501176631903583,
+      "loss": 0.9061,
+      "step": 8860
+    },
+    {
+      "epoch": 0.41985311537550346,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012499734680237684,
+      "loss": 1.1701,
+      "step": 8861
+    },
+    {
+      "epoch": 0.4199004975124378,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012498292673130334,
+      "loss": 0.6589,
+      "step": 8862
+    },
+    {
+      "epoch": 0.4199478796493722,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012496850610613512,
+      "loss": 1.3303,
+      "step": 8863
+    },
+    {
+      "epoch": 0.4199952617863066,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012495408492719205,
+      "loss": 0.1845,
+      "step": 8864
+    },
+    {
+      "epoch": 0.4200426439232409,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00012493966319479397,
+      "loss": 0.0253,
+      "step": 8865
+    },
+    {
+      "epoch": 0.4200900260601753,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001249252409092607,
+      "loss": 1.3075,
+      "step": 8866
+    },
+    {
+      "epoch": 0.4201374081971097,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012491081807091214,
+      "loss": 0.6048,
+      "step": 8867
+    },
+    {
+      "epoch": 0.42018479033404404,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001248963946800682,
+      "loss": 0.9905,
+      "step": 8868
+    },
+    {
+      "epoch": 0.42023217247097844,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001248819707370487,
+      "loss": 1.0346,
+      "step": 8869
+    },
+    {
+      "epoch": 0.42027955460791283,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00012486754624217361,
+      "loss": 0.0315,
+      "step": 8870
+    },
+    {
+      "epoch": 0.42032693674484717,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001248531211957628,
+      "loss": 0.9402,
+      "step": 8871
+    },
+    {
+      "epoch": 0.42037431888178156,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012483869559813627,
+      "loss": 0.748,
+      "step": 8872
+    },
+    {
+      "epoch": 0.42042170101871595,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012482426944961396,
+      "loss": 1.1636,
+      "step": 8873
+    },
+    {
+      "epoch": 0.42046908315565035,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012480984275051575,
+      "loss": 0.0295,
+      "step": 8874
+    },
+    {
+      "epoch": 0.4205164652925847,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00012479541550116166,
+      "loss": 0.026,
+      "step": 8875
+    },
+    {
+      "epoch": 0.4205638474295191,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012478098770187166,
+      "loss": 0.3643,
+      "step": 8876
+    },
+    {
+      "epoch": 0.42061122956645347,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012476655935296575,
+      "loss": 1.3416,
+      "step": 8877
+    },
+    {
+      "epoch": 0.4206586117033878,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012475213045476394,
+      "loss": 1.367,
+      "step": 8878
+    },
+    {
+      "epoch": 0.4207059938403222,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00012473770100758624,
+      "loss": 0.0348,
+      "step": 8879
+    },
+    {
+      "epoch": 0.4207533759772566,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001247232710117527,
+      "loss": 0.4657,
+      "step": 8880
+    },
+    {
+      "epoch": 0.42080075811419093,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012470884046758332,
+      "loss": 0.5723,
+      "step": 8881
+    },
+    {
+      "epoch": 0.4208481402511253,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012469440937539817,
+      "loss": 0.9915,
+      "step": 8882
+    },
+    {
+      "epoch": 0.4208955223880597,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012467997773551735,
+      "loss": 1.0745,
+      "step": 8883
+    },
+    {
+      "epoch": 0.42094290452499405,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012466554554826088,
+      "loss": 1.0999,
+      "step": 8884
+    },
+    {
+      "epoch": 0.42099028666192845,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001246511128139489,
+      "loss": 1.0777,
+      "step": 8885
+    },
+    {
+      "epoch": 0.42103766879886284,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001246366795329015,
+      "loss": 1.0084,
+      "step": 8886
+    },
+    {
+      "epoch": 0.4210850509357972,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012462224570543876,
+      "loss": 0.646,
+      "step": 8887
+    },
+    {
+      "epoch": 0.42113243307273157,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00012460781133188088,
+      "loss": 0.1525,
+      "step": 8888
+    },
+    {
+      "epoch": 0.42117981520966596,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012459337641254796,
+      "loss": 0.1222,
+      "step": 8889
+    },
+    {
+      "epoch": 0.42122719734660036,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001245789409477601,
+      "loss": 0.7078,
+      "step": 8890
+    },
+    {
+      "epoch": 0.4212745794835347,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00012456450493783752,
+      "loss": 0.5646,
+      "step": 8891
+    },
+    {
+      "epoch": 0.4213219616204691,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012455006838310042,
+      "loss": 0.8092,
+      "step": 8892
+    },
+    {
+      "epoch": 0.4213693437574035,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00012453563128386893,
+      "loss": 0.9258,
+      "step": 8893
+    },
+    {
+      "epoch": 0.4214167258943378,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012452119364046325,
+      "loss": 1.4296,
+      "step": 8894
+    },
+    {
+      "epoch": 0.4214641080312722,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00012450675545320366,
+      "loss": 1.3191,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4215114901682066,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00012449231672241032,
+      "loss": 0.0116,
+      "step": 8896
+    },
+    {
+      "epoch": 0.42155887230514094,
+      "grad_norm": 0.002197265625,
+      "learning_rate": 0.0001244778774484035,
+      "loss": 0.0001,
+      "step": 8897
+    },
+    {
+      "epoch": 0.42160625444207533,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001244634376315034,
+      "loss": 0.0157,
+      "step": 8898
+    },
+    {
+      "epoch": 0.4216536365790097,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001244489972720303,
+      "loss": 1.2617,
+      "step": 8899
+    },
+    {
+      "epoch": 0.42170101871594406,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012443455637030452,
+      "loss": 1.3783,
+      "step": 8900
+    },
+    {
+      "epoch": 0.42174840085287846,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012442011492664628,
+      "loss": 1.0486,
+      "step": 8901
+    },
+    {
+      "epoch": 0.42179578298981285,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001244056729413759,
+      "loss": 1.0188,
+      "step": 8902
+    },
+    {
+      "epoch": 0.42184316512674724,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001243912304148137,
+      "loss": 1.1449,
+      "step": 8903
+    },
+    {
+      "epoch": 0.4218905472636816,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012437678734728,
+      "loss": 0.5396,
+      "step": 8904
+    },
+    {
+      "epoch": 0.421937929400616,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00012436234373909512,
+      "loss": 0.8482,
+      "step": 8905
+    },
+    {
+      "epoch": 0.42198531153755037,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001243478995905794,
+      "loss": 0.0132,
+      "step": 8906
+    },
+    {
+      "epoch": 0.4220326936744847,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012433345490205322,
+      "loss": 0.5625,
+      "step": 8907
+    },
+    {
+      "epoch": 0.4220800758114191,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00012431900967383686,
+      "loss": 0.0742,
+      "step": 8908
+    },
+    {
+      "epoch": 0.4221274579483535,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012430456390625082,
+      "loss": 0.529,
+      "step": 8909
+    },
+    {
+      "epoch": 0.42217484008528783,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00012429011759961545,
+      "loss": 0.6306,
+      "step": 8910
+    },
+    {
+      "epoch": 0.4222222222222222,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012427567075425113,
+      "loss": 0.1279,
+      "step": 8911
+    },
+    {
+      "epoch": 0.4222696043591566,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001242612233704783,
+      "loss": 1.2095,
+      "step": 8912
+    },
+    {
+      "epoch": 0.42231698649609095,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00012424677544861738,
+      "loss": 0.959,
+      "step": 8913
+    },
+    {
+      "epoch": 0.42236436863302534,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00012423232698898878,
+      "loss": 0.9641,
+      "step": 8914
+    },
+    {
+      "epoch": 0.42241175076995974,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.000124217877991913,
+      "loss": 0.016,
+      "step": 8915
+    },
+    {
+      "epoch": 0.4224591329068941,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012420342845771048,
+      "loss": 0.5453,
+      "step": 8916
+    },
+    {
+      "epoch": 0.42250651504382847,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012418897838670166,
+      "loss": 0.8954,
+      "step": 8917
+    },
+    {
+      "epoch": 0.42255389718076286,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012417452777920712,
+      "loss": 1.2048,
+      "step": 8918
+    },
+    {
+      "epoch": 0.42260127931769725,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001241600766355473,
+      "loss": 0.01,
+      "step": 8919
+    },
+    {
+      "epoch": 0.4226486614546316,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012414562495604268,
+      "loss": 1.2205,
+      "step": 8920
+    },
+    {
+      "epoch": 0.422696043591566,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00012413117274101386,
+      "loss": 1.1835,
+      "step": 8921
+    },
+    {
+      "epoch": 0.4227434257285004,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00012411671999078128,
+      "loss": 0.0057,
+      "step": 8922
+    },
+    {
+      "epoch": 0.4227908078654347,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001241022667056656,
+      "loss": 0.8422,
+      "step": 8923
+    },
+    {
+      "epoch": 0.4228381900023691,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012408781288598725,
+      "loss": 0.0834,
+      "step": 8924
+    },
+    {
+      "epoch": 0.4228855721393035,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001240733585320669,
+      "loss": 0.8948,
+      "step": 8925
+    },
+    {
+      "epoch": 0.42293295427623784,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001240589036442251,
+      "loss": 1.1781,
+      "step": 8926
+    },
+    {
+      "epoch": 0.42298033641317223,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012404444822278242,
+      "loss": 0.6403,
+      "step": 8927
+    },
+    {
+      "epoch": 0.4230277185501066,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012402999226805955,
+      "loss": 0.6801,
+      "step": 8928
+    },
+    {
+      "epoch": 0.42307510068704096,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012401553578037698,
+      "loss": 0.0082,
+      "step": 8929
+    },
+    {
+      "epoch": 0.42312248282397535,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012400107876005544,
+      "loss": 0.8599,
+      "step": 8930
+    },
+    {
+      "epoch": 0.42316986496090975,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00012398662120741553,
+      "loss": 0.2746,
+      "step": 8931
+    },
+    {
+      "epoch": 0.42321724709784414,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001239721631227779,
+      "loss": 1.0153,
+      "step": 8932
+    },
+    {
+      "epoch": 0.4232646292347785,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00012395770450646324,
+      "loss": 0.7281,
+      "step": 8933
+    },
+    {
+      "epoch": 0.42331201137171287,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012394324535879223,
+      "loss": 0.9411,
+      "step": 8934
+    },
+    {
+      "epoch": 0.42335939350864726,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00012392878568008554,
+      "loss": 0.0147,
+      "step": 8935
+    },
+    {
+      "epoch": 0.4234067756455816,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012391432547066383,
+      "loss": 1.3564,
+      "step": 8936
+    },
+    {
+      "epoch": 0.423454157782516,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001238998647308479,
+      "loss": 1.171,
+      "step": 8937
+    },
+    {
+      "epoch": 0.4235015399194504,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001238854034609584,
+      "loss": 1.0655,
+      "step": 8938
+    },
+    {
+      "epoch": 0.4235489220563847,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001238709416613161,
+      "loss": 0.6354,
+      "step": 8939
+    },
+    {
+      "epoch": 0.4235963041933191,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00012385647933224172,
+      "loss": 1.0352,
+      "step": 8940
+    },
+    {
+      "epoch": 0.4236436863302535,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012384201647405607,
+      "loss": 0.0452,
+      "step": 8941
+    },
+    {
+      "epoch": 0.42369106846718785,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00012382755308707987,
+      "loss": 0.2959,
+      "step": 8942
+    },
+    {
+      "epoch": 0.42373845060412224,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012381308917163393,
+      "loss": 0.7554,
+      "step": 8943
+    },
+    {
+      "epoch": 0.42378583274105663,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012379862472803904,
+      "loss": 0.5578,
+      "step": 8944
+    },
+    {
+      "epoch": 0.42383321487799097,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000123784159756616,
+      "loss": 0.9,
+      "step": 8945
+    },
+    {
+      "epoch": 0.42388059701492536,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012376969425768563,
+      "loss": 1.1052,
+      "step": 8946
+    },
+    {
+      "epoch": 0.42392797915185976,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012375522823156876,
+      "loss": 1.1587,
+      "step": 8947
+    },
+    {
+      "epoch": 0.42397536128879415,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001237407616785862,
+      "loss": 1.0681,
+      "step": 8948
+    },
+    {
+      "epoch": 0.4240227434257285,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00012372629459905888,
+      "loss": 0.0124,
+      "step": 8949
+    },
+    {
+      "epoch": 0.4240701255626629,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001237118269933076,
+      "loss": 0.7137,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4241175076995973,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012369735886165326,
+      "loss": 0.7155,
+      "step": 8951
+    },
+    {
+      "epoch": 0.4241648898365316,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00012368289020441676,
+      "loss": 0.0344,
+      "step": 8952
+    },
+    {
+      "epoch": 0.424212271973466,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012366842102191894,
+      "loss": 0.7535,
+      "step": 8953
+    },
+    {
+      "epoch": 0.4242596541104004,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001236539513144808,
+      "loss": 0.624,
+      "step": 8954
+    },
+    {
+      "epoch": 0.42430703624733473,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00012363948108242315,
+      "loss": 0.4886,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4243544183842691,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012362501032606702,
+      "loss": 0.7854,
+      "step": 8956
+    },
+    {
+      "epoch": 0.4244018005212035,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00012361053904573333,
+      "loss": 1.3908,
+      "step": 8957
+    },
+    {
+      "epoch": 0.42444918265813786,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012359606724174303,
+      "loss": 0.9789,
+      "step": 8958
+    },
+    {
+      "epoch": 0.42449656479507225,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001235815949144171,
+      "loss": 0.03,
+      "step": 8959
+    },
+    {
+      "epoch": 0.42454394693200664,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00012356712206407653,
+      "loss": 0.806,
+      "step": 8960
+    },
+    {
+      "epoch": 0.42459132906894104,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001235526486910423,
+      "loss": 0.1725,
+      "step": 8961
+    },
+    {
+      "epoch": 0.4246387112058754,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00012353817479563535,
+      "loss": 0.9963,
+      "step": 8962
+    },
+    {
+      "epoch": 0.42468609334280977,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0001235237003781768,
+      "loss": 0.0063,
+      "step": 8963
+    },
+    {
+      "epoch": 0.42473347547974416,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00012350922543898757,
+      "loss": 0.0165,
+      "step": 8964
+    },
+    {
+      "epoch": 0.4247808576166785,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012349474997838883,
+      "loss": 1.1218,
+      "step": 8965
+    },
+    {
+      "epoch": 0.4248282397536129,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012348027399670155,
+      "loss": 1.1432,
+      "step": 8966
+    },
+    {
+      "epoch": 0.4248756218905473,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00012346579749424679,
+      "loss": 0.0073,
+      "step": 8967
+    },
+    {
+      "epoch": 0.4249230040274816,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012345132047134565,
+      "loss": 1.2107,
+      "step": 8968
+    },
+    {
+      "epoch": 0.424970386164416,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00012343684292831917,
+      "loss": 0.0673,
+      "step": 8969
+    },
+    {
+      "epoch": 0.4250177683013504,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001234223648654885,
+      "loss": 1.0352,
+      "step": 8970
+    },
+    {
+      "epoch": 0.42506515043828474,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012340788628317472,
+      "loss": 0.9669,
+      "step": 8971
+    },
+    {
+      "epoch": 0.42511253257521914,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000123393407181699,
+      "loss": 0.526,
+      "step": 8972
+    },
+    {
+      "epoch": 0.42515991471215353,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00012337892756138235,
+      "loss": 0.1482,
+      "step": 8973
+    },
+    {
+      "epoch": 0.42520729684908787,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00012336444742254603,
+      "loss": 0.4275,
+      "step": 8974
+    },
+    {
+      "epoch": 0.42525467898602226,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00012334996676551115,
+      "loss": 0.082,
+      "step": 8975
+    },
+    {
+      "epoch": 0.42530206112295665,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001233354855905989,
+      "loss": 0.0929,
+      "step": 8976
+    },
+    {
+      "epoch": 0.42534944325989105,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001233210038981304,
+      "loss": 0.9906,
+      "step": 8977
+    },
+    {
+      "epoch": 0.4253968253968254,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012330652168842686,
+      "loss": 0.8739,
+      "step": 8978
+    },
+    {
+      "epoch": 0.4254442075337598,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00012329203896180953,
+      "loss": 0.9353,
+      "step": 8979
+    },
+    {
+      "epoch": 0.42549158967069417,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001232775557185996,
+      "loss": 0.1222,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4255389718076285,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00012326307195911822,
+      "loss": 1.0315,
+      "step": 8981
+    },
+    {
+      "epoch": 0.4255863539445629,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001232485876836867,
+      "loss": 0.897,
+      "step": 8982
+    },
+    {
+      "epoch": 0.4256337360814973,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001232341028926263,
+      "loss": 1.1748,
+      "step": 8983
+    },
+    {
+      "epoch": 0.42568111821843163,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012321961758625824,
+      "loss": 0.0113,
+      "step": 8984
+    },
+    {
+      "epoch": 0.425728500355366,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012320513176490377,
+      "loss": 0.4843,
+      "step": 8985
+    },
+    {
+      "epoch": 0.4257758824923004,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001231906454288842,
+      "loss": 1.2123,
+      "step": 8986
+    },
+    {
+      "epoch": 0.42582326462923475,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00012317615857852083,
+      "loss": 0.4092,
+      "step": 8987
+    },
+    {
+      "epoch": 0.42587064676616915,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00012316167121413497,
+      "loss": 1.2504,
+      "step": 8988
+    },
+    {
+      "epoch": 0.42591802890310354,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012314718333604786,
+      "loss": 0.5383,
+      "step": 8989
+    },
+    {
+      "epoch": 0.42596541104003793,
+      "grad_norm": 0.0031585693359375,
+      "learning_rate": 0.00012313269494458088,
+      "loss": 0.0002,
+      "step": 8990
+    },
+    {
+      "epoch": 0.42601279317697227,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00012311820604005543,
+      "loss": 0.0202,
+      "step": 8991
+    },
+    {
+      "epoch": 0.42606017531390666,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012310371662279277,
+      "loss": 1.3952,
+      "step": 8992
+    },
+    {
+      "epoch": 0.42610755745084106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012308922669311426,
+      "loss": 0.7924,
+      "step": 8993
+    },
+    {
+      "epoch": 0.4261549395877754,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001230747362513413,
+      "loss": 1.1176,
+      "step": 8994
+    },
+    {
+      "epoch": 0.4262023217247098,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00012306024529779526,
+      "loss": 1.4522,
+      "step": 8995
+    },
+    {
+      "epoch": 0.4262497038616442,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012304575383279755,
+      "loss": 0.1515,
+      "step": 8996
+    },
+    {
+      "epoch": 0.4262970859985785,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.00012303126185666958,
+      "loss": 0.0047,
+      "step": 8997
+    },
+    {
+      "epoch": 0.4263444681355129,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00012301676936973273,
+      "loss": 0.2999,
+      "step": 8998
+    },
+    {
+      "epoch": 0.4263918502724473,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001230022763723085,
+      "loss": 1.2043,
+      "step": 8999
+    },
+    {
+      "epoch": 0.42643923240938164,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00012298778286471825,
+      "loss": 1.2351,
+      "step": 9000
+    },
+    {
+      "epoch": 0.42648661454631603,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012297328884728346,
+      "loss": 0.1498,
+      "step": 9001
+    },
+    {
+      "epoch": 0.4265339966832504,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00012295879432032558,
+      "loss": 0.043,
+      "step": 9002
+    },
+    {
+      "epoch": 0.42658137882018476,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0001229442992841661,
+      "loss": 1.1026,
+      "step": 9003
+    },
+    {
+      "epoch": 0.42662876095711916,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00012292980373912652,
+      "loss": 0.0639,
+      "step": 9004
+    },
+    {
+      "epoch": 0.42667614309405355,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001229153076855283,
+      "loss": 1.012,
+      "step": 9005
+    },
+    {
+      "epoch": 0.42672352523098794,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012290081112369298,
+      "loss": 1.2843,
+      "step": 9006
+    },
+    {
+      "epoch": 0.4267709073679223,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00012288631405394206,
+      "loss": 0.7494,
+      "step": 9007
+    },
+    {
+      "epoch": 0.4268182895048567,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012287181647659706,
+      "loss": 0.8909,
+      "step": 9008
+    },
+    {
+      "epoch": 0.42686567164179107,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012285731839197954,
+      "loss": 1.3663,
+      "step": 9009
+    },
+    {
+      "epoch": 0.4269130537787254,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012284281980041103,
+      "loss": 1.3305,
+      "step": 9010
+    },
+    {
+      "epoch": 0.4269604359156598,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012282832070221314,
+      "loss": 1.2073,
+      "step": 9011
+    },
+    {
+      "epoch": 0.4270078180525942,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001228138210977074,
+      "loss": 0.1763,
+      "step": 9012
+    },
+    {
+      "epoch": 0.4270552001895285,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001227993209872154,
+      "loss": 1.3054,
+      "step": 9013
+    },
+    {
+      "epoch": 0.4271025823264629,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012278482037105873,
+      "loss": 1.2087,
+      "step": 9014
+    },
+    {
+      "epoch": 0.4271499644633973,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012277031924955905,
+      "loss": 0.0055,
+      "step": 9015
+    },
+    {
+      "epoch": 0.42719734660033165,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001227558176230379,
+      "loss": 0.9413,
+      "step": 9016
+    },
+    {
+      "epoch": 0.42724472873726604,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00012274131549181697,
+      "loss": 0.0318,
+      "step": 9017
+    },
+    {
+      "epoch": 0.42729211087420044,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001227268128562179,
+      "loss": 0.8257,
+      "step": 9018
+    },
+    {
+      "epoch": 0.4273394930111348,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012271230971656235,
+      "loss": 1.3141,
+      "step": 9019
+    },
+    {
+      "epoch": 0.42738687514806917,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001226978060731719,
+      "loss": 1.0274,
+      "step": 9020
+    },
+    {
+      "epoch": 0.42743425728500356,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012268330192636832,
+      "loss": 0.0155,
+      "step": 9021
+    },
+    {
+      "epoch": 0.42748163942193795,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012266879727647326,
+      "loss": 0.8564,
+      "step": 9022
+    },
+    {
+      "epoch": 0.4275290215588723,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012265429212380847,
+      "loss": 1.0449,
+      "step": 9023
+    },
+    {
+      "epoch": 0.4275764036958067,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012263978646869555,
+      "loss": 1.0867,
+      "step": 9024
+    },
+    {
+      "epoch": 0.4276237858327411,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00012262528031145627,
+      "loss": 1.1932,
+      "step": 9025
+    },
+    {
+      "epoch": 0.4276711679696754,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012261077365241242,
+      "loss": 0.8738,
+      "step": 9026
+    },
+    {
+      "epoch": 0.4277185501066098,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012259626649188568,
+      "loss": 0.9239,
+      "step": 9027
+    },
+    {
+      "epoch": 0.4277659322435442,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001225817588301978,
+      "loss": 0.7741,
+      "step": 9028
+    },
+    {
+      "epoch": 0.42781331438047854,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001225672506676706,
+      "loss": 1.0425,
+      "step": 9029
+    },
+    {
+      "epoch": 0.42786069651741293,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001225527420046258,
+      "loss": 0.6123,
+      "step": 9030
+    },
+    {
+      "epoch": 0.4279080786543473,
+      "grad_norm": 0.021240234375,
+      "learning_rate": 0.0001225382328413852,
+      "loss": 0.0008,
+      "step": 9031
+    },
+    {
+      "epoch": 0.42795546079128166,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012252372317827056,
+      "loss": 0.8573,
+      "step": 9032
+    },
+    {
+      "epoch": 0.42800284292821605,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012250921301560377,
+      "loss": 1.1215,
+      "step": 9033
+    },
+    {
+      "epoch": 0.42805022506515045,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012249470235370664,
+      "loss": 0.0241,
+      "step": 9034
+    },
+    {
+      "epoch": 0.42809760720208484,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00012248019119290093,
+      "loss": 1.162,
+      "step": 9035
+    },
+    {
+      "epoch": 0.4281449893390192,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012246567953350852,
+      "loss": 1.4999,
+      "step": 9036
+    },
+    {
+      "epoch": 0.42819237147595357,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001224511673758513,
+      "loss": 0.0294,
+      "step": 9037
+    },
+    {
+      "epoch": 0.42823975361288796,
+      "grad_norm": 0.0301513671875,
+      "learning_rate": 0.00012243665472025112,
+      "loss": 0.0009,
+      "step": 9038
+    },
+    {
+      "epoch": 0.4282871357498223,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001224221415670298,
+      "loss": 1.0573,
+      "step": 9039
+    },
+    {
+      "epoch": 0.4283345178867567,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012240762791650923,
+      "loss": 0.7919,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4283819000236911,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001223931137690114,
+      "loss": 0.0177,
+      "step": 9041
+    },
+    {
+      "epoch": 0.4284292821606254,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00012237859912485815,
+      "loss": 1.1608,
+      "step": 9042
+    },
+    {
+      "epoch": 0.4284766642975598,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012236408398437135,
+      "loss": 0.5714,
+      "step": 9043
+    },
+    {
+      "epoch": 0.4285240464344942,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012234956834787303,
+      "loss": 0.0568,
+      "step": 9044
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00012233505221568512,
+      "loss": 1.2182,
+      "step": 9045
+    },
+    {
+      "epoch": 0.42861881070836294,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012232053558812952,
+      "loss": 0.1509,
+      "step": 9046
+    },
+    {
+      "epoch": 0.42866619284529733,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001223060184655282,
+      "loss": 1.013,
+      "step": 9047
+    },
+    {
+      "epoch": 0.42871357498223167,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00012229150084820315,
+      "loss": 1.0359,
+      "step": 9048
+    },
+    {
+      "epoch": 0.42876095711916606,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001222769827364764,
+      "loss": 1.0239,
+      "step": 9049
+    },
+    {
+      "epoch": 0.42880833925610046,
+      "grad_norm": 0.00162506103515625,
+      "learning_rate": 0.00012226246413066984,
+      "loss": 0.0001,
+      "step": 9050
+    },
+    {
+      "epoch": 0.42885572139303485,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012224794503110556,
+      "loss": 1.1826,
+      "step": 9051
+    },
+    {
+      "epoch": 0.4289031035299692,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00012223342543810556,
+      "loss": 0.7507,
+      "step": 9052
+    },
+    {
+      "epoch": 0.4289504856669036,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012221890535199186,
+      "loss": 0.4868,
+      "step": 9053
+    },
+    {
+      "epoch": 0.428997867803838,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012220438477308655,
+      "loss": 0.8132,
+      "step": 9054
+    },
+    {
+      "epoch": 0.4290452499407723,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012218986370171158,
+      "loss": 0.6508,
+      "step": 9055
+    },
+    {
+      "epoch": 0.4290926320777067,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00012217534213818907,
+      "loss": 0.0103,
+      "step": 9056
+    },
+    {
+      "epoch": 0.4291400142146411,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00012216082008284114,
+      "loss": 0.0368,
+      "step": 9057
+    },
+    {
+      "epoch": 0.42918739635157543,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00012214629753598978,
+      "loss": 0.1443,
+      "step": 9058
+    },
+    {
+      "epoch": 0.4292347784885098,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012213177449795713,
+      "loss": 0.7987,
+      "step": 9059
+    },
+    {
+      "epoch": 0.4292821606254442,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001221172509690653,
+      "loss": 0.7702,
+      "step": 9060
+    },
+    {
+      "epoch": 0.42932954276237856,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012210272694963644,
+      "loss": 0.7252,
+      "step": 9061
+    },
+    {
+      "epoch": 0.42937692489931295,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012208820243999263,
+      "loss": 0.817,
+      "step": 9062
+    },
+    {
+      "epoch": 0.42942430703624734,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00012207367744045597,
+      "loss": 0.2632,
+      "step": 9063
+    },
+    {
+      "epoch": 0.42947168917318174,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001220591519513487,
+      "loss": 1.2632,
+      "step": 9064
+    },
+    {
+      "epoch": 0.4295190713101161,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00012204462597299291,
+      "loss": 1.1329,
+      "step": 9065
+    },
+    {
+      "epoch": 0.42956645344705047,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012203009950571078,
+      "loss": 1.0932,
+      "step": 9066
+    },
+    {
+      "epoch": 0.42961383558398486,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012201557254982454,
+      "loss": 0.6318,
+      "step": 9067
+    },
+    {
+      "epoch": 0.4296612177209192,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00012200104510565635,
+      "loss": 0.05,
+      "step": 9068
+    },
+    {
+      "epoch": 0.4297085998578536,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001219865171735284,
+      "loss": 1.1741,
+      "step": 9069
+    },
+    {
+      "epoch": 0.429755981994788,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012197198875376295,
+      "loss": 1.007,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4298033641317223,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00012195745984668216,
+      "loss": 0.9968,
+      "step": 9071
+    },
+    {
+      "epoch": 0.4298507462686567,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012194293045260832,
+      "loss": 1.262,
+      "step": 9072
+    },
+    {
+      "epoch": 0.4298981284055911,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012192840057186363,
+      "loss": 1.5263,
+      "step": 9073
+    },
+    {
+      "epoch": 0.42994551054252544,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00012191387020477038,
+      "loss": 0.0267,
+      "step": 9074
+    },
+    {
+      "epoch": 0.42999289267945984,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012189933935165086,
+      "loss": 0.0294,
+      "step": 9075
+    },
+    {
+      "epoch": 0.43004027481639423,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012188480801282729,
+      "loss": 0.0222,
+      "step": 9076
+    },
+    {
+      "epoch": 0.43008765695332857,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012187027618862199,
+      "loss": 1.0799,
+      "step": 9077
+    },
+    {
+      "epoch": 0.43013503909026296,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012185574387935727,
+      "loss": 0.014,
+      "step": 9078
+    },
+    {
+      "epoch": 0.43018242122719735,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012184121108535543,
+      "loss": 0.6301,
+      "step": 9079
+    },
+    {
+      "epoch": 0.43022980336413175,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001218266778069388,
+      "loss": 0.2887,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4302771855010661,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00012181214404442967,
+      "loss": 0.0134,
+      "step": 9081
+    },
+    {
+      "epoch": 0.4303245676380005,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012179760979815045,
+      "loss": 0.9382,
+      "step": 9082
+    },
+    {
+      "epoch": 0.43037194977493487,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012178307506842345,
+      "loss": 0.7161,
+      "step": 9083
+    },
+    {
+      "epoch": 0.4304193319118692,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00012176853985557104,
+      "loss": 0.0159,
+      "step": 9084
+    },
+    {
+      "epoch": 0.4304667140488036,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00012175400415991563,
+      "loss": 0.1446,
+      "step": 9085
+    },
+    {
+      "epoch": 0.430514096185738,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00012173946798177956,
+      "loss": 0.7958,
+      "step": 9086
+    },
+    {
+      "epoch": 0.43056147832267233,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00012172493132148525,
+      "loss": 0.0021,
+      "step": 9087
+    },
+    {
+      "epoch": 0.4306088604596067,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001217103941793551,
+      "loss": 0.4472,
+      "step": 9088
+    },
+    {
+      "epoch": 0.4306562425965411,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012169585655571154,
+      "loss": 0.5643,
+      "step": 9089
+    },
+    {
+      "epoch": 0.43070362473347545,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012168131845087699,
+      "loss": 0.5358,
+      "step": 9090
+    },
+    {
+      "epoch": 0.43075100687040985,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012166677986517387,
+      "loss": 0.8892,
+      "step": 9091
+    },
+    {
+      "epoch": 0.43079838900734424,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00012165224079892467,
+      "loss": 1.2856,
+      "step": 9092
+    },
+    {
+      "epoch": 0.43084577114427863,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012163770125245185,
+      "loss": 1.1084,
+      "step": 9093
+    },
+    {
+      "epoch": 0.43089315328121297,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00012162316122607781,
+      "loss": 0.9945,
+      "step": 9094
+    },
+    {
+      "epoch": 0.43094053541814736,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00012160862072012515,
+      "loss": 1.2901,
+      "step": 9095
+    },
+    {
+      "epoch": 0.43098791755508176,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 0.00012159407973491625,
+      "loss": 0.0036,
+      "step": 9096
+    },
+    {
+      "epoch": 0.4310352996920161,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012157953827077367,
+      "loss": 1.0957,
+      "step": 9097
+    },
+    {
+      "epoch": 0.4310826818289505,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012156499632801994,
+      "loss": 1.5313,
+      "step": 9098
+    },
+    {
+      "epoch": 0.4311300639658849,
+      "grad_norm": 0.002532958984375,
+      "learning_rate": 0.00012155045390697754,
+      "loss": 0.0002,
+      "step": 9099
+    },
+    {
+      "epoch": 0.4311774461028192,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00012153591100796903,
+      "loss": 1.1189,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4312248282397536,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012152136763131697,
+      "loss": 0.5701,
+      "step": 9101
+    },
+    {
+      "epoch": 0.431272210376688,
+      "grad_norm": 0.0283203125,
+      "learning_rate": 0.00012150682377734388,
+      "loss": 0.0027,
+      "step": 9102
+    },
+    {
+      "epoch": 0.43131959251362234,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012149227944637235,
+      "loss": 0.9697,
+      "step": 9103
+    },
+    {
+      "epoch": 0.43136697465055673,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012147773463872496,
+      "loss": 0.0147,
+      "step": 9104
+    },
+    {
+      "epoch": 0.4314143567874911,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001214631893547243,
+      "loss": 1.3039,
+      "step": 9105
+    },
+    {
+      "epoch": 0.43146173892442546,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012144864359469294,
+      "loss": 0.8633,
+      "step": 9106
+    },
+    {
+      "epoch": 0.43150912106135986,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012143409735895353,
+      "loss": 0.8172,
+      "step": 9107
+    },
+    {
+      "epoch": 0.43155650319829425,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00012141955064782866,
+      "loss": 0.0044,
+      "step": 9108
+    },
+    {
+      "epoch": 0.43160388533522864,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012140500346164099,
+      "loss": 1.3199,
+      "step": 9109
+    },
+    {
+      "epoch": 0.431651267472163,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00012139045580071313,
+      "loss": 1.3542,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4316986496090974,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012137590766536775,
+      "loss": 0.7236,
+      "step": 9111
+    },
+    {
+      "epoch": 0.43174603174603177,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.00012136135905592749,
+      "loss": 0.0086,
+      "step": 9112
+    },
+    {
+      "epoch": 0.4317934138829661,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012134680997271504,
+      "loss": 1.334,
+      "step": 9113
+    },
+    {
+      "epoch": 0.4318407960199005,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001213322604160531,
+      "loss": 0.8218,
+      "step": 9114
+    },
+    {
+      "epoch": 0.4318881781568349,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00012131771038626434,
+      "loss": 0.8312,
+      "step": 9115
+    },
+    {
+      "epoch": 0.4319355602937692,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012130315988367145,
+      "loss": 0.7438,
+      "step": 9116
+    },
+    {
+      "epoch": 0.4319829424307036,
+      "grad_norm": 0.025634765625,
+      "learning_rate": 0.00012128860890859722,
+      "loss": 0.0011,
+      "step": 9117
+    },
+    {
+      "epoch": 0.432030324567638,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00012127405746136427,
+      "loss": 1.1243,
+      "step": 9118
+    },
+    {
+      "epoch": 0.43207770670457235,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00012125950554229539,
+      "loss": 1.2505,
+      "step": 9119
+    },
+    {
+      "epoch": 0.43212508884150674,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012124495315171334,
+      "loss": 1.7117,
+      "step": 9120
+    },
+    {
+      "epoch": 0.43217247097844114,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00012123040028994086,
+      "loss": 0.0113,
+      "step": 9121
+    },
+    {
+      "epoch": 0.43221985311537553,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012121584695730068,
+      "loss": 1.0387,
+      "step": 9122
+    },
+    {
+      "epoch": 0.43226723525230987,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00012120129315411564,
+      "loss": 0.7958,
+      "step": 9123
+    },
+    {
+      "epoch": 0.43231461738924426,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00012118673888070848,
+      "loss": 0.0095,
+      "step": 9124
+    },
+    {
+      "epoch": 0.43236199952617865,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00012117218413740206,
+      "loss": 0.1888,
+      "step": 9125
+    },
+    {
+      "epoch": 0.432409381663113,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00012115762892451915,
+      "loss": 0.0096,
+      "step": 9126
+    },
+    {
+      "epoch": 0.4324567638000474,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012114307324238253,
+      "loss": 1.0692,
+      "step": 9127
+    },
+    {
+      "epoch": 0.4325041459369818,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012112851709131505,
+      "loss": 0.5198,
+      "step": 9128
+    },
+    {
+      "epoch": 0.4325515280739161,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001211139604716396,
+      "loss": 0.9162,
+      "step": 9129
+    },
+    {
+      "epoch": 0.4325989102108505,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.00012109940338367897,
+      "loss": 0.004,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4326462923477849,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00012108484582775608,
+      "loss": 1.0419,
+      "step": 9131
+    },
+    {
+      "epoch": 0.43269367448471924,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00012107028780419374,
+      "loss": 0.6108,
+      "step": 9132
+    },
+    {
+      "epoch": 0.43274105662165363,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001210557293133149,
+      "loss": 0.0552,
+      "step": 9133
+    },
+    {
+      "epoch": 0.432788438758588,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012104117035544238,
+      "loss": 1.0201,
+      "step": 9134
+    },
+    {
+      "epoch": 0.43283582089552236,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00012102661093089911,
+      "loss": 0.4299,
+      "step": 9135
+    },
+    {
+      "epoch": 0.43288320303245675,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.00012101205104000801,
+      "loss": 0.0036,
+      "step": 9136
+    },
+    {
+      "epoch": 0.43293058516939115,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00012099749068309201,
+      "loss": 0.5321,
+      "step": 9137
+    },
+    {
+      "epoch": 0.43297796730632554,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00012098292986047402,
+      "loss": 0.017,
+      "step": 9138
+    },
+    {
+      "epoch": 0.4330253494432599,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000120968368572477,
+      "loss": 0.0328,
+      "step": 9139
+    },
+    {
+      "epoch": 0.43307273158019427,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001209538068194239,
+      "loss": 0.8218,
+      "step": 9140
+    },
+    {
+      "epoch": 0.43312011371712866,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001209392446016377,
+      "loss": 1.2307,
+      "step": 9141
+    },
+    {
+      "epoch": 0.433167495854063,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012092468191944133,
+      "loss": 0.5895,
+      "step": 9142
+    },
+    {
+      "epoch": 0.4332148779909974,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0001209101187731578,
+      "loss": 0.6921,
+      "step": 9143
+    },
+    {
+      "epoch": 0.4332622601279318,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012089555516311016,
+      "loss": 1.3407,
+      "step": 9144
+    },
+    {
+      "epoch": 0.4333096422648661,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001208809910896213,
+      "loss": 0.6376,
+      "step": 9145
+    },
+    {
+      "epoch": 0.4333570244018005,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00012086642655301432,
+      "loss": 0.0064,
+      "step": 9146
+    },
+    {
+      "epoch": 0.4334044065387349,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00012085186155361224,
+      "loss": 0.0053,
+      "step": 9147
+    },
+    {
+      "epoch": 0.43345178867566925,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012083729609173806,
+      "loss": 1.034,
+      "step": 9148
+    },
+    {
+      "epoch": 0.43349917081260364,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012082273016771488,
+      "loss": 0.196,
+      "step": 9149
+    },
+    {
+      "epoch": 0.43354655294953803,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001208081637818657,
+      "loss": 1.3181,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4335939350864724,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001207935969345136,
+      "loss": 1.2652,
+      "step": 9151
+    },
+    {
+      "epoch": 0.43364131722340676,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00012077902962598171,
+      "loss": 0.0348,
+      "step": 9152
+    },
+    {
+      "epoch": 0.43368869936034116,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00012076446185659302,
+      "loss": 0.0056,
+      "step": 9153
+    },
+    {
+      "epoch": 0.43373608149727555,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001207498936266707,
+      "loss": 0.772,
+      "step": 9154
+    },
+    {
+      "epoch": 0.4337834636342099,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012073532493653786,
+      "loss": 1.2163,
+      "step": 9155
+    },
+    {
+      "epoch": 0.4338308457711443,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012072075578651762,
+      "loss": 1.0997,
+      "step": 9156
+    },
+    {
+      "epoch": 0.4338782279080787,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012070618617693306,
+      "loss": 1.0526,
+      "step": 9157
+    },
+    {
+      "epoch": 0.433925610045013,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012069161610810734,
+      "loss": 0.8277,
+      "step": 9158
+    },
+    {
+      "epoch": 0.4339729921819474,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012067704558036362,
+      "loss": 1.0736,
+      "step": 9159
+    },
+    {
+      "epoch": 0.4340203743188818,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012066247459402507,
+      "loss": 0.7963,
+      "step": 9160
+    },
+    {
+      "epoch": 0.43406775645581613,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001206479031494148,
+      "loss": 1.3514,
+      "step": 9161
+    },
+    {
+      "epoch": 0.4341151385927505,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00012063333124685606,
+      "loss": 0.0038,
+      "step": 9162
+    },
+    {
+      "epoch": 0.4341625207296849,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00012061875888667203,
+      "loss": 0.009,
+      "step": 9163
+    },
+    {
+      "epoch": 0.43420990286661926,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012060418606918587,
+      "loss": 0.7726,
+      "step": 9164
+    },
+    {
+      "epoch": 0.43425728500355365,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012058961279472079,
+      "loss": 0.8138,
+      "step": 9165
+    },
+    {
+      "epoch": 0.43430466714048804,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00012057503906360004,
+      "loss": 0.5046,
+      "step": 9166
+    },
+    {
+      "epoch": 0.43435204927742244,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012056046487614687,
+      "loss": 1.3172,
+      "step": 9167
+    },
+    {
+      "epoch": 0.4343994314143568,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00012054589023268445,
+      "loss": 0.1465,
+      "step": 9168
+    },
+    {
+      "epoch": 0.43444681355129117,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00012053131513353608,
+      "loss": 0.811,
+      "step": 9169
+    },
+    {
+      "epoch": 0.43449419568822556,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012051673957902501,
+      "loss": 1.2989,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4345415778251599,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00012050216356947453,
+      "loss": 1.5614,
+      "step": 9171
+    },
+    {
+      "epoch": 0.4345889599620943,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012048758710520789,
+      "loss": 0.9579,
+      "step": 9172
+    },
+    {
+      "epoch": 0.4346363420990287,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012047301018654838,
+      "loss": 1.2662,
+      "step": 9173
+    },
+    {
+      "epoch": 0.434683724235963,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001204584328138193,
+      "loss": 1.0081,
+      "step": 9174
+    },
+    {
+      "epoch": 0.4347311063728974,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012044385498734398,
+      "loss": 0.835,
+      "step": 9175
+    },
+    {
+      "epoch": 0.4347784885098318,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012042927670744574,
+      "loss": 1.023,
+      "step": 9176
+    },
+    {
+      "epoch": 0.43482587064676614,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012041469797444788,
+      "loss": 0.7017,
+      "step": 9177
+    },
+    {
+      "epoch": 0.43487325278370054,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012040011878867379,
+      "loss": 0.9192,
+      "step": 9178
+    },
+    {
+      "epoch": 0.43492063492063493,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012038553915044679,
+      "loss": 0.8073,
+      "step": 9179
+    },
+    {
+      "epoch": 0.4349680170575693,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012037095906009026,
+      "loss": 0.0218,
+      "step": 9180
+    },
+    {
+      "epoch": 0.43501539919450366,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00012035637851792754,
+      "loss": 0.7896,
+      "step": 9181
+    },
+    {
+      "epoch": 0.43506278133143805,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012034179752428203,
+      "loss": 0.7577,
+      "step": 9182
+    },
+    {
+      "epoch": 0.43511016346837245,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012032721607947712,
+      "loss": 0.0411,
+      "step": 9183
+    },
+    {
+      "epoch": 0.4351575456053068,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012031263418383618,
+      "loss": 1.0111,
+      "step": 9184
+    },
+    {
+      "epoch": 0.4352049277422412,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00012029805183768268,
+      "loss": 1.3509,
+      "step": 9185
+    },
+    {
+      "epoch": 0.43525230987917557,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012028346904134003,
+      "loss": 1.2804,
+      "step": 9186
+    },
+    {
+      "epoch": 0.4352996920161099,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001202688857951316,
+      "loss": 0.8546,
+      "step": 9187
+    },
+    {
+      "epoch": 0.4353470741530443,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00012025430209938089,
+      "loss": 0.116,
+      "step": 9188
+    },
+    {
+      "epoch": 0.4353944562899787,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012023971795441133,
+      "loss": 0.5732,
+      "step": 9189
+    },
+    {
+      "epoch": 0.43544183842691303,
+      "grad_norm": 0.0306396484375,
+      "learning_rate": 0.0001202251333605464,
+      "loss": 0.0006,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4354892205638474,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012021054831810953,
+      "loss": 0.0279,
+      "step": 9191
+    },
+    {
+      "epoch": 0.4355366027007818,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00012019596282742424,
+      "loss": 0.0075,
+      "step": 9192
+    },
+    {
+      "epoch": 0.43558398483771615,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00012018137688881399,
+      "loss": 1.0802,
+      "step": 9193
+    },
+    {
+      "epoch": 0.43563136697465055,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012016679050260232,
+      "loss": 1.4969,
+      "step": 9194
+    },
+    {
+      "epoch": 0.43567874911158494,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00012015220366911273,
+      "loss": 0.711,
+      "step": 9195
+    },
+    {
+      "epoch": 0.43572613124851933,
+      "grad_norm": 0.0016021728515625,
+      "learning_rate": 0.0001201376163886687,
+      "loss": 0.0001,
+      "step": 9196
+    },
+    {
+      "epoch": 0.43577351338545367,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012012302866159377,
+      "loss": 0.0915,
+      "step": 9197
+    },
+    {
+      "epoch": 0.43582089552238806,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00012010844048821155,
+      "loss": 0.0159,
+      "step": 9198
+    },
+    {
+      "epoch": 0.43586827765932246,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001200938518688455,
+      "loss": 0.3029,
+      "step": 9199
+    },
+    {
+      "epoch": 0.4359156597962568,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00012007926280381921,
+      "loss": 0.9484,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4359630419331912,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012006467329345627,
+      "loss": 0.8423,
+      "step": 9201
+    },
+    {
+      "epoch": 0.4360104240701256,
+      "grad_norm": 0.0302734375,
+      "learning_rate": 0.00012005008333808028,
+      "loss": 0.0023,
+      "step": 9202
+    },
+    {
+      "epoch": 0.4360578062070599,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012003549293801477,
+      "loss": 0.8052,
+      "step": 9203
+    },
+    {
+      "epoch": 0.4361051883439943,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012002090209358336,
+      "loss": 0.1232,
+      "step": 9204
+    },
+    {
+      "epoch": 0.4361525704809287,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00012000631080510969,
+      "loss": 0.3767,
+      "step": 9205
+    },
+    {
+      "epoch": 0.43619995261786304,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011999171907291735,
+      "loss": 0.7324,
+      "step": 9206
+    },
+    {
+      "epoch": 0.43624733475479743,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011997712689732996,
+      "loss": 1.2838,
+      "step": 9207
+    },
+    {
+      "epoch": 0.4362947168917318,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011996253427867116,
+      "loss": 0.8453,
+      "step": 9208
+    },
+    {
+      "epoch": 0.4363420990286662,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011994794121726467,
+      "loss": 1.2646,
+      "step": 9209
+    },
+    {
+      "epoch": 0.43638948116560056,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011993334771343405,
+      "loss": 0.8129,
+      "step": 9210
+    },
+    {
+      "epoch": 0.43643686330253495,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011991875376750303,
+      "loss": 0.8692,
+      "step": 9211
+    },
+    {
+      "epoch": 0.43648424543946934,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011990415937979525,
+      "loss": 1.3055,
+      "step": 9212
+    },
+    {
+      "epoch": 0.4365316275764037,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011988956455063442,
+      "loss": 1.3004,
+      "step": 9213
+    },
+    {
+      "epoch": 0.4365790097133381,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00011987496928034423,
+      "loss": 0.9849,
+      "step": 9214
+    },
+    {
+      "epoch": 0.43662639185027247,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.00011986037356924839,
+      "loss": 0.0081,
+      "step": 9215
+    },
+    {
+      "epoch": 0.4366737739872068,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011984577741767062,
+      "loss": 0.7798,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4367211561241412,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011983118082593466,
+      "loss": 0.1623,
+      "step": 9217
+    },
+    {
+      "epoch": 0.4367685382610756,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00011981658379436423,
+      "loss": 1.3366,
+      "step": 9218
+    },
+    {
+      "epoch": 0.4368159203980099,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011980198632328307,
+      "loss": 1.4171,
+      "step": 9219
+    },
+    {
+      "epoch": 0.4368633025349443,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.00011978738841301494,
+      "loss": 0.0048,
+      "step": 9220
+    },
+    {
+      "epoch": 0.4369106846718787,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011977279006388365,
+      "loss": 0.0173,
+      "step": 9221
+    },
+    {
+      "epoch": 0.43695806680881305,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011975819127621289,
+      "loss": 0.6574,
+      "step": 9222
+    },
+    {
+      "epoch": 0.43700544894574744,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011974359205032648,
+      "loss": 1.1759,
+      "step": 9223
+    },
+    {
+      "epoch": 0.43705283108268184,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00011972899238654827,
+      "loss": 0.0073,
+      "step": 9224
+    },
+    {
+      "epoch": 0.43710021321961623,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00011971439228520203,
+      "loss": 0.3249,
+      "step": 9225
+    },
+    {
+      "epoch": 0.43714759535655057,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00011969979174661152,
+      "loss": 0.088,
+      "step": 9226
+    },
+    {
+      "epoch": 0.43719497749348496,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011968519077110065,
+      "loss": 1.2362,
+      "step": 9227
+    },
+    {
+      "epoch": 0.43724235963041935,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001196705893589932,
+      "loss": 1.108,
+      "step": 9228
+    },
+    {
+      "epoch": 0.4372897417673537,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011965598751061307,
+      "loss": 0.0033,
+      "step": 9229
+    },
+    {
+      "epoch": 0.4373371239042881,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011964138522628403,
+      "loss": 0.0473,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4373845060412225,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011962678250632999,
+      "loss": 1.0528,
+      "step": 9231
+    },
+    {
+      "epoch": 0.4374318881781568,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011961217935107482,
+      "loss": 0.8092,
+      "step": 9232
+    },
+    {
+      "epoch": 0.4374792703150912,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011959757576084243,
+      "loss": 1.3905,
+      "step": 9233
+    },
+    {
+      "epoch": 0.4375266524520256,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00011958297173595666,
+      "loss": 1.2362,
+      "step": 9234
+    },
+    {
+      "epoch": 0.43757403458895994,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011956836727674143,
+      "loss": 1.259,
+      "step": 9235
+    },
+    {
+      "epoch": 0.43762141672589433,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011955376238352069,
+      "loss": 0.779,
+      "step": 9236
+    },
+    {
+      "epoch": 0.4376687988628287,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.00011953915705661829,
+      "loss": 0.0048,
+      "step": 9237
+    },
+    {
+      "epoch": 0.4377161809997631,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001195245512963582,
+      "loss": 0.0321,
+      "step": 9238
+    },
+    {
+      "epoch": 0.43776356313669745,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011950994510306437,
+      "loss": 0.7033,
+      "step": 9239
+    },
+    {
+      "epoch": 0.43781094527363185,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00011949533847706076,
+      "loss": 1.2161,
+      "step": 9240
+    },
+    {
+      "epoch": 0.43785832741056624,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011948073141867126,
+      "loss": 0.9879,
+      "step": 9241
+    },
+    {
+      "epoch": 0.4379057095475006,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011946612392821992,
+      "loss": 0.4406,
+      "step": 9242
+    },
+    {
+      "epoch": 0.43795309168443497,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001194515160060307,
+      "loss": 1.2926,
+      "step": 9243
+    },
+    {
+      "epoch": 0.43800047382136936,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011943690765242757,
+      "loss": 0.8098,
+      "step": 9244
+    },
+    {
+      "epoch": 0.4380478559583037,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00011942229886773451,
+      "loss": 0.9731,
+      "step": 9245
+    },
+    {
+      "epoch": 0.4380952380952381,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011940768965227553,
+      "loss": 0.9692,
+      "step": 9246
+    },
+    {
+      "epoch": 0.4381426202321725,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011939308000637471,
+      "loss": 0.6649,
+      "step": 9247
+    },
+    {
+      "epoch": 0.4381900023691068,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011937846993035602,
+      "loss": 1.4241,
+      "step": 9248
+    },
+    {
+      "epoch": 0.4382373845060412,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011936385942454348,
+      "loss": 0.6976,
+      "step": 9249
+    },
+    {
+      "epoch": 0.4382847666429756,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011934924848926122,
+      "loss": 0.0258,
+      "step": 9250
+    },
+    {
+      "epoch": 0.43833214877990995,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00011933463712483318,
+      "loss": 0.5183,
+      "step": 9251
+    },
+    {
+      "epoch": 0.43837953091684434,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011932002533158351,
+      "loss": 0.0113,
+      "step": 9252
+    },
+    {
+      "epoch": 0.43842691305377873,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011930541310983625,
+      "loss": 0.9669,
+      "step": 9253
+    },
+    {
+      "epoch": 0.4384742951907131,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00011929080045991549,
+      "loss": 0.2126,
+      "step": 9254
+    },
+    {
+      "epoch": 0.43852167732764746,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00011927618738214534,
+      "loss": 0.1515,
+      "step": 9255
+    },
+    {
+      "epoch": 0.43856905946458186,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011926157387684986,
+      "loss": 1.0796,
+      "step": 9256
+    },
+    {
+      "epoch": 0.43861644160151625,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001192469599443532,
+      "loss": 1.2017,
+      "step": 9257
+    },
+    {
+      "epoch": 0.4386638237384506,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011923234558497948,
+      "loss": 0.6035,
+      "step": 9258
+    },
+    {
+      "epoch": 0.438711205875385,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011921773079905283,
+      "loss": 0.9509,
+      "step": 9259
+    },
+    {
+      "epoch": 0.4387585880123194,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011920311558689734,
+      "loss": 0.2578,
+      "step": 9260
+    },
+    {
+      "epoch": 0.4388059701492537,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011918849994883721,
+      "loss": 0.9338,
+      "step": 9261
+    },
+    {
+      "epoch": 0.4388533522861881,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00011917388388519661,
+      "loss": 1.1853,
+      "step": 9262
+    },
+    {
+      "epoch": 0.4389007344231225,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00011915926739629968,
+      "loss": 0.7985,
+      "step": 9263
+    },
+    {
+      "epoch": 0.43894811656005683,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001191446504824706,
+      "loss": 0.7152,
+      "step": 9264
+    },
+    {
+      "epoch": 0.4389954986969912,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011913003314403357,
+      "loss": 0.4354,
+      "step": 9265
+    },
+    {
+      "epoch": 0.4390428808339256,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011911541538131279,
+      "loss": 0.8526,
+      "step": 9266
+    },
+    {
+      "epoch": 0.43909026297086,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011910079719463247,
+      "loss": 0.1765,
+      "step": 9267
+    },
+    {
+      "epoch": 0.43913764510779435,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011908617858431679,
+      "loss": 0.8484,
+      "step": 9268
+    },
+    {
+      "epoch": 0.43918502724472874,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011907155955068999,
+      "loss": 0.1277,
+      "step": 9269
+    },
+    {
+      "epoch": 0.43923240938166314,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011905694009407631,
+      "loss": 0.1151,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4392797915185975,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011904232021480002,
+      "loss": 0.9624,
+      "step": 9271
+    },
+    {
+      "epoch": 0.43932717365553187,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00011902769991318534,
+      "loss": 0.7456,
+      "step": 9272
+    },
+    {
+      "epoch": 0.43937455579246626,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00011901307918955653,
+      "loss": 0.6132,
+      "step": 9273
+    },
+    {
+      "epoch": 0.4394219379294006,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 0.0001189984580442379,
+      "loss": 0.003,
+      "step": 9274
+    },
+    {
+      "epoch": 0.439469320066335,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011898383647755374,
+      "loss": 0.9202,
+      "step": 9275
+    },
+    {
+      "epoch": 0.4395167022032694,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011896921448982825,
+      "loss": 1.2101,
+      "step": 9276
+    },
+    {
+      "epoch": 0.4395640843402037,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011895459208138579,
+      "loss": 0.7341,
+      "step": 9277
+    },
+    {
+      "epoch": 0.4396114664771381,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011893996925255069,
+      "loss": 0.0061,
+      "step": 9278
+    },
+    {
+      "epoch": 0.4396588486140725,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.00011892534600364725,
+      "loss": 0.003,
+      "step": 9279
+    },
+    {
+      "epoch": 0.43970623075100684,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00011891072233499976,
+      "loss": 0.8399,
+      "step": 9280
+    },
+    {
+      "epoch": 0.43975361288794124,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001188960982469326,
+      "loss": 0.874,
+      "step": 9281
+    },
+    {
+      "epoch": 0.43980099502487563,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011888147373977014,
+      "loss": 0.9002,
+      "step": 9282
+    },
+    {
+      "epoch": 0.43984837716181,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00011886684881383669,
+      "loss": 1.1015,
+      "step": 9283
+    },
+    {
+      "epoch": 0.43989575929874436,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011885222346945661,
+      "loss": 0.5635,
+      "step": 9284
+    },
+    {
+      "epoch": 0.43994314143567875,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011883759770695431,
+      "loss": 0.8537,
+      "step": 9285
+    },
+    {
+      "epoch": 0.43999052357261315,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011882297152665416,
+      "loss": 0.7777,
+      "step": 9286
+    },
+    {
+      "epoch": 0.4400379057095475,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011880834492888056,
+      "loss": 1.4362,
+      "step": 9287
+    },
+    {
+      "epoch": 0.4400852878464819,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011879371791395785,
+      "loss": 0.007,
+      "step": 9288
+    },
+    {
+      "epoch": 0.44013266998341627,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00011877909048221056,
+      "loss": 1.1929,
+      "step": 9289
+    },
+    {
+      "epoch": 0.4401800521203506,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00011876446263396303,
+      "loss": 0.0123,
+      "step": 9290
+    },
+    {
+      "epoch": 0.440227434257285,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001187498343695397,
+      "loss": 0.0907,
+      "step": 9291
+    },
+    {
+      "epoch": 0.4402748163942194,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000118735205689265,
+      "loss": 1.025,
+      "step": 9292
+    },
+    {
+      "epoch": 0.44032219853115373,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001187205765934634,
+      "loss": 0.1393,
+      "step": 9293
+    },
+    {
+      "epoch": 0.4403695806680881,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011870594708245937,
+      "loss": 0.9188,
+      "step": 9294
+    },
+    {
+      "epoch": 0.4404169628050225,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00011869131715657733,
+      "loss": 0.0155,
+      "step": 9295
+    },
+    {
+      "epoch": 0.4404643449419569,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011867668681614179,
+      "loss": 0.6337,
+      "step": 9296
+    },
+    {
+      "epoch": 0.44051172707889125,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00011866205606147725,
+      "loss": 0.5353,
+      "step": 9297
+    },
+    {
+      "epoch": 0.44055910921582564,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00011864742489290819,
+      "loss": 1.0244,
+      "step": 9298
+    },
+    {
+      "epoch": 0.44060649135276003,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0001186327933107591,
+      "loss": 0.0026,
+      "step": 9299
+    },
+    {
+      "epoch": 0.44065387348969437,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011861816131535447,
+      "loss": 1.4961,
+      "step": 9300
+    },
+    {
+      "epoch": 0.44070125562662876,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011860352890701889,
+      "loss": 1.1084,
+      "step": 9301
+    },
+    {
+      "epoch": 0.44074863776356316,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011858889608607684,
+      "loss": 0.1376,
+      "step": 9302
+    },
+    {
+      "epoch": 0.4407960199004975,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011857426285285287,
+      "loss": 1.4545,
+      "step": 9303
+    },
+    {
+      "epoch": 0.4408434020374319,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00011855962920767152,
+      "loss": 0.0084,
+      "step": 9304
+    },
+    {
+      "epoch": 0.4408907841743663,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011854499515085741,
+      "loss": 0.9402,
+      "step": 9305
+    },
+    {
+      "epoch": 0.4409381663113006,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011853036068273504,
+      "loss": 0.8362,
+      "step": 9306
+    },
+    {
+      "epoch": 0.440985548448235,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000118515725803629,
+      "loss": 0.0034,
+      "step": 9307
+    },
+    {
+      "epoch": 0.4410329305851694,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00011850109051386387,
+      "loss": 0.9941,
+      "step": 9308
+    },
+    {
+      "epoch": 0.44108031272210374,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011848645481376428,
+      "loss": 1.1298,
+      "step": 9309
+    },
+    {
+      "epoch": 0.44112769485903813,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011847181870365481,
+      "loss": 0.8825,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4411750769959725,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011845718218386006,
+      "loss": 0.8857,
+      "step": 9311
+    },
+    {
+      "epoch": 0.4412224591329069,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00011844254525470468,
+      "loss": 0.6584,
+      "step": 9312
+    },
+    {
+      "epoch": 0.44126984126984126,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00011842790791651333,
+      "loss": 0.9399,
+      "step": 9313
+    },
+    {
+      "epoch": 0.44131722340677565,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011841327016961055,
+      "loss": 1.0528,
+      "step": 9314
+    },
+    {
+      "epoch": 0.44136460554371004,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00011839863201432106,
+      "loss": 0.0161,
+      "step": 9315
+    },
+    {
+      "epoch": 0.4414119876806444,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011838399345096953,
+      "loss": 0.6023,
+      "step": 9316
+    },
+    {
+      "epoch": 0.4414593698175788,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.00011836935447988061,
+      "loss": 0.005,
+      "step": 9317
+    },
+    {
+      "epoch": 0.44150675195451317,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00011835471510137895,
+      "loss": 0.0147,
+      "step": 9318
+    },
+    {
+      "epoch": 0.4415541340914475,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011834007531578927,
+      "loss": 0.6816,
+      "step": 9319
+    },
+    {
+      "epoch": 0.4416015162283819,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011832543512343628,
+      "loss": 0.6251,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4416488983653163,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011831079452464464,
+      "loss": 1.2253,
+      "step": 9321
+    },
+    {
+      "epoch": 0.4416962805022506,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00011829615351973906,
+      "loss": 1.0539,
+      "step": 9322
+    },
+    {
+      "epoch": 0.441743662639185,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001182815121090443,
+      "loss": 1.1218,
+      "step": 9323
+    },
+    {
+      "epoch": 0.4417910447761194,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001182668702928851,
+      "loss": 0.6269,
+      "step": 9324
+    },
+    {
+      "epoch": 0.4418384269130538,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011825222807158613,
+      "loss": 1.0764,
+      "step": 9325
+    },
+    {
+      "epoch": 0.44188580904998814,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011823758544547221,
+      "loss": 0.1038,
+      "step": 9326
+    },
+    {
+      "epoch": 0.44193319118692254,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011822294241486805,
+      "loss": 0.1845,
+      "step": 9327
+    },
+    {
+      "epoch": 0.44198057332385693,
+      "grad_norm": 0.02978515625,
+      "learning_rate": 0.00011820829898009847,
+      "loss": 0.0009,
+      "step": 9328
+    },
+    {
+      "epoch": 0.44202795546079127,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011819365514148819,
+      "loss": 1.1411,
+      "step": 9329
+    },
+    {
+      "epoch": 0.44207533759772566,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011817901089936201,
+      "loss": 0.0144,
+      "step": 9330
+    },
+    {
+      "epoch": 0.44212271973466005,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011816436625404474,
+      "loss": 0.0671,
+      "step": 9331
+    },
+    {
+      "epoch": 0.4421701018715944,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011814972120586118,
+      "loss": 1.2879,
+      "step": 9332
+    },
+    {
+      "epoch": 0.4422174840085288,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011813507575513614,
+      "loss": 0.9976,
+      "step": 9333
+    },
+    {
+      "epoch": 0.4422648661454632,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011812042990219441,
+      "loss": 0.8429,
+      "step": 9334
+    },
+    {
+      "epoch": 0.4423122482823975,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011810578364736089,
+      "loss": 1.0488,
+      "step": 9335
+    },
+    {
+      "epoch": 0.4423596304193319,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011809113699096033,
+      "loss": 1.1678,
+      "step": 9336
+    },
+    {
+      "epoch": 0.4424070125562663,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011807648993331766,
+      "loss": 0.9482,
+      "step": 9337
+    },
+    {
+      "epoch": 0.44245439469320064,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011806184247475766,
+      "loss": 0.8572,
+      "step": 9338
+    },
+    {
+      "epoch": 0.44250177683013503,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011804719461560525,
+      "loss": 0.8175,
+      "step": 9339
+    },
+    {
+      "epoch": 0.4425491589670694,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011803254635618531,
+      "loss": 1.0617,
+      "step": 9340
+    },
+    {
+      "epoch": 0.4425965411040038,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011801789769682266,
+      "loss": 0.6566,
+      "step": 9341
+    },
+    {
+      "epoch": 0.44264392324093815,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011800324863784225,
+      "loss": 1.1249,
+      "step": 9342
+    },
+    {
+      "epoch": 0.44269130537787255,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011798859917956897,
+      "loss": 0.5442,
+      "step": 9343
+    },
+    {
+      "epoch": 0.44273868751480694,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011797394932232769,
+      "loss": 0.6662,
+      "step": 9344
+    },
+    {
+      "epoch": 0.4427860696517413,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00011795929906644339,
+      "loss": 1.1757,
+      "step": 9345
+    },
+    {
+      "epoch": 0.44283345178867567,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00011794464841224095,
+      "loss": 0.2145,
+      "step": 9346
+    },
+    {
+      "epoch": 0.44288083392561006,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011792999736004533,
+      "loss": 0.75,
+      "step": 9347
+    },
+    {
+      "epoch": 0.4429282160625444,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011791534591018148,
+      "loss": 1.4978,
+      "step": 9348
+    },
+    {
+      "epoch": 0.4429755981994788,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011790069406297431,
+      "loss": 0.9863,
+      "step": 9349
+    },
+    {
+      "epoch": 0.4430229803364132,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011788604181874881,
+      "loss": 0.8307,
+      "step": 9350
+    },
+    {
+      "epoch": 0.4430703624733475,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011787138917782999,
+      "loss": 1.3348,
+      "step": 9351
+    },
+    {
+      "epoch": 0.4431177446102819,
+      "grad_norm": 0.004852294921875,
+      "learning_rate": 0.00011785673614054277,
+      "loss": 0.0002,
+      "step": 9352
+    },
+    {
+      "epoch": 0.4431651267472163,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00011784208270721216,
+      "loss": 0.1688,
+      "step": 9353
+    },
+    {
+      "epoch": 0.4432125088841507,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011782742887816316,
+      "loss": 1.3078,
+      "step": 9354
+    },
+    {
+      "epoch": 0.44325989102108504,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001178127746537208,
+      "loss": 1.1072,
+      "step": 9355
+    },
+    {
+      "epoch": 0.44330727315801943,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011779812003421004,
+      "loss": 0.9367,
+      "step": 9356
+    },
+    {
+      "epoch": 0.4433546552949538,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011778346501995594,
+      "loss": 0.7521,
+      "step": 9357
+    },
+    {
+      "epoch": 0.44340203743188816,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00011776880961128356,
+      "loss": 0.9882,
+      "step": 9358
+    },
+    {
+      "epoch": 0.44344941956882256,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011775415380851787,
+      "loss": 1.1479,
+      "step": 9359
+    },
+    {
+      "epoch": 0.44349680170575695,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000117739497611984,
+      "loss": 0.1426,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4435441838426913,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011772484102200693,
+      "loss": 1.0151,
+      "step": 9361
+    },
+    {
+      "epoch": 0.4435915659796257,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011771018403891176,
+      "loss": 1.0357,
+      "step": 9362
+    },
+    {
+      "epoch": 0.4436389481165601,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00011769552666302361,
+      "loss": 1.2855,
+      "step": 9363
+    },
+    {
+      "epoch": 0.4436863302534944,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011768086889466752,
+      "loss": 1.3557,
+      "step": 9364
+    },
+    {
+      "epoch": 0.4437337123904288,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011766621073416857,
+      "loss": 0.1146,
+      "step": 9365
+    },
+    {
+      "epoch": 0.4437810945273632,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011765155218185188,
+      "loss": 0.685,
+      "step": 9366
+    },
+    {
+      "epoch": 0.44382847666429753,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011763689323804258,
+      "loss": 0.8738,
+      "step": 9367
+    },
+    {
+      "epoch": 0.4438758588012319,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00011762223390306579,
+      "loss": 0.9329,
+      "step": 9368
+    },
+    {
+      "epoch": 0.4439232409381663,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011760757417724659,
+      "loss": 0.0706,
+      "step": 9369
+    },
+    {
+      "epoch": 0.4439706230751007,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00011759291406091018,
+      "loss": 0.0074,
+      "step": 9370
+    },
+    {
+      "epoch": 0.44401800521203505,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00011757825355438165,
+      "loss": 1.3688,
+      "step": 9371
+    },
+    {
+      "epoch": 0.44406538734896944,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011756359265798618,
+      "loss": 0.2759,
+      "step": 9372
+    },
+    {
+      "epoch": 0.44411276948590384,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 0.00011754893137204895,
+      "loss": 0.0032,
+      "step": 9373
+    },
+    {
+      "epoch": 0.4441601516228382,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001175342696968951,
+      "loss": 0.7585,
+      "step": 9374
+    },
+    {
+      "epoch": 0.44420753375977257,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00011751960763284981,
+      "loss": 0.0044,
+      "step": 9375
+    },
+    {
+      "epoch": 0.44425491589670696,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011750494518023834,
+      "loss": 0.7121,
+      "step": 9376
+    },
+    {
+      "epoch": 0.4443022980336413,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00011749028233938577,
+      "loss": 0.8757,
+      "step": 9377
+    },
+    {
+      "epoch": 0.4443496801705757,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00011747561911061742,
+      "loss": 0.1564,
+      "step": 9378
+    },
+    {
+      "epoch": 0.4443970623075101,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011746095549425841,
+      "loss": 1.078,
+      "step": 9379
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00011744629149063403,
+      "loss": 0.6784,
+      "step": 9380
+    },
+    {
+      "epoch": 0.4444918265813788,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011743162710006947,
+      "loss": 0.051,
+      "step": 9381
+    },
+    {
+      "epoch": 0.4445392087183132,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00011741696232288998,
+      "loss": 0.983,
+      "step": 9382
+    },
+    {
+      "epoch": 0.4445865908552476,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011740229715942083,
+      "loss": 0.6347,
+      "step": 9383
+    },
+    {
+      "epoch": 0.44463397299218194,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011738763160998729,
+      "loss": 0.9181,
+      "step": 9384
+    },
+    {
+      "epoch": 0.44468135512911633,
+      "grad_norm": 0.0023345947265625,
+      "learning_rate": 0.00011737296567491459,
+      "loss": 0.0001,
+      "step": 9385
+    },
+    {
+      "epoch": 0.4447287372660507,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000117358299354528,
+      "loss": 0.0823,
+      "step": 9386
+    },
+    {
+      "epoch": 0.44477611940298506,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011734363264915282,
+      "loss": 0.7144,
+      "step": 9387
+    },
+    {
+      "epoch": 0.44482350153991945,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00011732896555911435,
+      "loss": 0.1086,
+      "step": 9388
+    },
+    {
+      "epoch": 0.44487088367685385,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011731429808473788,
+      "loss": 0.1366,
+      "step": 9389
+    },
+    {
+      "epoch": 0.4449182658137882,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00011729963022634872,
+      "loss": 1.0817,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4449656479507226,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011728496198427217,
+      "loss": 0.7142,
+      "step": 9391
+    },
+    {
+      "epoch": 0.44501303008765697,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011727029335883361,
+      "loss": 0.649,
+      "step": 9392
+    },
+    {
+      "epoch": 0.4450604122245913,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011725562435035833,
+      "loss": 1.0864,
+      "step": 9393
+    },
+    {
+      "epoch": 0.4451077943615257,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011724095495917167,
+      "loss": 0.5156,
+      "step": 9394
+    },
+    {
+      "epoch": 0.4451551764984601,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00011722628518559899,
+      "loss": 0.0558,
+      "step": 9395
+    },
+    {
+      "epoch": 0.44520255863539443,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011721161502996565,
+      "loss": 0.5618,
+      "step": 9396
+    },
+    {
+      "epoch": 0.4452499407723288,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00011719694449259706,
+      "loss": 0.7703,
+      "step": 9397
+    },
+    {
+      "epoch": 0.4452973229092632,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001171822735738185,
+      "loss": 0.9965,
+      "step": 9398
+    },
+    {
+      "epoch": 0.4453447050461976,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011716760227395545,
+      "loss": 1.0266,
+      "step": 9399
+    },
+    {
+      "epoch": 0.44539208718313195,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00011715293059333327,
+      "loss": 0.0017,
+      "step": 9400
+    },
+    {
+      "epoch": 0.44543946932006634,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011713825853227738,
+      "loss": 1.2956,
+      "step": 9401
+    },
+    {
+      "epoch": 0.44548685145700073,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011712358609111311,
+      "loss": 1.0393,
+      "step": 9402
+    },
+    {
+      "epoch": 0.44553423359393507,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00011710891327016597,
+      "loss": 1.3742,
+      "step": 9403
+    },
+    {
+      "epoch": 0.44558161573086946,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011709424006976138,
+      "loss": 1.0638,
+      "step": 9404
+    },
+    {
+      "epoch": 0.44562899786780386,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001170795664902247,
+      "loss": 0.94,
+      "step": 9405
+    },
+    {
+      "epoch": 0.4456763800047382,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011706489253188147,
+      "loss": 0.5283,
+      "step": 9406
+    },
+    {
+      "epoch": 0.4457237621416726,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001170502181950571,
+      "loss": 1.1417,
+      "step": 9407
+    },
+    {
+      "epoch": 0.445771144278607,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00011703554348007704,
+      "loss": 0.9421,
+      "step": 9408
+    },
+    {
+      "epoch": 0.4458185264155413,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011702086838726679,
+      "loss": 0.7954,
+      "step": 9409
+    },
+    {
+      "epoch": 0.4458659085524757,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011700619291695178,
+      "loss": 1.0548,
+      "step": 9410
+    },
+    {
+      "epoch": 0.4459132906894101,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011699151706945753,
+      "loss": 0.0486,
+      "step": 9411
+    },
+    {
+      "epoch": 0.4459606728263445,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00011697684084510956,
+      "loss": 0.8602,
+      "step": 9412
+    },
+    {
+      "epoch": 0.44600805496327883,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001169621642442333,
+      "loss": 0.7969,
+      "step": 9413
+    },
+    {
+      "epoch": 0.4460554371002132,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011694748726715432,
+      "loss": 0.1151,
+      "step": 9414
+    },
+    {
+      "epoch": 0.4461028192371476,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00011693280991419815,
+      "loss": 0.0798,
+      "step": 9415
+    },
+    {
+      "epoch": 0.44615020137408196,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011691813218569027,
+      "loss": 0.7561,
+      "step": 9416
+    },
+    {
+      "epoch": 0.44619758351101635,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011690345408195624,
+      "loss": 1.4186,
+      "step": 9417
+    },
+    {
+      "epoch": 0.44624496564795074,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011688877560332158,
+      "loss": 0.9857,
+      "step": 9418
+    },
+    {
+      "epoch": 0.4462923477848851,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011687409675011189,
+      "loss": 0.8451,
+      "step": 9419
+    },
+    {
+      "epoch": 0.4463397299218195,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011685941752265272,
+      "loss": 0.5844,
+      "step": 9420
+    },
+    {
+      "epoch": 0.44638711205875387,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001168447379212696,
+      "loss": 0.0114,
+      "step": 9421
+    },
+    {
+      "epoch": 0.4464344941956882,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00011683005794628814,
+      "loss": 0.017,
+      "step": 9422
+    },
+    {
+      "epoch": 0.4464818763326226,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00011681537759803393,
+      "loss": 0.0175,
+      "step": 9423
+    },
+    {
+      "epoch": 0.446529258469557,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001168006968768326,
+      "loss": 0.9998,
+      "step": 9424
+    },
+    {
+      "epoch": 0.4465766406064913,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011678601578300965,
+      "loss": 0.4831,
+      "step": 9425
+    },
+    {
+      "epoch": 0.4466240227434257,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011677133431689075,
+      "loss": 0.8602,
+      "step": 9426
+    },
+    {
+      "epoch": 0.4466714048803601,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011675665247880154,
+      "loss": 0.6564,
+      "step": 9427
+    },
+    {
+      "epoch": 0.4467187870172945,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011674197026906762,
+      "loss": 0.7945,
+      "step": 9428
+    },
+    {
+      "epoch": 0.44676616915422884,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011672728768801461,
+      "loss": 0.4851,
+      "step": 9429
+    },
+    {
+      "epoch": 0.44681355129116324,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011671260473596819,
+      "loss": 0.5367,
+      "step": 9430
+    },
+    {
+      "epoch": 0.44686093342809763,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011669792141325402,
+      "loss": 0.5629,
+      "step": 9431
+    },
+    {
+      "epoch": 0.44690831556503197,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011668323772019774,
+      "loss": 0.0414,
+      "step": 9432
+    },
+    {
+      "epoch": 0.44695569770196636,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00011666855365712499,
+      "loss": 1.2052,
+      "step": 9433
+    },
+    {
+      "epoch": 0.44700307983890075,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011665386922436148,
+      "loss": 0.7529,
+      "step": 9434
+    },
+    {
+      "epoch": 0.4470504619758351,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011663918442223292,
+      "loss": 1.1898,
+      "step": 9435
+    },
+    {
+      "epoch": 0.4470978441127695,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00011662449925106493,
+      "loss": 0.0194,
+      "step": 9436
+    },
+    {
+      "epoch": 0.4471452262497039,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00011660981371118327,
+      "loss": 0.4117,
+      "step": 9437
+    },
+    {
+      "epoch": 0.4471926083866382,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011659512780291364,
+      "loss": 1.0694,
+      "step": 9438
+    },
+    {
+      "epoch": 0.4472399905235726,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011658044152658179,
+      "loss": 0.128,
+      "step": 9439
+    },
+    {
+      "epoch": 0.447287372660507,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011656575488251333,
+      "loss": 0.9505,
+      "step": 9440
+    },
+    {
+      "epoch": 0.4473347547974414,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00011655106787103411,
+      "loss": 0.016,
+      "step": 9441
+    },
+    {
+      "epoch": 0.44738213693437573,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011653638049246982,
+      "loss": 0.5005,
+      "step": 9442
+    },
+    {
+      "epoch": 0.4474295190713101,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011652169274714626,
+      "loss": 1.0562,
+      "step": 9443
+    },
+    {
+      "epoch": 0.4474769012082445,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001165070046353891,
+      "loss": 0.2267,
+      "step": 9444
+    },
+    {
+      "epoch": 0.44752428334517885,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001164923161575242,
+      "loss": 1.0425,
+      "step": 9445
+    },
+    {
+      "epoch": 0.44757166548211325,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001164776273138773,
+      "loss": 0.9655,
+      "step": 9446
+    },
+    {
+      "epoch": 0.44761904761904764,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011646293810477418,
+      "loss": 1.1615,
+      "step": 9447
+    },
+    {
+      "epoch": 0.447666429755982,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001164482485305406,
+      "loss": 1.0926,
+      "step": 9448
+    },
+    {
+      "epoch": 0.44771381189291637,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011643355859150237,
+      "loss": 0.1574,
+      "step": 9449
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011641886828798536,
+      "loss": 0.8106,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4478085761667851,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011640417762031533,
+      "loss": 0.1354,
+      "step": 9451
+    },
+    {
+      "epoch": 0.4478559583037195,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011638948658881808,
+      "loss": 0.5431,
+      "step": 9452
+    },
+    {
+      "epoch": 0.4479033404406539,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011637479519381949,
+      "loss": 1.068,
+      "step": 9453
+    },
+    {
+      "epoch": 0.4479507225775882,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001163601034356454,
+      "loss": 1.0824,
+      "step": 9454
+    },
+    {
+      "epoch": 0.4479981047145226,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011634541131462167,
+      "loss": 0.9222,
+      "step": 9455
+    },
+    {
+      "epoch": 0.448045486851457,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011633071883107407,
+      "loss": 0.1449,
+      "step": 9456
+    },
+    {
+      "epoch": 0.4480928689883914,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011631602598532854,
+      "loss": 0.99,
+      "step": 9457
+    },
+    {
+      "epoch": 0.44814025112532574,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011630133277771096,
+      "loss": 0.6691,
+      "step": 9458
+    },
+    {
+      "epoch": 0.44818763326226013,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011628663920854714,
+      "loss": 0.7797,
+      "step": 9459
+    },
+    {
+      "epoch": 0.4482350153991945,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00011627194527816304,
+      "loss": 0.6866,
+      "step": 9460
+    },
+    {
+      "epoch": 0.44828239753612886,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001162572509868845,
+      "loss": 1.1991,
+      "step": 9461
+    },
+    {
+      "epoch": 0.44832977967306326,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011624255633503749,
+      "loss": 1.1431,
+      "step": 9462
+    },
+    {
+      "epoch": 0.44837716180999765,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011622786132294784,
+      "loss": 1.2445,
+      "step": 9463
+    },
+    {
+      "epoch": 0.448424543946932,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011621316595094151,
+      "loss": 0.8336,
+      "step": 9464
+    },
+    {
+      "epoch": 0.4484719260838664,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011619847021934445,
+      "loss": 1.1235,
+      "step": 9465
+    },
+    {
+      "epoch": 0.44851930822080077,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011618377412848258,
+      "loss": 1.0955,
+      "step": 9466
+    },
+    {
+      "epoch": 0.4485666903577351,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00011616907767868181,
+      "loss": 0.2217,
+      "step": 9467
+    },
+    {
+      "epoch": 0.4486140724946695,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011615438087026816,
+      "loss": 0.8414,
+      "step": 9468
+    },
+    {
+      "epoch": 0.4486614546316039,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00011613968370356754,
+      "loss": 1.3303,
+      "step": 9469
+    },
+    {
+      "epoch": 0.4487088367685383,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001161249861789059,
+      "loss": 0.9061,
+      "step": 9470
+    },
+    {
+      "epoch": 0.4487562189054726,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011611028829660925,
+      "loss": 0.924,
+      "step": 9471
+    },
+    {
+      "epoch": 0.448803601042407,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00011609559005700356,
+      "loss": 1.097,
+      "step": 9472
+    },
+    {
+      "epoch": 0.4488509831793414,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011608089146041487,
+      "loss": 0.8251,
+      "step": 9473
+    },
+    {
+      "epoch": 0.44889836531627575,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0001160661925071691,
+      "loss": 0.0071,
+      "step": 9474
+    },
+    {
+      "epoch": 0.44894574745321014,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011605149319759228,
+      "loss": 1.0481,
+      "step": 9475
+    },
+    {
+      "epoch": 0.44899312959014454,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011603679353201044,
+      "loss": 1.073,
+      "step": 9476
+    },
+    {
+      "epoch": 0.4490405117270789,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00011602209351074963,
+      "loss": 0.9888,
+      "step": 9477
+    },
+    {
+      "epoch": 0.44908789386401327,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011600739313413587,
+      "loss": 0.9214,
+      "step": 9478
+    },
+    {
+      "epoch": 0.44913527600094766,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011599269240249513,
+      "loss": 0.9524,
+      "step": 9479
+    },
+    {
+      "epoch": 0.449182658137882,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011597799131615352,
+      "loss": 0.1399,
+      "step": 9480
+    },
+    {
+      "epoch": 0.4492300402748164,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0001159632898754371,
+      "loss": 0.0011,
+      "step": 9481
+    },
+    {
+      "epoch": 0.4492774224117508,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011594858808067188,
+      "loss": 0.6639,
+      "step": 9482
+    },
+    {
+      "epoch": 0.4493248045486851,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 0.000115933885932184,
+      "loss": 0.0029,
+      "step": 9483
+    },
+    {
+      "epoch": 0.4493721866856195,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011591918343029946,
+      "loss": 0.8705,
+      "step": 9484
+    },
+    {
+      "epoch": 0.4494195688225539,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.00011590448057534445,
+      "loss": 0.007,
+      "step": 9485
+    },
+    {
+      "epoch": 0.4494669509594883,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00011588977736764497,
+      "loss": 0.0142,
+      "step": 9486
+    },
+    {
+      "epoch": 0.44951433309642264,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011587507380752713,
+      "loss": 0.7774,
+      "step": 9487
+    },
+    {
+      "epoch": 0.44956171523335703,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011586036989531707,
+      "loss": 0.823,
+      "step": 9488
+    },
+    {
+      "epoch": 0.4496090973702914,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011584566563134094,
+      "loss": 0.9035,
+      "step": 9489
+    },
+    {
+      "epoch": 0.44965647950722576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011583096101592479,
+      "loss": 1.0848,
+      "step": 9490
+    },
+    {
+      "epoch": 0.44970386164416015,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011581625604939477,
+      "loss": 1.6401,
+      "step": 9491
+    },
+    {
+      "epoch": 0.44975124378109455,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011580155073207707,
+      "loss": 1.2194,
+      "step": 9492
+    },
+    {
+      "epoch": 0.4497986259180289,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001157868450642978,
+      "loss": 0.0073,
+      "step": 9493
+    },
+    {
+      "epoch": 0.4498460080549633,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00011577213904638313,
+      "loss": 1.1233,
+      "step": 9494
+    },
+    {
+      "epoch": 0.44989339019189767,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001157574326786592,
+      "loss": 0.1325,
+      "step": 9495
+    },
+    {
+      "epoch": 0.449940772328832,
+      "grad_norm": 0.0274658203125,
+      "learning_rate": 0.00011574272596145221,
+      "loss": 0.0009,
+      "step": 9496
+    },
+    {
+      "epoch": 0.4499881544657664,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011572801889508831,
+      "loss": 0.1232,
+      "step": 9497
+    },
+    {
+      "epoch": 0.4500355366027008,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001157133114798937,
+      "loss": 0.1474,
+      "step": 9498
+    },
+    {
+      "epoch": 0.4500829187396352,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011569860371619459,
+      "loss": 0.1166,
+      "step": 9499
+    },
+    {
+      "epoch": 0.4501303008765695,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00011568389560431719,
+      "loss": 0.7246,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4501776830135039,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011566918714458765,
+      "loss": 0.4216,
+      "step": 9501
+    },
+    {
+      "epoch": 0.4502250651504383,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011565447833733227,
+      "loss": 0.0335,
+      "step": 9502
+    },
+    {
+      "epoch": 0.45027244728737265,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0001156397691828772,
+      "loss": 0.0089,
+      "step": 9503
+    },
+    {
+      "epoch": 0.45031982942430704,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00011562505968154876,
+      "loss": 0.4767,
+      "step": 9504
+    },
+    {
+      "epoch": 0.45036721156124143,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011561034983367307,
+      "loss": 0.5419,
+      "step": 9505
+    },
+    {
+      "epoch": 0.45041459369817577,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011559563963957649,
+      "loss": 0.4571,
+      "step": 9506
+    },
+    {
+      "epoch": 0.45046197583511016,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001155809290995852,
+      "loss": 0.8394,
+      "step": 9507
+    },
+    {
+      "epoch": 0.45050935797204456,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00011556621821402556,
+      "loss": 0.008,
+      "step": 9508
+    },
+    {
+      "epoch": 0.4505567401089789,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00011555150698322372,
+      "loss": 1.2913,
+      "step": 9509
+    },
+    {
+      "epoch": 0.4506041222459133,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00011553679540750605,
+      "loss": 0.0034,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4506515043828477,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001155220834871988,
+      "loss": 0.0453,
+      "step": 9511
+    },
+    {
+      "epoch": 0.450698886519782,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001155073712226283,
+      "loss": 0.1167,
+      "step": 9512
+    },
+    {
+      "epoch": 0.4507462686567164,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011549265861412077,
+      "loss": 1.2963,
+      "step": 9513
+    },
+    {
+      "epoch": 0.4507936507936508,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011547794566200261,
+      "loss": 1.3112,
+      "step": 9514
+    },
+    {
+      "epoch": 0.4508410329305852,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001154632323666001,
+      "loss": 0.1733,
+      "step": 9515
+    },
+    {
+      "epoch": 0.45088841506751953,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011544851872823956,
+      "loss": 0.8656,
+      "step": 9516
+    },
+    {
+      "epoch": 0.4509357972044539,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00011543380474724735,
+      "loss": 1.2624,
+      "step": 9517
+    },
+    {
+      "epoch": 0.4509831793413883,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011541909042394974,
+      "loss": 1.0939,
+      "step": 9518
+    },
+    {
+      "epoch": 0.45103056147832266,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00011540437575867315,
+      "loss": 0.6567,
+      "step": 9519
+    },
+    {
+      "epoch": 0.45107794361525705,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00011538966075174396,
+      "loss": 0.0171,
+      "step": 9520
+    },
+    {
+      "epoch": 0.45112532575219144,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011537494540348843,
+      "loss": 0.7626,
+      "step": 9521
+    },
+    {
+      "epoch": 0.4511727078891258,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.000115360229714233,
+      "loss": 1.0279,
+      "step": 9522
+    },
+    {
+      "epoch": 0.45122009002606017,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011534551368430404,
+      "loss": 0.9124,
+      "step": 9523
+    },
+    {
+      "epoch": 0.45126747216299457,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011533079731402793,
+      "loss": 0.5132,
+      "step": 9524
+    },
+    {
+      "epoch": 0.4513148542999289,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011531608060373109,
+      "loss": 1.246,
+      "step": 9525
+    },
+    {
+      "epoch": 0.4513622364368633,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011530136355373986,
+      "loss": 1.0366,
+      "step": 9526
+    },
+    {
+      "epoch": 0.4514096185737977,
+      "grad_norm": 0.01202392578125,
+      "learning_rate": 0.00011528664616438072,
+      "loss": 0.0007,
+      "step": 9527
+    },
+    {
+      "epoch": 0.4514570007107321,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00011527192843598002,
+      "loss": 1.3601,
+      "step": 9528
+    },
+    {
+      "epoch": 0.4515043828476664,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011525721036886421,
+      "loss": 0.1206,
+      "step": 9529
+    },
+    {
+      "epoch": 0.4515517649846008,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00011524249196335974,
+      "loss": 0.0013,
+      "step": 9530
+    },
+    {
+      "epoch": 0.4515991471215352,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011522777321979303,
+      "loss": 0.4378,
+      "step": 9531
+    },
+    {
+      "epoch": 0.45164652925846954,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00011521305413849055,
+      "loss": 0.9599,
+      "step": 9532
+    },
+    {
+      "epoch": 0.45169391139540394,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011519833471977873,
+      "loss": 1.0324,
+      "step": 9533
+    },
+    {
+      "epoch": 0.45174129353233833,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000115183614963984,
+      "loss": 0.0441,
+      "step": 9534
+    },
+    {
+      "epoch": 0.45178867566927267,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001151688948714329,
+      "loss": 1.2256,
+      "step": 9535
+    },
+    {
+      "epoch": 0.45183605780620706,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011515417444245185,
+      "loss": 1.162,
+      "step": 9536
+    },
+    {
+      "epoch": 0.45188343994314145,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011513945367736733,
+      "loss": 0.1426,
+      "step": 9537
+    },
+    {
+      "epoch": 0.4519308220800758,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 0.00011512473257650592,
+      "loss": 0.0041,
+      "step": 9538
+    },
+    {
+      "epoch": 0.4519782042170102,
+      "grad_norm": 0.0277099609375,
+      "learning_rate": 0.00011511001114019399,
+      "loss": 0.002,
+      "step": 9539
+    },
+    {
+      "epoch": 0.4520255863539446,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001150952893687581,
+      "loss": 0.0135,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4520729684908789,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011508056726252482,
+      "loss": 1.2194,
+      "step": 9541
+    },
+    {
+      "epoch": 0.4521203506278133,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011506584482182061,
+      "loss": 0.9896,
+      "step": 9542
+    },
+    {
+      "epoch": 0.4521677327647477,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000115051122046972,
+      "loss": 0.9118,
+      "step": 9543
+    },
+    {
+      "epoch": 0.4522151149016821,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011503639893830553,
+      "loss": 1.2259,
+      "step": 9544
+    },
+    {
+      "epoch": 0.45226249703861643,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00011502167549614773,
+      "loss": 0.0313,
+      "step": 9545
+    },
+    {
+      "epoch": 0.4523098791755508,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00011500695172082521,
+      "loss": 0.1573,
+      "step": 9546
+    },
+    {
+      "epoch": 0.4523572613124852,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011499222761266446,
+      "loss": 0.8081,
+      "step": 9547
+    },
+    {
+      "epoch": 0.45240464344941955,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011497750317199208,
+      "loss": 0.8415,
+      "step": 9548
+    },
+    {
+      "epoch": 0.45245202558635395,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00011496277839913463,
+      "loss": 0.1733,
+      "step": 9549
+    },
+    {
+      "epoch": 0.45249940772328834,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011494805329441871,
+      "loss": 1.1652,
+      "step": 9550
+    },
+    {
+      "epoch": 0.4525467898602227,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011493332785817086,
+      "loss": 1.0033,
+      "step": 9551
+    },
+    {
+      "epoch": 0.45259417199715707,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001149186020907177,
+      "loss": 1.1452,
+      "step": 9552
+    },
+    {
+      "epoch": 0.45264155413409146,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00011490387599238586,
+      "loss": 0.017,
+      "step": 9553
+    },
+    {
+      "epoch": 0.4526889362710258,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00011488914956350191,
+      "loss": 1.0636,
+      "step": 9554
+    },
+    {
+      "epoch": 0.4527363184079602,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011487442280439249,
+      "loss": 0.8769,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4527837005448946,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001148596957153842,
+      "loss": 0.8713,
+      "step": 9556
+    },
+    {
+      "epoch": 0.452831082681829,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011484496829680371,
+      "loss": 1.2281,
+      "step": 9557
+    },
+    {
+      "epoch": 0.4528784648187633,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.00011483024054897764,
+      "loss": 0.0025,
+      "step": 9558
+    },
+    {
+      "epoch": 0.4529258469556977,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00011481551247223261,
+      "loss": 0.0514,
+      "step": 9559
+    },
+    {
+      "epoch": 0.4529732290926321,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001148007840668953,
+      "loss": 0.5776,
+      "step": 9560
+    },
+    {
+      "epoch": 0.45302061122956644,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011478605533329239,
+      "loss": 0.1592,
+      "step": 9561
+    },
+    {
+      "epoch": 0.45306799336650083,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011477132627175046,
+      "loss": 0.8947,
+      "step": 9562
+    },
+    {
+      "epoch": 0.4531153755034352,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001147565968825963,
+      "loss": 0.0262,
+      "step": 9563
+    },
+    {
+      "epoch": 0.45316275764036956,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011474186716615651,
+      "loss": 0.9152,
+      "step": 9564
+    },
+    {
+      "epoch": 0.45321013977730396,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011472713712275784,
+      "loss": 0.8025,
+      "step": 9565
+    },
+    {
+      "epoch": 0.45325752191423835,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011471240675272694,
+      "loss": 0.4409,
+      "step": 9566
+    },
+    {
+      "epoch": 0.4533049040511727,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011469767605639052,
+      "loss": 1.0586,
+      "step": 9567
+    },
+    {
+      "epoch": 0.4533522861881071,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00011468294503407531,
+      "loss": 0.1058,
+      "step": 9568
+    },
+    {
+      "epoch": 0.45339966832504147,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011466821368610803,
+      "loss": 0.0348,
+      "step": 9569
+    },
+    {
+      "epoch": 0.4534470504619758,
+      "grad_norm": 0.0303955078125,
+      "learning_rate": 0.00011465348201281538,
+      "loss": 0.0017,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4534944325989102,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011463875001452409,
+      "loss": 0.9173,
+      "step": 9571
+    },
+    {
+      "epoch": 0.4535418147358446,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00011462401769156095,
+      "loss": 0.0327,
+      "step": 9572
+    },
+    {
+      "epoch": 0.453589196872779,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011460928504425267,
+      "loss": 1.0341,
+      "step": 9573
+    },
+    {
+      "epoch": 0.4536365790097133,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.000114594552072926,
+      "loss": 0.8655,
+      "step": 9574
+    },
+    {
+      "epoch": 0.4536839611466477,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011457981877790769,
+      "loss": 1.1325,
+      "step": 9575
+    },
+    {
+      "epoch": 0.4537313432835821,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011456508515952456,
+      "loss": 1.2329,
+      "step": 9576
+    },
+    {
+      "epoch": 0.45377872542051645,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00011455035121810334,
+      "loss": 0.3296,
+      "step": 9577
+    },
+    {
+      "epoch": 0.45382610755745084,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011453561695397081,
+      "loss": 0.0984,
+      "step": 9578
+    },
+    {
+      "epoch": 0.45387348969438523,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00011452088236745382,
+      "loss": 1.1948,
+      "step": 9579
+    },
+    {
+      "epoch": 0.4539208718313196,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00011450614745887911,
+      "loss": 1.0574,
+      "step": 9580
+    },
+    {
+      "epoch": 0.45396825396825397,
+      "grad_norm": 0.02001953125,
+      "learning_rate": 0.00011449141222857351,
+      "loss": 0.001,
+      "step": 9581
+    },
+    {
+      "epoch": 0.45401563610518836,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011447667667686379,
+      "loss": 0.9499,
+      "step": 9582
+    },
+    {
+      "epoch": 0.4540630182421227,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011446194080407682,
+      "loss": 1.3364,
+      "step": 9583
+    },
+    {
+      "epoch": 0.4541104003790571,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011444720461053942,
+      "loss": 0.6499,
+      "step": 9584
+    },
+    {
+      "epoch": 0.4541577825159915,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011443246809657839,
+      "loss": 0.2543,
+      "step": 9585
+    },
+    {
+      "epoch": 0.4542051646529259,
+      "grad_norm": 0.00191497802734375,
+      "learning_rate": 0.0001144177312625206,
+      "loss": 0.0001,
+      "step": 9586
+    },
+    {
+      "epoch": 0.4542525467898602,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00011440299410869286,
+      "loss": 0.147,
+      "step": 9587
+    },
+    {
+      "epoch": 0.4542999289267946,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00011438825663542209,
+      "loss": 0.0698,
+      "step": 9588
+    },
+    {
+      "epoch": 0.454347311063729,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00011437351884303513,
+      "loss": 0.9221,
+      "step": 9589
+    },
+    {
+      "epoch": 0.45439469320066334,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00011435878073185879,
+      "loss": 0.6449,
+      "step": 9590
+    },
+    {
+      "epoch": 0.45444207533759773,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00011434404230222,
+      "loss": 1.0036,
+      "step": 9591
+    },
+    {
+      "epoch": 0.4544894574745321,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00011432930355444564,
+      "loss": 1.5581,
+      "step": 9592
+    },
+    {
+      "epoch": 0.45453683961146646,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00011431456448886257,
+      "loss": 0.0236,
+      "step": 9593
+    },
+    {
+      "epoch": 0.45458422174840085,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001142998251057977,
+      "loss": 0.1308,
+      "step": 9594
+    },
+    {
+      "epoch": 0.45463160388533524,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011428508540557799,
+      "loss": 1.1786,
+      "step": 9595
+    },
+    {
+      "epoch": 0.4546789860222696,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00011427034538853028,
+      "loss": 0.9578,
+      "step": 9596
+    },
+    {
+      "epoch": 0.454726368159204,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001142556050549815,
+      "loss": 0.8847,
+      "step": 9597
+    },
+    {
+      "epoch": 0.45477375029613837,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011424086440525856,
+      "loss": 1.0814,
+      "step": 9598
+    },
+    {
+      "epoch": 0.4548211324330727,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011422612343968844,
+      "loss": 1.1423,
+      "step": 9599
+    },
+    {
+      "epoch": 0.4548685145700071,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011421138215859806,
+      "loss": 0.0239,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4549158967069415,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011419664056231436,
+      "loss": 0.4021,
+      "step": 9601
+    },
+    {
+      "epoch": 0.4549632788438759,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011418189865116429,
+      "loss": 1.1078,
+      "step": 9602
+    },
+    {
+      "epoch": 0.4550106609808102,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011416715642547482,
+      "loss": 0.1314,
+      "step": 9603
+    },
+    {
+      "epoch": 0.4550580431177446,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011415241388557292,
+      "loss": 0.8406,
+      "step": 9604
+    },
+    {
+      "epoch": 0.455105425254679,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011413767103178553,
+      "loss": 1.1282,
+      "step": 9605
+    },
+    {
+      "epoch": 0.45515280739161335,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011412292786443965,
+      "loss": 0.4674,
+      "step": 9606
+    },
+    {
+      "epoch": 0.45520018952854774,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001141081843838623,
+      "loss": 1.5249,
+      "step": 9607
+    },
+    {
+      "epoch": 0.45524757166548213,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011409344059038043,
+      "loss": 1.1895,
+      "step": 9608
+    },
+    {
+      "epoch": 0.45529495380241647,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011407869648432105,
+      "loss": 0.8841,
+      "step": 9609
+    },
+    {
+      "epoch": 0.45534233593935086,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011406395206601119,
+      "loss": 0.8649,
+      "step": 9610
+    },
+    {
+      "epoch": 0.45538971807628525,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011404920733577786,
+      "loss": 0.3384,
+      "step": 9611
+    },
+    {
+      "epoch": 0.4554371002132196,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011403446229394809,
+      "loss": 0.8103,
+      "step": 9612
+    },
+    {
+      "epoch": 0.455484482350154,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011401971694084887,
+      "loss": 1.1253,
+      "step": 9613
+    },
+    {
+      "epoch": 0.4555318644870884,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011400497127680727,
+      "loss": 0.8887,
+      "step": 9614
+    },
+    {
+      "epoch": 0.45557924662402277,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00011399022530215033,
+      "loss": 0.0379,
+      "step": 9615
+    },
+    {
+      "epoch": 0.4556266287609571,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.00011397547901720508,
+      "loss": 0.0031,
+      "step": 9616
+    },
+    {
+      "epoch": 0.4556740108978915,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00011396073242229859,
+      "loss": 0.0252,
+      "step": 9617
+    },
+    {
+      "epoch": 0.4557213930348259,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011394598551775794,
+      "loss": 0.8843,
+      "step": 9618
+    },
+    {
+      "epoch": 0.45576877517176023,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011393123830391019,
+      "loss": 0.6872,
+      "step": 9619
+    },
+    {
+      "epoch": 0.4558161573086946,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001139164907810824,
+      "loss": 0.0144,
+      "step": 9620
+    },
+    {
+      "epoch": 0.455863539445629,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011390174294960165,
+      "loss": 1.0954,
+      "step": 9621
+    },
+    {
+      "epoch": 0.45591092158256336,
+      "grad_norm": 0.02392578125,
+      "learning_rate": 0.00011388699480979507,
+      "loss": 0.0009,
+      "step": 9622
+    },
+    {
+      "epoch": 0.45595830371949775,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011387224636198977,
+      "loss": 0.9338,
+      "step": 9623
+    },
+    {
+      "epoch": 0.45600568585643214,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00011385749760651276,
+      "loss": 0.8699,
+      "step": 9624
+    },
+    {
+      "epoch": 0.4560530679933665,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011384274854369124,
+      "loss": 1.0061,
+      "step": 9625
+    },
+    {
+      "epoch": 0.45610045013030087,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011382799917385232,
+      "loss": 0.9334,
+      "step": 9626
+    },
+    {
+      "epoch": 0.45614783226723526,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001138132494973231,
+      "loss": 1.0398,
+      "step": 9627
+    },
+    {
+      "epoch": 0.4561952144041696,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00011379849951443071,
+      "loss": 0.8752,
+      "step": 9628
+    },
+    {
+      "epoch": 0.456242596541104,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011378374922550228,
+      "loss": 0.0754,
+      "step": 9629
+    },
+    {
+      "epoch": 0.4562899786780384,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00011376899863086501,
+      "loss": 0.0124,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4563373608149728,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011375424773084599,
+      "loss": 0.8615,
+      "step": 9631
+    },
+    {
+      "epoch": 0.4563847429519071,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001137394965257724,
+      "loss": 0.9877,
+      "step": 9632
+    },
+    {
+      "epoch": 0.4564321250888415,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00011372474501597144,
+      "loss": 0.7045,
+      "step": 9633
+    },
+    {
+      "epoch": 0.4564795072257759,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00011370999320177022,
+      "loss": 0.0186,
+      "step": 9634
+    },
+    {
+      "epoch": 0.45652688936271024,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00011369524108349601,
+      "loss": 0.1419,
+      "step": 9635
+    },
+    {
+      "epoch": 0.45657427149964463,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00011368048866147589,
+      "loss": 0.9008,
+      "step": 9636
+    },
+    {
+      "epoch": 0.45662165363657903,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001136657359360371,
+      "loss": 0.913,
+      "step": 9637
+    },
+    {
+      "epoch": 0.45666903577351337,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011365098290750689,
+      "loss": 1.543,
+      "step": 9638
+    },
+    {
+      "epoch": 0.45671641791044776,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011363622957621238,
+      "loss": 0.655,
+      "step": 9639
+    },
+    {
+      "epoch": 0.45676380004738215,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00011362147594248082,
+      "loss": 0.0694,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4568111821843165,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011360672200663946,
+      "loss": 0.7387,
+      "step": 9641
+    },
+    {
+      "epoch": 0.4568585643212509,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011359196776901548,
+      "loss": 0.4316,
+      "step": 9642
+    },
+    {
+      "epoch": 0.4569059464581853,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00011357721322993615,
+      "loss": 0.5524,
+      "step": 9643
+    },
+    {
+      "epoch": 0.4569533285951196,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011356245838972868,
+      "loss": 1.012,
+      "step": 9644
+    },
+    {
+      "epoch": 0.457000710732054,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001135477032487203,
+      "loss": 0.9261,
+      "step": 9645
+    },
+    {
+      "epoch": 0.4570480928689884,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011353294780723836,
+      "loss": 0.5779,
+      "step": 9646
+    },
+    {
+      "epoch": 0.4570954750059228,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00011351819206561,
+      "loss": 0.0075,
+      "step": 9647
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011350343602416254,
+      "loss": 0.8601,
+      "step": 9648
+    },
+    {
+      "epoch": 0.4571902392797915,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011348867968322327,
+      "loss": 0.982,
+      "step": 9649
+    },
+    {
+      "epoch": 0.4572376214167259,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00011347392304311944,
+      "loss": 0.8338,
+      "step": 9650
+    },
+    {
+      "epoch": 0.45728500355366025,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011345916610417837,
+      "loss": 0.1226,
+      "step": 9651
+    },
+    {
+      "epoch": 0.45733238569059464,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011344440886672732,
+      "loss": 0.7333,
+      "step": 9652
+    },
+    {
+      "epoch": 0.45737976782752904,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00011342965133109361,
+      "loss": 0.08,
+      "step": 9653
+    },
+    {
+      "epoch": 0.4574271499644634,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00011341489349760453,
+      "loss": 1.1604,
+      "step": 9654
+    },
+    {
+      "epoch": 0.45747453210139777,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011340013536658738,
+      "loss": 0.7503,
+      "step": 9655
+    },
+    {
+      "epoch": 0.45752191423833216,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001133853769383695,
+      "loss": 0.3503,
+      "step": 9656
+    },
+    {
+      "epoch": 0.4575692963752665,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011337061821327825,
+      "loss": 0.1039,
+      "step": 9657
+    },
+    {
+      "epoch": 0.4576166785122009,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011335585919164092,
+      "loss": 0.5785,
+      "step": 9658
+    },
+    {
+      "epoch": 0.4576640606491353,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011334109987378485,
+      "loss": 1.3708,
+      "step": 9659
+    },
+    {
+      "epoch": 0.4577114427860697,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00011332634026003741,
+      "loss": 1.051,
+      "step": 9660
+    },
+    {
+      "epoch": 0.457758824923004,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011331158035072593,
+      "loss": 0.8012,
+      "step": 9661
+    },
+    {
+      "epoch": 0.4578062070599384,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00011329682014617777,
+      "loss": 0.8043,
+      "step": 9662
+    },
+    {
+      "epoch": 0.4578535891968728,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00011328205964672029,
+      "loss": 0.0104,
+      "step": 9663
+    },
+    {
+      "epoch": 0.45790097133380714,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00011326729885268088,
+      "loss": 0.0549,
+      "step": 9664
+    },
+    {
+      "epoch": 0.45794835347074153,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011325253776438695,
+      "loss": 0.7109,
+      "step": 9665
+    },
+    {
+      "epoch": 0.4579957356076759,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00011323777638216582,
+      "loss": 0.0102,
+      "step": 9666
+    },
+    {
+      "epoch": 0.45804311774461026,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0001132230147063449,
+      "loss": 0.0063,
+      "step": 9667
+    },
+    {
+      "epoch": 0.45809049988154465,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011320825273725162,
+      "loss": 0.0473,
+      "step": 9668
+    },
+    {
+      "epoch": 0.45813788201847905,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011319349047521337,
+      "loss": 1.3497,
+      "step": 9669
+    },
+    {
+      "epoch": 0.4581852641554134,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011317872792055752,
+      "loss": 1.1318,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4582326462923478,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011316396507361153,
+      "loss": 1.4103,
+      "step": 9671
+    },
+    {
+      "epoch": 0.45828002842928217,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011314920193470284,
+      "loss": 0.015,
+      "step": 9672
+    },
+    {
+      "epoch": 0.4583274105662165,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011313443850415884,
+      "loss": 1.2142,
+      "step": 9673
+    },
+    {
+      "epoch": 0.4583747927031509,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.000113119674782307,
+      "loss": 0.119,
+      "step": 9674
+    },
+    {
+      "epoch": 0.4584221748400853,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011310491076947474,
+      "loss": 0.6329,
+      "step": 9675
+    },
+    {
+      "epoch": 0.4584695569770197,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011309014646598953,
+      "loss": 0.9227,
+      "step": 9676
+    },
+    {
+      "epoch": 0.458516939113954,
+      "grad_norm": 0.001556396484375,
+      "learning_rate": 0.00011307538187217879,
+      "loss": 0.0001,
+      "step": 9677
+    },
+    {
+      "epoch": 0.4585643212508884,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011306061698837,
+      "loss": 1.0653,
+      "step": 9678
+    },
+    {
+      "epoch": 0.4586117033878228,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011304585181489065,
+      "loss": 0.3294,
+      "step": 9679
+    },
+    {
+      "epoch": 0.45865908552475715,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001130310863520682,
+      "loss": 0.9124,
+      "step": 9680
+    },
+    {
+      "epoch": 0.45870646766169154,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011301632060023015,
+      "loss": 1.1995,
+      "step": 9681
+    },
+    {
+      "epoch": 0.45875384979862593,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011300155455970396,
+      "loss": 1.0626,
+      "step": 9682
+    },
+    {
+      "epoch": 0.45880123193556027,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011298678823081714,
+      "loss": 1.5004,
+      "step": 9683
+    },
+    {
+      "epoch": 0.45884861407249466,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001129720216138972,
+      "loss": 0.0103,
+      "step": 9684
+    },
+    {
+      "epoch": 0.45889599620942906,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011295725470927163,
+      "loss": 0.5508,
+      "step": 9685
+    },
+    {
+      "epoch": 0.4589433783463634,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011294248751726795,
+      "loss": 0.1358,
+      "step": 9686
+    },
+    {
+      "epoch": 0.4589907604832978,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011292772003821366,
+      "loss": 0.8689,
+      "step": 9687
+    },
+    {
+      "epoch": 0.4590381426202322,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011291295227243634,
+      "loss": 1.5178,
+      "step": 9688
+    },
+    {
+      "epoch": 0.4590855247571666,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00011289818422026348,
+      "loss": 0.5054,
+      "step": 9689
+    },
+    {
+      "epoch": 0.4591329068941009,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00011288341588202266,
+      "loss": 0.0367,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4591802890310353,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011286864725804136,
+      "loss": 0.5464,
+      "step": 9691
+    },
+    {
+      "epoch": 0.4592276711679697,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001128538783486472,
+      "loss": 1.1001,
+      "step": 9692
+    },
+    {
+      "epoch": 0.45927505330490404,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00011283910915416771,
+      "loss": 0.1493,
+      "step": 9693
+    },
+    {
+      "epoch": 0.45932243544183843,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011282433967493042,
+      "loss": 0.6702,
+      "step": 9694
+    },
+    {
+      "epoch": 0.4593698175787728,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.00011280956991126297,
+      "loss": 0.0083,
+      "step": 9695
+    },
+    {
+      "epoch": 0.45941719971570716,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00011279479986349288,
+      "loss": 0.1184,
+      "step": 9696
+    },
+    {
+      "epoch": 0.45946458185264155,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.00011278002953194777,
+      "loss": 0.003,
+      "step": 9697
+    },
+    {
+      "epoch": 0.45951196398957594,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00011276525891695521,
+      "loss": 0.1499,
+      "step": 9698
+    },
+    {
+      "epoch": 0.4595593461265103,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00011275048801884284,
+      "loss": 0.8919,
+      "step": 9699
+    },
+    {
+      "epoch": 0.4596067282634447,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0001127357168379382,
+      "loss": 2.1475,
+      "step": 9700
+    },
+    {
+      "epoch": 0.45965411040037907,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0001127209453745689,
+      "loss": 0.0036,
+      "step": 9701
+    },
+    {
+      "epoch": 0.4597014925373134,
+      "grad_norm": 0.00628662109375,
+      "learning_rate": 0.0001127061736290626,
+      "loss": 0.0002,
+      "step": 9702
+    },
+    {
+      "epoch": 0.4597488746742478,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00011269140160174694,
+      "loss": 0.0031,
+      "step": 9703
+    },
+    {
+      "epoch": 0.4597962568111822,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011267662929294946,
+      "loss": 0.5549,
+      "step": 9704
+    },
+    {
+      "epoch": 0.4598436389481166,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011266185670299785,
+      "loss": 0.8531,
+      "step": 9705
+    },
+    {
+      "epoch": 0.4598910210850509,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00011264708383221978,
+      "loss": 0.0104,
+      "step": 9706
+    },
+    {
+      "epoch": 0.4599384032219853,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00011263231068094285,
+      "loss": 1.5082,
+      "step": 9707
+    },
+    {
+      "epoch": 0.4599857853589197,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011261753724949471,
+      "loss": 0.7843,
+      "step": 9708
+    },
+    {
+      "epoch": 0.46003316749585405,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011260276353820303,
+      "loss": 0.9899,
+      "step": 9709
+    },
+    {
+      "epoch": 0.46008054963278844,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011258798954739547,
+      "loss": 0.4392,
+      "step": 9710
+    },
+    {
+      "epoch": 0.46012793176972283,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011257321527739974,
+      "loss": 0.7733,
+      "step": 9711
+    },
+    {
+      "epoch": 0.46017531390665717,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00011255844072854347,
+      "loss": 0.9897,
+      "step": 9712
+    },
+    {
+      "epoch": 0.46022269604359156,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00011254366590115435,
+      "loss": 0.3876,
+      "step": 9713
+    },
+    {
+      "epoch": 0.46027007818052595,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011252889079556011,
+      "loss": 1.0065,
+      "step": 9714
+    },
+    {
+      "epoch": 0.4603174603174603,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011251411541208843,
+      "loss": 0.8447,
+      "step": 9715
+    },
+    {
+      "epoch": 0.4603648424543947,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011249933975106697,
+      "loss": 0.7846,
+      "step": 9716
+    },
+    {
+      "epoch": 0.4604122245913291,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00011248456381282344,
+      "loss": 1.0592,
+      "step": 9717
+    },
+    {
+      "epoch": 0.46045960672826347,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011246978759768563,
+      "loss": 1.2718,
+      "step": 9718
+    },
+    {
+      "epoch": 0.4605069888651978,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0001124550111059812,
+      "loss": 0.0379,
+      "step": 9719
+    },
+    {
+      "epoch": 0.4605543710021322,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00011244023433803788,
+      "loss": 0.1691,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4606017531390666,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00011242545729418342,
+      "loss": 0.0345,
+      "step": 9721
+    },
+    {
+      "epoch": 0.46064913527600093,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011241067997474557,
+      "loss": 1.0231,
+      "step": 9722
+    },
+    {
+      "epoch": 0.4606965174129353,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011239590238005204,
+      "loss": 1.3441,
+      "step": 9723
+    },
+    {
+      "epoch": 0.4607438995498697,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011238112451043059,
+      "loss": 1.337,
+      "step": 9724
+    },
+    {
+      "epoch": 0.46079128168680406,
+      "grad_norm": 0.033203125,
+      "learning_rate": 0.00011236634636620899,
+      "loss": 0.0008,
+      "step": 9725
+    },
+    {
+      "epoch": 0.46083866382373845,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011235156794771502,
+      "loss": 0.7237,
+      "step": 9726
+    },
+    {
+      "epoch": 0.46088604596067284,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0001123367892552764,
+      "loss": 0.0114,
+      "step": 9727
+    },
+    {
+      "epoch": 0.4609334280976072,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011232201028922093,
+      "loss": 1.2869,
+      "step": 9728
+    },
+    {
+      "epoch": 0.46098081023454157,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00011230723104987644,
+      "loss": 0.0132,
+      "step": 9729
+    },
+    {
+      "epoch": 0.46102819237147596,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00011229245153757067,
+      "loss": 0.8768,
+      "step": 9730
+    },
+    {
+      "epoch": 0.4610755745084103,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011227767175263138,
+      "loss": 2.0491,
+      "step": 9731
+    },
+    {
+      "epoch": 0.4611229566453447,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011226289169538642,
+      "loss": 0.728,
+      "step": 9732
+    },
+    {
+      "epoch": 0.4611703387822791,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 0.00011224811136616358,
+      "loss": 0.0014,
+      "step": 9733
+    },
+    {
+      "epoch": 0.4612177209192135,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011223333076529071,
+      "loss": 1.6146,
+      "step": 9734
+    },
+    {
+      "epoch": 0.4612651030561478,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00011221854989309555,
+      "loss": 0.0202,
+      "step": 9735
+    },
+    {
+      "epoch": 0.4613124851930822,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011220376874990599,
+      "loss": 0.8589,
+      "step": 9736
+    },
+    {
+      "epoch": 0.4613598673300166,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011218898733604985,
+      "loss": 1.0107,
+      "step": 9737
+    },
+    {
+      "epoch": 0.46140724946695094,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.00011217420565185495,
+      "loss": 0.004,
+      "step": 9738
+    },
+    {
+      "epoch": 0.46145463160388533,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00011215942369764912,
+      "loss": 0.7273,
+      "step": 9739
+    },
+    {
+      "epoch": 0.4615020137408197,
+      "grad_norm": 0.02587890625,
+      "learning_rate": 0.00011214464147376022,
+      "loss": 0.001,
+      "step": 9740
+    },
+    {
+      "epoch": 0.46154939587775407,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011212985898051613,
+      "loss": 0.9893,
+      "step": 9741
+    },
+    {
+      "epoch": 0.46159677801468846,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011211507621824467,
+      "loss": 0.6534,
+      "step": 9742
+    },
+    {
+      "epoch": 0.46164416015162285,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00011210029318727374,
+      "loss": 1.0455,
+      "step": 9743
+    },
+    {
+      "epoch": 0.4616915422885572,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011208550988793116,
+      "loss": 1.008,
+      "step": 9744
+    },
+    {
+      "epoch": 0.4617389244254916,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.00011207072632054489,
+      "loss": 0.0047,
+      "step": 9745
+    },
+    {
+      "epoch": 0.461786306562426,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011205594248544275,
+      "loss": 0.013,
+      "step": 9746
+    },
+    {
+      "epoch": 0.46183368869936037,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00011204115838295263,
+      "loss": 0.021,
+      "step": 9747
+    },
+    {
+      "epoch": 0.4618810708362947,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011202637401340246,
+      "loss": 0.5427,
+      "step": 9748
+    },
+    {
+      "epoch": 0.4619284529732291,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011201158937712013,
+      "loss": 0.1054,
+      "step": 9749
+    },
+    {
+      "epoch": 0.4619758351101635,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00011199680447443352,
+      "loss": 0.0799,
+      "step": 9750
+    },
+    {
+      "epoch": 0.46202321724709783,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011198201930567059,
+      "loss": 0.6151,
+      "step": 9751
+    },
+    {
+      "epoch": 0.4620705993840322,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00011196723387115922,
+      "loss": 1.362,
+      "step": 9752
+    },
+    {
+      "epoch": 0.4621179815209666,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00011195244817122736,
+      "loss": 0.6715,
+      "step": 9753
+    },
+    {
+      "epoch": 0.46216536365790095,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011193766220620291,
+      "loss": 1.1283,
+      "step": 9754
+    },
+    {
+      "epoch": 0.46221274579483534,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011192287597641385,
+      "loss": 1.0834,
+      "step": 9755
+    },
+    {
+      "epoch": 0.46226012793176974,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00011190808948218807,
+      "loss": 0.0404,
+      "step": 9756
+    },
+    {
+      "epoch": 0.4623075100687041,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00011189330272385359,
+      "loss": 0.5915,
+      "step": 9757
+    },
+    {
+      "epoch": 0.46235489220563847,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00011187851570173831,
+      "loss": 0.1652,
+      "step": 9758
+    },
+    {
+      "epoch": 0.46240227434257286,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011186372841617019,
+      "loss": 0.0703,
+      "step": 9759
+    },
+    {
+      "epoch": 0.4624496564795072,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00011184894086747722,
+      "loss": 1.0824,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4624970386164416,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011183415305598737,
+      "loss": 0.9035,
+      "step": 9761
+    },
+    {
+      "epoch": 0.462544420753376,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001118193649820286,
+      "loss": 1.2383,
+      "step": 9762
+    },
+    {
+      "epoch": 0.4625918028903104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001118045766459289,
+      "loss": 1.1803,
+      "step": 9763
+    },
+    {
+      "epoch": 0.4626391850272447,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011178978804801627,
+      "loss": 0.7214,
+      "step": 9764
+    },
+    {
+      "epoch": 0.4626865671641791,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011177499918861869,
+      "loss": 0.0074,
+      "step": 9765
+    },
+    {
+      "epoch": 0.4627339493011135,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011176021006806418,
+      "loss": 0.9706,
+      "step": 9766
+    },
+    {
+      "epoch": 0.46278133143804784,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00011174542068668073,
+      "loss": 1.4274,
+      "step": 9767
+    },
+    {
+      "epoch": 0.46282871357498223,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011173063104479638,
+      "loss": 1.291,
+      "step": 9768
+    },
+    {
+      "epoch": 0.4628760957119166,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011171584114273912,
+      "loss": 0.6011,
+      "step": 9769
+    },
+    {
+      "epoch": 0.46292347784885096,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00011170105098083696,
+      "loss": 1.3595,
+      "step": 9770
+    },
+    {
+      "epoch": 0.46297085998578535,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011168626055941798,
+      "loss": 0.6471,
+      "step": 9771
+    },
+    {
+      "epoch": 0.46301824212271975,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011167146987881019,
+      "loss": 0.9433,
+      "step": 9772
+    },
+    {
+      "epoch": 0.4630656242596541,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001116566789393416,
+      "loss": 0.0392,
+      "step": 9773
+    },
+    {
+      "epoch": 0.4631130063965885,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001116418877413403,
+      "loss": 0.1455,
+      "step": 9774
+    },
+    {
+      "epoch": 0.46316038853352287,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011162709628513435,
+      "loss": 0.5217,
+      "step": 9775
+    },
+    {
+      "epoch": 0.46320777067045726,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00011161230457105176,
+      "loss": 0.6512,
+      "step": 9776
+    },
+    {
+      "epoch": 0.4632551528073916,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011159751259942066,
+      "loss": 1.0447,
+      "step": 9777
+    },
+    {
+      "epoch": 0.463302534944326,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011158272037056905,
+      "loss": 0.8032,
+      "step": 9778
+    },
+    {
+      "epoch": 0.4633499170812604,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011156792788482504,
+      "loss": 1.1034,
+      "step": 9779
+    },
+    {
+      "epoch": 0.4633972992181947,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011155313514251673,
+      "loss": 1.1353,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4634446813551291,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00011153834214397219,
+      "loss": 0.8526,
+      "step": 9781
+    },
+    {
+      "epoch": 0.4634920634920635,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011152354888951948,
+      "loss": 0.4919,
+      "step": 9782
+    },
+    {
+      "epoch": 0.46353944562899785,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011150875537948677,
+      "loss": 1.0076,
+      "step": 9783
+    },
+    {
+      "epoch": 0.46358682776593224,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00011149396161420211,
+      "loss": 1.3346,
+      "step": 9784
+    },
+    {
+      "epoch": 0.46363420990286663,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00011147916759399362,
+      "loss": 1.3384,
+      "step": 9785
+    },
+    {
+      "epoch": 0.46368159203980097,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011146437331918939,
+      "loss": 1.0324,
+      "step": 9786
+    },
+    {
+      "epoch": 0.46372897417673536,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011144957879011764,
+      "loss": 0.759,
+      "step": 9787
+    },
+    {
+      "epoch": 0.46377635631366976,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00011143478400710636,
+      "loss": 1.0396,
+      "step": 9788
+    },
+    {
+      "epoch": 0.4638237384506041,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011141998897048378,
+      "loss": 0.1246,
+      "step": 9789
+    },
+    {
+      "epoch": 0.4638711205875385,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.000111405193680578,
+      "loss": 0.8911,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4639185027244729,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011139039813771719,
+      "loss": 1.1938,
+      "step": 9791
+    },
+    {
+      "epoch": 0.4639658848614073,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001113756023422295,
+      "loss": 0.6653,
+      "step": 9792
+    },
+    {
+      "epoch": 0.4640132669983416,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00011136080629444303,
+      "loss": 0.0119,
+      "step": 9793
+    },
+    {
+      "epoch": 0.464060649135276,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00011134600999468596,
+      "loss": 0.8786,
+      "step": 9794
+    },
+    {
+      "epoch": 0.4641080312722104,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00011133121344328652,
+      "loss": 0.7408,
+      "step": 9795
+    },
+    {
+      "epoch": 0.46415541340914473,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011131641664057282,
+      "loss": 0.0056,
+      "step": 9796
+    },
+    {
+      "epoch": 0.4642027955460791,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00011130161958687304,
+      "loss": 1.5517,
+      "step": 9797
+    },
+    {
+      "epoch": 0.4642501776830135,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011128682228251538,
+      "loss": 1.2288,
+      "step": 9798
+    },
+    {
+      "epoch": 0.46429755981994786,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011127202472782802,
+      "loss": 1.3046,
+      "step": 9799
+    },
+    {
+      "epoch": 0.46434494195688225,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011125722692313918,
+      "loss": 0.8764,
+      "step": 9800
+    },
+    {
+      "epoch": 0.46439232409381664,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011124242886877703,
+      "loss": 0.7067,
+      "step": 9801
+    },
+    {
+      "epoch": 0.464439706230751,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011122763056506975,
+      "loss": 0.0462,
+      "step": 9802
+    },
+    {
+      "epoch": 0.4644870883676854,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011121283201234562,
+      "loss": 1.4981,
+      "step": 9803
+    },
+    {
+      "epoch": 0.46453447050461977,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011119803321093281,
+      "loss": 1.1497,
+      "step": 9804
+    },
+    {
+      "epoch": 0.46458185264155416,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011118323416115955,
+      "loss": 0.4918,
+      "step": 9805
+    },
+    {
+      "epoch": 0.4646292347784885,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011116843486335407,
+      "loss": 0.8862,
+      "step": 9806
+    },
+    {
+      "epoch": 0.4646766169154229,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001111536353178446,
+      "loss": 1.2946,
+      "step": 9807
+    },
+    {
+      "epoch": 0.4647239990523573,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011113883552495938,
+      "loss": 1.6293,
+      "step": 9808
+    },
+    {
+      "epoch": 0.4647713811892916,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011112403548502664,
+      "loss": 0.7951,
+      "step": 9809
+    },
+    {
+      "epoch": 0.464818763326226,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011110923519837466,
+      "loss": 0.8851,
+      "step": 9810
+    },
+    {
+      "epoch": 0.4648661454631604,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00011109443466533167,
+      "loss": 0.0103,
+      "step": 9811
+    },
+    {
+      "epoch": 0.46491352760009474,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00011107963388622594,
+      "loss": 0.8671,
+      "step": 9812
+    },
+    {
+      "epoch": 0.46496090973702914,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001110648328613857,
+      "loss": 0.6951,
+      "step": 9813
+    },
+    {
+      "epoch": 0.46500829187396353,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001110500315911393,
+      "loss": 1.1671,
+      "step": 9814
+    },
+    {
+      "epoch": 0.46505567401089787,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011103523007581494,
+      "loss": 1.1655,
+      "step": 9815
+    },
+    {
+      "epoch": 0.46510305614783226,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00011102042831574095,
+      "loss": 0.0534,
+      "step": 9816
+    },
+    {
+      "epoch": 0.46515043828476665,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011100562631124558,
+      "loss": 1.0242,
+      "step": 9817
+    },
+    {
+      "epoch": 0.465197820421701,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011099082406265715,
+      "loss": 0.5828,
+      "step": 9818
+    },
+    {
+      "epoch": 0.4652452025586354,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011097602157030394,
+      "loss": 1.1612,
+      "step": 9819
+    },
+    {
+      "epoch": 0.4652925846955698,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011096121883451427,
+      "loss": 0.5682,
+      "step": 9820
+    },
+    {
+      "epoch": 0.46533996683250417,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011094641585561645,
+      "loss": 0.653,
+      "step": 9821
+    },
+    {
+      "epoch": 0.4653873489694385,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011093161263393876,
+      "loss": 1.2252,
+      "step": 9822
+    },
+    {
+      "epoch": 0.4654347311063729,
+      "grad_norm": 0.027099609375,
+      "learning_rate": 0.00011091680916980957,
+      "loss": 0.0018,
+      "step": 9823
+    },
+    {
+      "epoch": 0.4654821132433073,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00011090200546355718,
+      "loss": 0.9737,
+      "step": 9824
+    },
+    {
+      "epoch": 0.46552949538024163,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011088720151550991,
+      "loss": 0.8436,
+      "step": 9825
+    },
+    {
+      "epoch": 0.465576877517176,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00011087239732599612,
+      "loss": 0.0059,
+      "step": 9826
+    },
+    {
+      "epoch": 0.4656242596541104,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00011085759289534412,
+      "loss": 0.0125,
+      "step": 9827
+    },
+    {
+      "epoch": 0.46567164179104475,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011084278822388228,
+      "loss": 1.3405,
+      "step": 9828
+    },
+    {
+      "epoch": 0.46571902392797915,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00011082798331193898,
+      "loss": 1.1513,
+      "step": 9829
+    },
+    {
+      "epoch": 0.46576640606491354,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0001108131781598425,
+      "loss": 1.184,
+      "step": 9830
+    },
+    {
+      "epoch": 0.4658137882018479,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011079837276792125,
+      "loss": 0.7758,
+      "step": 9831
+    },
+    {
+      "epoch": 0.46586117033878227,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00011078356713650361,
+      "loss": 0.0137,
+      "step": 9832
+    },
+    {
+      "epoch": 0.46590855247571666,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00011076876126591796,
+      "loss": 1.1446,
+      "step": 9833
+    },
+    {
+      "epoch": 0.46595593461265106,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00011075395515649261,
+      "loss": 0.0185,
+      "step": 9834
+    },
+    {
+      "epoch": 0.4660033167495854,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.000110739148808556,
+      "loss": 0.614,
+      "step": 9835
+    },
+    {
+      "epoch": 0.4660506988865198,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00011072434222243651,
+      "loss": 0.7432,
+      "step": 9836
+    },
+    {
+      "epoch": 0.4660980810234542,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011070953539846254,
+      "loss": 0.9163,
+      "step": 9837
+    },
+    {
+      "epoch": 0.4661454631603885,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011069472833696249,
+      "loss": 0.8119,
+      "step": 9838
+    },
+    {
+      "epoch": 0.4661928452973229,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00011067992103826474,
+      "loss": 0.0054,
+      "step": 9839
+    },
+    {
+      "epoch": 0.4662402274342573,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011066511350269773,
+      "loss": 0.6951,
+      "step": 9840
+    },
+    {
+      "epoch": 0.46628760957119164,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011065030573058987,
+      "loss": 0.88,
+      "step": 9841
+    },
+    {
+      "epoch": 0.46633499170812603,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011063549772226955,
+      "loss": 1.2468,
+      "step": 9842
+    },
+    {
+      "epoch": 0.4663823738450604,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011062068947806521,
+      "loss": 0.1288,
+      "step": 9843
+    },
+    {
+      "epoch": 0.46642975598199476,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011060588099830533,
+      "loss": 1.4737,
+      "step": 9844
+    },
+    {
+      "epoch": 0.46647713811892916,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00011059107228331829,
+      "loss": 1.4271,
+      "step": 9845
+    },
+    {
+      "epoch": 0.46652452025586355,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011057626333343251,
+      "loss": 0.6395,
+      "step": 9846
+    },
+    {
+      "epoch": 0.4665719023927979,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011056145414897654,
+      "loss": 0.7846,
+      "step": 9847
+    },
+    {
+      "epoch": 0.4666192845297323,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00011054664473027871,
+      "loss": 0.8299,
+      "step": 9848
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011053183507766758,
+      "loss": 0.9766,
+      "step": 9849
+    },
+    {
+      "epoch": 0.46671404880360107,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011051702519147153,
+      "loss": 0.1263,
+      "step": 9850
+    },
+    {
+      "epoch": 0.4667614309405354,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00011050221507201908,
+      "loss": 0.0947,
+      "step": 9851
+    },
+    {
+      "epoch": 0.4668088130774698,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011048740471963868,
+      "loss": 0.1709,
+      "step": 9852
+    },
+    {
+      "epoch": 0.4668561952144042,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011047259413465882,
+      "loss": 0.7236,
+      "step": 9853
+    },
+    {
+      "epoch": 0.46690357735133853,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011045778331740797,
+      "loss": 1.0616,
+      "step": 9854
+    },
+    {
+      "epoch": 0.4669509594882729,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011044297226821463,
+      "loss": 1.0608,
+      "step": 9855
+    },
+    {
+      "epoch": 0.4669983416252073,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011042816098740732,
+      "loss": 0.7149,
+      "step": 9856
+    },
+    {
+      "epoch": 0.46704572376214165,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011041334947531445,
+      "loss": 1.1244,
+      "step": 9857
+    },
+    {
+      "epoch": 0.46709310589907604,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011039853773226461,
+      "loss": 0.0236,
+      "step": 9858
+    },
+    {
+      "epoch": 0.46714048803601044,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011038372575858625,
+      "loss": 0.9527,
+      "step": 9859
+    },
+    {
+      "epoch": 0.4671878701729448,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00011036891355460795,
+      "loss": 0.8516,
+      "step": 9860
+    },
+    {
+      "epoch": 0.46723525230987917,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011035410112065819,
+      "loss": 0.7951,
+      "step": 9861
+    },
+    {
+      "epoch": 0.46728263444681356,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011033928845706545,
+      "loss": 0.8273,
+      "step": 9862
+    },
+    {
+      "epoch": 0.46733001658374795,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00011032447556415838,
+      "loss": 1.5249,
+      "step": 9863
+    },
+    {
+      "epoch": 0.4673773987206823,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001103096624422654,
+      "loss": 1.0421,
+      "step": 9864
+    },
+    {
+      "epoch": 0.4674247808576167,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011029484909171508,
+      "loss": 1.0365,
+      "step": 9865
+    },
+    {
+      "epoch": 0.4674721629945511,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011028003551283597,
+      "loss": 1.1412,
+      "step": 9866
+    },
+    {
+      "epoch": 0.4675195451314854,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00011026522170595663,
+      "loss": 1.4166,
+      "step": 9867
+    },
+    {
+      "epoch": 0.4675669272684198,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011025040767140562,
+      "loss": 0.9339,
+      "step": 9868
+    },
+    {
+      "epoch": 0.4676143094053542,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00011023559340951146,
+      "loss": 0.0557,
+      "step": 9869
+    },
+    {
+      "epoch": 0.46766169154228854,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00011022077892060274,
+      "loss": 0.7653,
+      "step": 9870
+    },
+    {
+      "epoch": 0.46770907367922293,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011020596420500807,
+      "loss": 0.6284,
+      "step": 9871
+    },
+    {
+      "epoch": 0.4677564558161573,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011019114926305597,
+      "loss": 0.6233,
+      "step": 9872
+    },
+    {
+      "epoch": 0.46780383795309166,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011017633409507502,
+      "loss": 0.7435,
+      "step": 9873
+    },
+    {
+      "epoch": 0.46785122009002605,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001101615187013938,
+      "loss": 1.2383,
+      "step": 9874
+    },
+    {
+      "epoch": 0.46789860222696045,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00011014670308234096,
+      "loss": 1.1249,
+      "step": 9875
+    },
+    {
+      "epoch": 0.4679459843638948,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011013188723824504,
+      "loss": 0.1879,
+      "step": 9876
+    },
+    {
+      "epoch": 0.4679933665008292,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00011011707116943463,
+      "loss": 1.1596,
+      "step": 9877
+    },
+    {
+      "epoch": 0.46804074863776357,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011010225487623837,
+      "loss": 0.6584,
+      "step": 9878
+    },
+    {
+      "epoch": 0.46808813077469796,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001100874383589849,
+      "loss": 1.255,
+      "step": 9879
+    },
+    {
+      "epoch": 0.4681355129116323,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00011007262161800276,
+      "loss": 0.9886,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4681828950485667,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.00011005780465362057,
+      "loss": 0.0037,
+      "step": 9881
+    },
+    {
+      "epoch": 0.4682302771855011,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011004298746616701,
+      "loss": 1.0849,
+      "step": 9882
+    },
+    {
+      "epoch": 0.4682776593224354,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001100281700559707,
+      "loss": 0.7927,
+      "step": 9883
+    },
+    {
+      "epoch": 0.4683250414593698,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011001335242336023,
+      "loss": 0.0405,
+      "step": 9884
+    },
+    {
+      "epoch": 0.4683724235963042,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010999853456866429,
+      "loss": 0.0701,
+      "step": 9885
+    },
+    {
+      "epoch": 0.46841980573323855,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001099837164922115,
+      "loss": 1.1635,
+      "step": 9886
+    },
+    {
+      "epoch": 0.46846718787017294,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010996889819433053,
+      "loss": 1.3223,
+      "step": 9887
+    },
+    {
+      "epoch": 0.46851457000710733,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010995407967535,
+      "loss": 0.6921,
+      "step": 9888
+    },
+    {
+      "epoch": 0.46856195214404167,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00010993926093559859,
+      "loss": 0.0444,
+      "step": 9889
+    },
+    {
+      "epoch": 0.46860933428097606,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00010992444197540495,
+      "loss": 0.0119,
+      "step": 9890
+    },
+    {
+      "epoch": 0.46865671641791046,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00010990962279509775,
+      "loss": 0.0671,
+      "step": 9891
+    },
+    {
+      "epoch": 0.46870409855484485,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010989480339500569,
+      "loss": 1.021,
+      "step": 9892
+    },
+    {
+      "epoch": 0.4687514806917792,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001098799837754574,
+      "loss": 0.3846,
+      "step": 9893
+    },
+    {
+      "epoch": 0.4687988628287136,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010986516393678164,
+      "loss": 1.0384,
+      "step": 9894
+    },
+    {
+      "epoch": 0.468846244965648,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00010985034387930704,
+      "loss": 0.3264,
+      "step": 9895
+    },
+    {
+      "epoch": 0.4688936271025823,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0001098355236033623,
+      "loss": 0.0473,
+      "step": 9896
+    },
+    {
+      "epoch": 0.4689410092395167,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001098207031092761,
+      "loss": 1.0076,
+      "step": 9897
+    },
+    {
+      "epoch": 0.4689883913764511,
+      "grad_norm": 0.030517578125,
+      "learning_rate": 0.0001098058823973772,
+      "loss": 0.0017,
+      "step": 9898
+    },
+    {
+      "epoch": 0.46903577351338543,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00010979106146799425,
+      "loss": 0.8646,
+      "step": 9899
+    },
+    {
+      "epoch": 0.4690831556503198,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010977624032145597,
+      "loss": 1.15,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4691305377872542,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010976141895809111,
+      "loss": 0.8969,
+      "step": 9901
+    },
+    {
+      "epoch": 0.46917791992418856,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00010974659737822842,
+      "loss": 0.3929,
+      "step": 9902
+    },
+    {
+      "epoch": 0.46922530206112295,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00010973177558219651,
+      "loss": 1.0758,
+      "step": 9903
+    },
+    {
+      "epoch": 0.46927268419805734,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010971695357032423,
+      "loss": 0.696,
+      "step": 9904
+    },
+    {
+      "epoch": 0.4693200663349917,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010970213134294023,
+      "loss": 0.1151,
+      "step": 9905
+    },
+    {
+      "epoch": 0.4693674484719261,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010968730890037333,
+      "loss": 0.7491,
+      "step": 9906
+    },
+    {
+      "epoch": 0.46941483060886047,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010967248624295221,
+      "loss": 0.4829,
+      "step": 9907
+    },
+    {
+      "epoch": 0.46946221274579486,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010965766337100567,
+      "loss": 1.2206,
+      "step": 9908
+    },
+    {
+      "epoch": 0.4695095948827292,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010964284028486245,
+      "loss": 0.8439,
+      "step": 9909
+    },
+    {
+      "epoch": 0.4695569770196636,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00010962801698485128,
+      "loss": 0.3366,
+      "step": 9910
+    },
+    {
+      "epoch": 0.469604359156598,
+      "grad_norm": 0.00140380859375,
+      "learning_rate": 0.00010961319347130095,
+      "loss": 0.0001,
+      "step": 9911
+    },
+    {
+      "epoch": 0.4696517412935323,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010959836974454023,
+      "loss": 0.5039,
+      "step": 9912
+    },
+    {
+      "epoch": 0.4696991234304667,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010958354580489791,
+      "loss": 0.8668,
+      "step": 9913
+    },
+    {
+      "epoch": 0.4697465055674011,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.00010956872165270273,
+      "loss": 0.0019,
+      "step": 9914
+    },
+    {
+      "epoch": 0.46979388770433544,
+      "grad_norm": 0.004180908203125,
+      "learning_rate": 0.0001095538972882835,
+      "loss": 0.0002,
+      "step": 9915
+    },
+    {
+      "epoch": 0.46984126984126984,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000109539072711969,
+      "loss": 0.6317,
+      "step": 9916
+    },
+    {
+      "epoch": 0.46988865197820423,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010952424792408804,
+      "loss": 0.9663,
+      "step": 9917
+    },
+    {
+      "epoch": 0.46993603411513857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00010950942292496942,
+      "loss": 0.026,
+      "step": 9918
+    },
+    {
+      "epoch": 0.46998341625207296,
+      "grad_norm": 0.0208740234375,
+      "learning_rate": 0.0001094945977149419,
+      "loss": 0.0009,
+      "step": 9919
+    },
+    {
+      "epoch": 0.47003079838900735,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010947977229433433,
+      "loss": 1.3223,
+      "step": 9920
+    },
+    {
+      "epoch": 0.47007818052594175,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010946494666347551,
+      "loss": 0.8865,
+      "step": 9921
+    },
+    {
+      "epoch": 0.4701255626628761,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010945012082269423,
+      "loss": 1.0039,
+      "step": 9922
+    },
+    {
+      "epoch": 0.4701729447998105,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010943529477231936,
+      "loss": 1.3037,
+      "step": 9923
+    },
+    {
+      "epoch": 0.47022032693674487,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00010942046851267968,
+      "loss": 0.0193,
+      "step": 9924
+    },
+    {
+      "epoch": 0.4702677090736792,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010940564204410408,
+      "loss": 0.7356,
+      "step": 9925
+    },
+    {
+      "epoch": 0.4703150912106136,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010939081536692135,
+      "loss": 0.8879,
+      "step": 9926
+    },
+    {
+      "epoch": 0.470362473347548,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010937598848146032,
+      "loss": 0.0262,
+      "step": 9927
+    },
+    {
+      "epoch": 0.47040985548448233,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010936116138804985,
+      "loss": 0.7974,
+      "step": 9928
+    },
+    {
+      "epoch": 0.4704572376214167,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010934633408701883,
+      "loss": 0.715,
+      "step": 9929
+    },
+    {
+      "epoch": 0.4705046197583511,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00010933150657869602,
+      "loss": 1.0631,
+      "step": 9930
+    },
+    {
+      "epoch": 0.47055200189528545,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010931667886341035,
+      "loss": 0.0301,
+      "step": 9931
+    },
+    {
+      "epoch": 0.47059938403221985,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010930185094149068,
+      "loss": 0.9386,
+      "step": 9932
+    },
+    {
+      "epoch": 0.47064676616915424,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010928702281326586,
+      "loss": 1.4621,
+      "step": 9933
+    },
+    {
+      "epoch": 0.4706941483060886,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010927219447906478,
+      "loss": 0.9693,
+      "step": 9934
+    },
+    {
+      "epoch": 0.47074153044302297,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010925736593921627,
+      "loss": 0.1278,
+      "step": 9935
+    },
+    {
+      "epoch": 0.47078891257995736,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00010924253719404929,
+      "loss": 0.1093,
+      "step": 9936
+    },
+    {
+      "epoch": 0.47083629471689176,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010922770824389264,
+      "loss": 0.8521,
+      "step": 9937
+    },
+    {
+      "epoch": 0.4708836768538261,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010921287908907525,
+      "loss": 0.1137,
+      "step": 9938
+    },
+    {
+      "epoch": 0.4709310589907605,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000109198049729926,
+      "loss": 0.0876,
+      "step": 9939
+    },
+    {
+      "epoch": 0.4709784411276949,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010918322016677385,
+      "loss": 1.2775,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4710258232646292,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010916839039994766,
+      "loss": 0.374,
+      "step": 9941
+    },
+    {
+      "epoch": 0.4710732054015636,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00010915356042977632,
+      "loss": 0.096,
+      "step": 9942
+    },
+    {
+      "epoch": 0.471120587538498,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00010913873025658874,
+      "loss": 1.0659,
+      "step": 9943
+    },
+    {
+      "epoch": 0.47116796967543234,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010912389988071388,
+      "loss": 0.206,
+      "step": 9944
+    },
+    {
+      "epoch": 0.47121535181236673,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010910906930248061,
+      "loss": 0.7639,
+      "step": 9945
+    },
+    {
+      "epoch": 0.4712627339493011,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00010909423852221792,
+      "loss": 0.8366,
+      "step": 9946
+    },
+    {
+      "epoch": 0.47131011608623546,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00010907940754025468,
+      "loss": 0.0766,
+      "step": 9947
+    },
+    {
+      "epoch": 0.47135749822316986,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010906457635691987,
+      "loss": 0.1732,
+      "step": 9948
+    },
+    {
+      "epoch": 0.47140488036010425,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010904974497254241,
+      "loss": 0.0058,
+      "step": 9949
+    },
+    {
+      "epoch": 0.47145226249703864,
+      "grad_norm": 0.0019683837890625,
+      "learning_rate": 0.00010903491338745124,
+      "loss": 0.0001,
+      "step": 9950
+    },
+    {
+      "epoch": 0.471499644633973,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001090200816019753,
+      "loss": 0.5352,
+      "step": 9951
+    },
+    {
+      "epoch": 0.4715470267709074,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010900524961644361,
+      "loss": 0.7032,
+      "step": 9952
+    },
+    {
+      "epoch": 0.47159440890784177,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010899041743118501,
+      "loss": 0.8182,
+      "step": 9953
+    },
+    {
+      "epoch": 0.4716417910447761,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00010897558504652856,
+      "loss": 1.6718,
+      "step": 9954
+    },
+    {
+      "epoch": 0.4716891731817105,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0001089607524628032,
+      "loss": 0.0083,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4717365553186449,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00010894591968033787,
+      "loss": 0.019,
+      "step": 9956
+    },
+    {
+      "epoch": 0.4717839374555792,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010893108669946162,
+      "loss": 0.0577,
+      "step": 9957
+    },
+    {
+      "epoch": 0.4718313195925136,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010891625352050332,
+      "loss": 0.9744,
+      "step": 9958
+    },
+    {
+      "epoch": 0.471878701729448,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00010890142014379205,
+      "loss": 1.0241,
+      "step": 9959
+    },
+    {
+      "epoch": 0.47192608386638235,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010888658656965675,
+      "loss": 0.6482,
+      "step": 9960
+    },
+    {
+      "epoch": 0.47197346600331674,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00010887175279842643,
+      "loss": 0.1113,
+      "step": 9961
+    },
+    {
+      "epoch": 0.47202084814025114,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010885691883043008,
+      "loss": 1.0869,
+      "step": 9962
+    },
+    {
+      "epoch": 0.4720682302771855,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0001088420846659967,
+      "loss": 0.9253,
+      "step": 9963
+    },
+    {
+      "epoch": 0.47211561241411987,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00010882725030545531,
+      "loss": 0.3006,
+      "step": 9964
+    },
+    {
+      "epoch": 0.47216299455105426,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010881241574913492,
+      "loss": 0.8257,
+      "step": 9965
+    },
+    {
+      "epoch": 0.47221037668798865,
+      "grad_norm": 0.000728607177734375,
+      "learning_rate": 0.00010879758099736453,
+      "loss": 0.0001,
+      "step": 9966
+    },
+    {
+      "epoch": 0.472257758824923,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010878274605047317,
+      "loss": 0.0298,
+      "step": 9967
+    },
+    {
+      "epoch": 0.4723051409618574,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00010876791090878981,
+      "loss": 0.008,
+      "step": 9968
+    },
+    {
+      "epoch": 0.4723525230987918,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00010875307557264356,
+      "loss": 0.4073,
+      "step": 9969
+    },
+    {
+      "epoch": 0.4723999052357261,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00010873824004236342,
+      "loss": 1.0676,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4724472873726605,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010872340431827841,
+      "loss": 1.1719,
+      "step": 9971
+    },
+    {
+      "epoch": 0.4724946695095949,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001087085684007176,
+      "loss": 1.0475,
+      "step": 9972
+    },
+    {
+      "epoch": 0.47254205164652924,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00010869373229001001,
+      "loss": 1.3784,
+      "step": 9973
+    },
+    {
+      "epoch": 0.47258943378346363,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001086788959864847,
+      "loss": 0.148,
+      "step": 9974
+    },
+    {
+      "epoch": 0.472636815920398,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010866405949047074,
+      "loss": 1.0741,
+      "step": 9975
+    },
+    {
+      "epoch": 0.47268419805733236,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010864922280229714,
+      "loss": 0.9483,
+      "step": 9976
+    },
+    {
+      "epoch": 0.47273158019426675,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010863438592229299,
+      "loss": 1.1607,
+      "step": 9977
+    },
+    {
+      "epoch": 0.47277896233120115,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00010861954885078738,
+      "loss": 1.3892,
+      "step": 9978
+    },
+    {
+      "epoch": 0.47282634446813554,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00010860471158810934,
+      "loss": 1.0734,
+      "step": 9979
+    },
+    {
+      "epoch": 0.4728737266050699,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010858987413458797,
+      "loss": 1.1007,
+      "step": 9980
+    },
+    {
+      "epoch": 0.47292110874200427,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00010857503649055234,
+      "loss": 1.2357,
+      "step": 9981
+    },
+    {
+      "epoch": 0.47296849087893866,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010856019865633156,
+      "loss": 0.1384,
+      "step": 9982
+    },
+    {
+      "epoch": 0.473015873015873,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010854536063225465,
+      "loss": 0.9593,
+      "step": 9983
+    },
+    {
+      "epoch": 0.4730632551528074,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010853052241865076,
+      "loss": 1.2983,
+      "step": 9984
+    },
+    {
+      "epoch": 0.4731106372897418,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010851568401584895,
+      "loss": 0.8646,
+      "step": 9985
+    },
+    {
+      "epoch": 0.4731580194266761,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010850084542417837,
+      "loss": 1.2175,
+      "step": 9986
+    },
+    {
+      "epoch": 0.4732054015636105,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010848600664396807,
+      "loss": 1.2886,
+      "step": 9987
+    },
+    {
+      "epoch": 0.4732527837005449,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010847116767554719,
+      "loss": 0.8181,
+      "step": 9988
+    },
+    {
+      "epoch": 0.47330016583747925,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00010845632851924485,
+      "loss": 1.0204,
+      "step": 9989
+    },
+    {
+      "epoch": 0.47334754797441364,
+      "grad_norm": 0.01513671875,
+      "learning_rate": 0.00010844148917539014,
+      "loss": 0.0006,
+      "step": 9990
+    },
+    {
+      "epoch": 0.47339493011134803,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010842664964431219,
+      "loss": 0.1346,
+      "step": 9991
+    },
+    {
+      "epoch": 0.47344231224828237,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001084118099263401,
+      "loss": 0.9228,
+      "step": 9992
+    },
+    {
+      "epoch": 0.47348969438521676,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010839697002180305,
+      "loss": 0.9449,
+      "step": 9993
+    },
+    {
+      "epoch": 0.47353707652215116,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00010838212993103016,
+      "loss": 0.7914,
+      "step": 9994
+    },
+    {
+      "epoch": 0.47358445865908555,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00010836728965435054,
+      "loss": 0.0168,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4736318407960199,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010835244919209337,
+      "loss": 0.7113,
+      "step": 9996
+    },
+    {
+      "epoch": 0.4736792229329543,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010833760854458774,
+      "loss": 0.6393,
+      "step": 9997
+    },
+    {
+      "epoch": 0.4737266050698887,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00010832276771216288,
+      "loss": 0.2326,
+      "step": 9998
+    },
+    {
+      "epoch": 0.473773987206823,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00010830792669514784,
+      "loss": 1.2877,
+      "step": 9999
+    },
+    {
+      "epoch": 0.4738213693437574,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010829308549387187,
+      "loss": 0.5168,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4738687514806918,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010827824410866409,
+      "loss": 0.8626,
+      "step": 10001
+    },
+    {
+      "epoch": 0.47391613361762613,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010826340253985368,
+      "loss": 0.9617,
+      "step": 10002
+    },
+    {
+      "epoch": 0.4739635157545605,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010824856078776979,
+      "loss": 0.7941,
+      "step": 10003
+    },
+    {
+      "epoch": 0.4740108978914949,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010823371885274163,
+      "loss": 1.194,
+      "step": 10004
+    },
+    {
+      "epoch": 0.47405828002842926,
+      "grad_norm": 0.007568359375,
+      "learning_rate": 0.00010821887673509834,
+      "loss": 0.0004,
+      "step": 10005
+    },
+    {
+      "epoch": 0.47410566216536365,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010820403443516911,
+      "loss": 1.1485,
+      "step": 10006
+    },
+    {
+      "epoch": 0.47415304430229804,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010818919195328315,
+      "loss": 0.0852,
+      "step": 10007
+    },
+    {
+      "epoch": 0.47420042643923244,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010817434928976962,
+      "loss": 0.7768,
+      "step": 10008
+    },
+    {
+      "epoch": 0.4742478085761668,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010815950644495774,
+      "loss": 0.8561,
+      "step": 10009
+    },
+    {
+      "epoch": 0.47429519071310117,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00010814466341917669,
+      "loss": 0.0363,
+      "step": 10010
+    },
+    {
+      "epoch": 0.47434257285003556,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010812982021275569,
+      "loss": 0.5881,
+      "step": 10011
+    },
+    {
+      "epoch": 0.4743899549869699,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00010811497682602393,
+      "loss": 0.626,
+      "step": 10012
+    },
+    {
+      "epoch": 0.4744373371239043,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010810013325931065,
+      "loss": 0.7399,
+      "step": 10013
+    },
+    {
+      "epoch": 0.4744847192608387,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010808528951294504,
+      "loss": 0.7295,
+      "step": 10014
+    },
+    {
+      "epoch": 0.474532101397773,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010807044558725627,
+      "loss": 0.5209,
+      "step": 10015
+    },
+    {
+      "epoch": 0.4745794835347074,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010805560148257365,
+      "loss": 0.6022,
+      "step": 10016
+    },
+    {
+      "epoch": 0.4746268656716418,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010804075719922638,
+      "loss": 1.0504,
+      "step": 10017
+    },
+    {
+      "epoch": 0.47467424780857614,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010802591273754365,
+      "loss": 0.1208,
+      "step": 10018
+    },
+    {
+      "epoch": 0.47472162994551054,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010801106809785475,
+      "loss": 0.6259,
+      "step": 10019
+    },
+    {
+      "epoch": 0.47476901208244493,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010799622328048888,
+      "loss": 1.0677,
+      "step": 10020
+    },
+    {
+      "epoch": 0.47481639421937927,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010798137828577533,
+      "loss": 1.0154,
+      "step": 10021
+    },
+    {
+      "epoch": 0.47486377635631366,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010796653311404326,
+      "loss": 0.2195,
+      "step": 10022
+    },
+    {
+      "epoch": 0.47491115849324805,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010795168776562198,
+      "loss": 0.2982,
+      "step": 10023
+    },
+    {
+      "epoch": 0.47495854063018245,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010793684224084077,
+      "loss": 1.2982,
+      "step": 10024
+    },
+    {
+      "epoch": 0.4750059227671168,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00010792199654002881,
+      "loss": 1.3599,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4750533049040512,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010790715066351542,
+      "loss": 0.8199,
+      "step": 10026
+    },
+    {
+      "epoch": 0.47510068704098557,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010789230461162984,
+      "loss": 0.9086,
+      "step": 10027
+    },
+    {
+      "epoch": 0.4751480691779199,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0001078774583847014,
+      "loss": 1.1019,
+      "step": 10028
+    },
+    {
+      "epoch": 0.4751954513148543,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010786261198305929,
+      "loss": 0.8015,
+      "step": 10029
+    },
+    {
+      "epoch": 0.4752428334517887,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010784776540703281,
+      "loss": 1.0677,
+      "step": 10030
+    },
+    {
+      "epoch": 0.47529021558872303,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00010783291865695127,
+      "loss": 1.1803,
+      "step": 10031
+    },
+    {
+      "epoch": 0.4753375977256574,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010781807173314394,
+      "loss": 0.953,
+      "step": 10032
+    },
+    {
+      "epoch": 0.4753849798625918,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010780322463594008,
+      "loss": 0.0117,
+      "step": 10033
+    },
+    {
+      "epoch": 0.47543236199952615,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010778837736566902,
+      "loss": 1.2231,
+      "step": 10034
+    },
+    {
+      "epoch": 0.47547974413646055,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00010777352992266006,
+      "loss": 0.1227,
+      "step": 10035
+    },
+    {
+      "epoch": 0.47552712627339494,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010775868230724249,
+      "loss": 0.9482,
+      "step": 10036
+    },
+    {
+      "epoch": 0.47557450841032933,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010774383451974559,
+      "loss": 1.0058,
+      "step": 10037
+    },
+    {
+      "epoch": 0.47562189054726367,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001077289865604987,
+      "loss": 0.145,
+      "step": 10038
+    },
+    {
+      "epoch": 0.47566927268419806,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010771413842983111,
+      "loss": 1.3951,
+      "step": 10039
+    },
+    {
+      "epoch": 0.47571665482113246,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00010769929012807216,
+      "loss": 0.9192,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4757640369580668,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010768444165555114,
+      "loss": 0.8916,
+      "step": 10041
+    },
+    {
+      "epoch": 0.4758114190950012,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001076695930125974,
+      "loss": 1.0961,
+      "step": 10042
+    },
+    {
+      "epoch": 0.4758588012319356,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00010765474419954027,
+      "loss": 0.5015,
+      "step": 10043
+    },
+    {
+      "epoch": 0.4759061833688699,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00010763989521670905,
+      "loss": 1.1383,
+      "step": 10044
+    },
+    {
+      "epoch": 0.4759535655058043,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010762504606443308,
+      "loss": 1.0595,
+      "step": 10045
+    },
+    {
+      "epoch": 0.4760009476427387,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010761019674304173,
+      "loss": 1.2456,
+      "step": 10046
+    },
+    {
+      "epoch": 0.47604832977967304,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010759534725286434,
+      "loss": 0.9614,
+      "step": 10047
+    },
+    {
+      "epoch": 0.47609571191660743,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001075804975942302,
+      "loss": 0.1372,
+      "step": 10048
+    },
+    {
+      "epoch": 0.4761430940535418,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0001075656477674687,
+      "loss": 1.1255,
+      "step": 10049
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001075507977729092,
+      "loss": 0.3984,
+      "step": 10050
+    },
+    {
+      "epoch": 0.47623785832741056,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010753594761088109,
+      "loss": 0.7424,
+      "step": 10051
+    },
+    {
+      "epoch": 0.47628524046434495,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00010752109728171363,
+      "loss": 0.8824,
+      "step": 10052
+    },
+    {
+      "epoch": 0.47633262260127934,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00010750624678573627,
+      "loss": 0.1074,
+      "step": 10053
+    },
+    {
+      "epoch": 0.4763800047382137,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010749139612327832,
+      "loss": 0.9001,
+      "step": 10054
+    },
+    {
+      "epoch": 0.4764273868751481,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00010747654529466921,
+      "loss": 0.4796,
+      "step": 10055
+    },
+    {
+      "epoch": 0.47647476901208247,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001074616943002383,
+      "loss": 0.7125,
+      "step": 10056
+    },
+    {
+      "epoch": 0.4765221511490168,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00010744684314031492,
+      "loss": 1.3499,
+      "step": 10057
+    },
+    {
+      "epoch": 0.4765695332859512,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010743199181522853,
+      "loss": 0.8206,
+      "step": 10058
+    },
+    {
+      "epoch": 0.4766169154228856,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001074171403253085,
+      "loss": 0.0325,
+      "step": 10059
+    },
+    {
+      "epoch": 0.4766642975598199,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00010740228867088415,
+      "loss": 0.8872,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4767116796967543,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010738743685228492,
+      "loss": 0.4424,
+      "step": 10061
+    },
+    {
+      "epoch": 0.4767590618336887,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010737258486984024,
+      "loss": 0.1372,
+      "step": 10062
+    },
+    {
+      "epoch": 0.47680644397062305,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00010735773272387945,
+      "loss": 0.0906,
+      "step": 10063
+    },
+    {
+      "epoch": 0.47685382610755744,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.000107342880414732,
+      "loss": 0.7328,
+      "step": 10064
+    },
+    {
+      "epoch": 0.47690120824449184,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00010732802794272727,
+      "loss": 0.0246,
+      "step": 10065
+    },
+    {
+      "epoch": 0.47694859038142623,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010731317530819471,
+      "loss": 0.0452,
+      "step": 10066
+    },
+    {
+      "epoch": 0.47699597251836057,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010729832251146372,
+      "loss": 0.0289,
+      "step": 10067
+    },
+    {
+      "epoch": 0.47704335465529496,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010728346955286368,
+      "loss": 0.9414,
+      "step": 10068
+    },
+    {
+      "epoch": 0.47709073679222935,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.00010726861643272406,
+      "loss": 0.0023,
+      "step": 10069
+    },
+    {
+      "epoch": 0.4771381189291637,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010725376315137427,
+      "loss": 1.0377,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4771855010660981,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010723890970914371,
+      "loss": 0.5354,
+      "step": 10071
+    },
+    {
+      "epoch": 0.4772328832030325,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010722405610636188,
+      "loss": 0.5258,
+      "step": 10072
+    },
+    {
+      "epoch": 0.4772802653399668,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010720920234335816,
+      "loss": 0.9203,
+      "step": 10073
+    },
+    {
+      "epoch": 0.4773276474769012,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00010719434842046203,
+      "loss": 0.1152,
+      "step": 10074
+    },
+    {
+      "epoch": 0.4773750296138356,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00010717949433800292,
+      "loss": 1.1106,
+      "step": 10075
+    },
+    {
+      "epoch": 0.47742241175076994,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010716464009631024,
+      "loss": 1.0645,
+      "step": 10076
+    },
+    {
+      "epoch": 0.47746979388770433,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010714978569571347,
+      "loss": 0.8776,
+      "step": 10077
+    },
+    {
+      "epoch": 0.4775171760246387,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010713493113654212,
+      "loss": 0.5713,
+      "step": 10078
+    },
+    {
+      "epoch": 0.47756455816157306,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010712007641912556,
+      "loss": 1.1019,
+      "step": 10079
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010710522154379328,
+      "loss": 1.1372,
+      "step": 10080
+    },
+    {
+      "epoch": 0.47765932243544185,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010709036651087478,
+      "loss": 0.7831,
+      "step": 10081
+    },
+    {
+      "epoch": 0.47770670457237624,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00010707551132069949,
+      "loss": 0.8754,
+      "step": 10082
+    },
+    {
+      "epoch": 0.4777540867093106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010706065597359692,
+      "loss": 1.0916,
+      "step": 10083
+    },
+    {
+      "epoch": 0.47780146884624497,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00010704580046989648,
+      "loss": 1.004,
+      "step": 10084
+    },
+    {
+      "epoch": 0.47784885098317936,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001070309448099277,
+      "loss": 0.3348,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4778962331201137,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00010701608899402009,
+      "loss": 0.8956,
+      "step": 10086
+    },
+    {
+      "epoch": 0.4779436152570481,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010700123302250307,
+      "loss": 0.1105,
+      "step": 10087
+    },
+    {
+      "epoch": 0.4779909973939825,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010698637689570614,
+      "loss": 0.8867,
+      "step": 10088
+    },
+    {
+      "epoch": 0.4780383795309168,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010697152061395885,
+      "loss": 0.0139,
+      "step": 10089
+    },
+    {
+      "epoch": 0.4780857616678512,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010695666417759062,
+      "loss": 1.2305,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4781331438047856,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010694180758693104,
+      "loss": 0.1164,
+      "step": 10091
+    },
+    {
+      "epoch": 0.47818052594171995,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010692695084230952,
+      "loss": 0.5308,
+      "step": 10092
+    },
+    {
+      "epoch": 0.47822790807865434,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010691209394405564,
+      "loss": 1.223,
+      "step": 10093
+    },
+    {
+      "epoch": 0.47827529021558873,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010689723689249884,
+      "loss": 1.0544,
+      "step": 10094
+    },
+    {
+      "epoch": 0.4783226723525231,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010688237968796869,
+      "loss": 1.0804,
+      "step": 10095
+    },
+    {
+      "epoch": 0.47837005448945746,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010686752233079467,
+      "loss": 1.0493,
+      "step": 10096
+    },
+    {
+      "epoch": 0.47841743662639186,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00010685266482130636,
+      "loss": 0.8879,
+      "step": 10097
+    },
+    {
+      "epoch": 0.47846481876332625,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00010683780715983323,
+      "loss": 0.1609,
+      "step": 10098
+    },
+    {
+      "epoch": 0.4785122009002606,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010682294934670482,
+      "loss": 0.3973,
+      "step": 10099
+    },
+    {
+      "epoch": 0.478559583037195,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00010680809138225065,
+      "loss": 1.0785,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4786069651741294,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010679323326680029,
+      "loss": 0.8985,
+      "step": 10101
+    },
+    {
+      "epoch": 0.4786543473110637,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010677837500068322,
+      "loss": 0.8989,
+      "step": 10102
+    },
+    {
+      "epoch": 0.4787017294479981,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.000106763516584229,
+      "loss": 0.0173,
+      "step": 10103
+    },
+    {
+      "epoch": 0.4787491115849325,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00010674865801776723,
+      "loss": 0.0018,
+      "step": 10104
+    },
+    {
+      "epoch": 0.47879649372186683,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010673379930162739,
+      "loss": 1.0238,
+      "step": 10105
+    },
+    {
+      "epoch": 0.4788438758588012,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00010671894043613908,
+      "loss": 0.7204,
+      "step": 10106
+    },
+    {
+      "epoch": 0.4788912579957356,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001067040814216318,
+      "loss": 0.3304,
+      "step": 10107
+    },
+    {
+      "epoch": 0.47893864013266996,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00010668922225843512,
+      "loss": 0.129,
+      "step": 10108
+    },
+    {
+      "epoch": 0.47898602226960435,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010667436294687865,
+      "loss": 1.1233,
+      "step": 10109
+    },
+    {
+      "epoch": 0.47903340440653874,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001066595034872919,
+      "loss": 1.0682,
+      "step": 10110
+    },
+    {
+      "epoch": 0.47908078654347314,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00010664464388000445,
+      "loss": 1.2712,
+      "step": 10111
+    },
+    {
+      "epoch": 0.4791281686804075,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001066297841253459,
+      "loss": 0.521,
+      "step": 10112
+    },
+    {
+      "epoch": 0.47917555081734187,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00010661492422364578,
+      "loss": 0.5304,
+      "step": 10113
+    },
+    {
+      "epoch": 0.47922293295427626,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010660006417523372,
+      "loss": 0.8213,
+      "step": 10114
+    },
+    {
+      "epoch": 0.4792703150912106,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010658520398043923,
+      "loss": 0.638,
+      "step": 10115
+    },
+    {
+      "epoch": 0.479317697228145,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010657034363959195,
+      "loss": 0.9732,
+      "step": 10116
+    },
+    {
+      "epoch": 0.4793650793650794,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010655548315302144,
+      "loss": 0.8203,
+      "step": 10117
+    },
+    {
+      "epoch": 0.4794124615020137,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00010654062252105728,
+      "loss": 0.0159,
+      "step": 10118
+    },
+    {
+      "epoch": 0.4794598436389481,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010652576174402909,
+      "loss": 1.1486,
+      "step": 10119
+    },
+    {
+      "epoch": 0.4795072257758825,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00010651090082226647,
+      "loss": 0.0382,
+      "step": 10120
+    },
+    {
+      "epoch": 0.47955460791281684,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010649603975609898,
+      "loss": 1.292,
+      "step": 10121
+    },
+    {
+      "epoch": 0.47960199004975124,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00010648117854585627,
+      "loss": 0.0801,
+      "step": 10122
+    },
+    {
+      "epoch": 0.47964937218668563,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001064663171918679,
+      "loss": 0.7903,
+      "step": 10123
+    },
+    {
+      "epoch": 0.47969675432362,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010645145569446353,
+      "loss": 0.3721,
+      "step": 10124
+    },
+    {
+      "epoch": 0.47974413646055436,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010643659405397273,
+      "loss": 0.8507,
+      "step": 10125
+    },
+    {
+      "epoch": 0.47979151859748875,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010642173227072511,
+      "loss": 0.1726,
+      "step": 10126
+    },
+    {
+      "epoch": 0.47983890073442315,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010640687034505034,
+      "loss": 0.3973,
+      "step": 10127
+    },
+    {
+      "epoch": 0.4798862828713575,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.000106392008277278,
+      "loss": 1.1665,
+      "step": 10128
+    },
+    {
+      "epoch": 0.4799336650082919,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.00010637714606773773,
+      "loss": 0.0156,
+      "step": 10129
+    },
+    {
+      "epoch": 0.47998104714522627,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.00010636228371675916,
+      "loss": 0.0057,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4800284292821606,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0001063474212246719,
+      "loss": 0.8653,
+      "step": 10131
+    },
+    {
+      "epoch": 0.480075811419095,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001063325585918056,
+      "loss": 0.1549,
+      "step": 10132
+    },
+    {
+      "epoch": 0.4801231935560294,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001063176958184899,
+      "loss": 1.2178,
+      "step": 10133
+    },
+    {
+      "epoch": 0.48017057569296373,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001063028329050544,
+      "loss": 0.6577,
+      "step": 10134
+    },
+    {
+      "epoch": 0.4802179578298981,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010628796985182883,
+      "loss": 0.7462,
+      "step": 10135
+    },
+    {
+      "epoch": 0.4802653399668325,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00010627310665914276,
+      "loss": 0.9332,
+      "step": 10136
+    },
+    {
+      "epoch": 0.48031272210376685,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010625824332732586,
+      "loss": 0.5907,
+      "step": 10137
+    },
+    {
+      "epoch": 0.48036010424070125,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00010624337985670782,
+      "loss": 0.2605,
+      "step": 10138
+    },
+    {
+      "epoch": 0.48040748637763564,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010622851624761828,
+      "loss": 1.0526,
+      "step": 10139
+    },
+    {
+      "epoch": 0.48045486851457003,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00010621365250038682,
+      "loss": 1.29,
+      "step": 10140
+    },
+    {
+      "epoch": 0.48050225065150437,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010619878861534318,
+      "loss": 1.2478,
+      "step": 10141
+    },
+    {
+      "epoch": 0.48054963278843876,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.00010618392459281703,
+      "loss": 0.0014,
+      "step": 10142
+    },
+    {
+      "epoch": 0.48059701492537316,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010616906043313803,
+      "loss": 0.7464,
+      "step": 10143
+    },
+    {
+      "epoch": 0.4806443970623075,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00010615419613663582,
+      "loss": 0.859,
+      "step": 10144
+    },
+    {
+      "epoch": 0.4806917791992419,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001061393317036401,
+      "loss": 1.2051,
+      "step": 10145
+    },
+    {
+      "epoch": 0.4807391613361763,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00010612446713448054,
+      "loss": 0.9745,
+      "step": 10146
+    },
+    {
+      "epoch": 0.4807865434731106,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010610960242948687,
+      "loss": 1.1956,
+      "step": 10147
+    },
+    {
+      "epoch": 0.480833925610045,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010609473758898867,
+      "loss": 1.363,
+      "step": 10148
+    },
+    {
+      "epoch": 0.4808813077469794,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00010607987261331567,
+      "loss": 1.1725,
+      "step": 10149
+    },
+    {
+      "epoch": 0.48092868988391374,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010606500750279761,
+      "loss": 1.0699,
+      "step": 10150
+    },
+    {
+      "epoch": 0.48097607202084813,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010605014225776412,
+      "loss": 0.0789,
+      "step": 10151
+    },
+    {
+      "epoch": 0.4810234541577825,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00010603527687854494,
+      "loss": 1.2713,
+      "step": 10152
+    },
+    {
+      "epoch": 0.4810708362947169,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010602041136546971,
+      "loss": 1.409,
+      "step": 10153
+    },
+    {
+      "epoch": 0.48111821843165126,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010600554571886823,
+      "loss": 1.0618,
+      "step": 10154
+    },
+    {
+      "epoch": 0.48116560056858565,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001059906799390701,
+      "loss": 1.3923,
+      "step": 10155
+    },
+    {
+      "epoch": 0.48121298270552004,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010597581402640508,
+      "loss": 0.1376,
+      "step": 10156
+    },
+    {
+      "epoch": 0.4812603648424544,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010596094798120286,
+      "loss": 0.9505,
+      "step": 10157
+    },
+    {
+      "epoch": 0.4813077469793888,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010594608180379317,
+      "loss": 0.0368,
+      "step": 10158
+    },
+    {
+      "epoch": 0.48135512911632317,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010593121549450573,
+      "loss": 0.9188,
+      "step": 10159
+    },
+    {
+      "epoch": 0.4814025112532575,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00010591634905367023,
+      "loss": 1.0409,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4814498933901919,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010590148248161641,
+      "loss": 1.4114,
+      "step": 10161
+    },
+    {
+      "epoch": 0.4814972755271263,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010588661577867403,
+      "loss": 0.8748,
+      "step": 10162
+    },
+    {
+      "epoch": 0.4815446576640606,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010587174894517276,
+      "loss": 0.0263,
+      "step": 10163
+    },
+    {
+      "epoch": 0.481592039800995,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010585688198144232,
+      "loss": 1.2197,
+      "step": 10164
+    },
+    {
+      "epoch": 0.4816394219379294,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010584201488781251,
+      "loss": 1.2011,
+      "step": 10165
+    },
+    {
+      "epoch": 0.48168680407486375,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010582714766461304,
+      "loss": 0.7421,
+      "step": 10166
+    },
+    {
+      "epoch": 0.48173418621179814,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0001058122803121736,
+      "loss": 0.0014,
+      "step": 10167
+    },
+    {
+      "epoch": 0.48178156834873254,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000105797412830824,
+      "loss": 0.9324,
+      "step": 10168
+    },
+    {
+      "epoch": 0.48182895048566693,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.00010578254522089397,
+      "loss": 0.0013,
+      "step": 10169
+    },
+    {
+      "epoch": 0.48187633262260127,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010576767748271326,
+      "loss": 1.0638,
+      "step": 10170
+    },
+    {
+      "epoch": 0.48192371475953566,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010575280961661159,
+      "loss": 1.0468,
+      "step": 10171
+    },
+    {
+      "epoch": 0.48197109689647005,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0001057379416229187,
+      "loss": 0.0099,
+      "step": 10172
+    },
+    {
+      "epoch": 0.4820184790334044,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010572307350196439,
+      "loss": 1.5398,
+      "step": 10173
+    },
+    {
+      "epoch": 0.4820658611703388,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00010570820525407843,
+      "loss": 0.0073,
+      "step": 10174
+    },
+    {
+      "epoch": 0.4821132433072732,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010569333687959053,
+      "loss": 1.2509,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4821606254442075,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00010567846837883049,
+      "loss": 1.298,
+      "step": 10176
+    },
+    {
+      "epoch": 0.4822080075811419,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001056635997521281,
+      "loss": 0.7765,
+      "step": 10177
+    },
+    {
+      "epoch": 0.4822553897180763,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010564873099981308,
+      "loss": 1.2411,
+      "step": 10178
+    },
+    {
+      "epoch": 0.48230277185501064,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001056338621222152,
+      "loss": 0.0974,
+      "step": 10179
+    },
+    {
+      "epoch": 0.48235015399194503,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00010561899311966428,
+      "loss": 0.0739,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4823975361288794,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.00010560412399249009,
+      "loss": 0.0032,
+      "step": 10181
+    },
+    {
+      "epoch": 0.4824449182658138,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010558925474102237,
+      "loss": 0.6145,
+      "step": 10182
+    },
+    {
+      "epoch": 0.48249230040274815,
+      "grad_norm": 0.019775390625,
+      "learning_rate": 0.00010557438536559095,
+      "loss": 0.0009,
+      "step": 10183
+    },
+    {
+      "epoch": 0.48253968253968255,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010555951586652557,
+      "loss": 1.0975,
+      "step": 10184
+    },
+    {
+      "epoch": 0.48258706467661694,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00010554464624415612,
+      "loss": 0.0134,
+      "step": 10185
+    },
+    {
+      "epoch": 0.4826344468135513,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010552977649881226,
+      "loss": 0.8918,
+      "step": 10186
+    },
+    {
+      "epoch": 0.48268182895048567,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010551490663082386,
+      "loss": 1.1871,
+      "step": 10187
+    },
+    {
+      "epoch": 0.48272921108742006,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001055000366405207,
+      "loss": 0.7529,
+      "step": 10188
+    },
+    {
+      "epoch": 0.4827765932243544,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010548516652823262,
+      "loss": 0.1328,
+      "step": 10189
+    },
+    {
+      "epoch": 0.4828239753612888,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010547029629428937,
+      "loss": 1.2239,
+      "step": 10190
+    },
+    {
+      "epoch": 0.4828713574982232,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010545542593902076,
+      "loss": 0.712,
+      "step": 10191
+    },
+    {
+      "epoch": 0.4829187396351575,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010544055546275664,
+      "loss": 0.5598,
+      "step": 10192
+    },
+    {
+      "epoch": 0.4829661217720919,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010542568486582678,
+      "loss": 0.9865,
+      "step": 10193
+    },
+    {
+      "epoch": 0.4830135039090263,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000105410814148561,
+      "loss": 0.9094,
+      "step": 10194
+    },
+    {
+      "epoch": 0.48306088604596065,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010539594331128912,
+      "loss": 1.296,
+      "step": 10195
+    },
+    {
+      "epoch": 0.48310826818289504,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00010538107235434101,
+      "loss": 0.005,
+      "step": 10196
+    },
+    {
+      "epoch": 0.48315565031982943,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00010536620127804639,
+      "loss": 0.0074,
+      "step": 10197
+    },
+    {
+      "epoch": 0.4832030324567638,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010535133008273517,
+      "loss": 0.886,
+      "step": 10198
+    },
+    {
+      "epoch": 0.48325041459369816,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010533645876873715,
+      "loss": 1.3604,
+      "step": 10199
+    },
+    {
+      "epoch": 0.48329779673063256,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010532158733638216,
+      "loss": 0.2285,
+      "step": 10200
+    },
+    {
+      "epoch": 0.48334517886756695,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00010530671578600005,
+      "loss": 0.1313,
+      "step": 10201
+    },
+    {
+      "epoch": 0.4833925610045013,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010529184411792059,
+      "loss": 1.0717,
+      "step": 10202
+    },
+    {
+      "epoch": 0.4834399431414357,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010527697233247369,
+      "loss": 0.6953,
+      "step": 10203
+    },
+    {
+      "epoch": 0.4834873252783701,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010526210042998916,
+      "loss": 0.8257,
+      "step": 10204
+    },
+    {
+      "epoch": 0.4835347074153044,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010524722841079684,
+      "loss": 0.245,
+      "step": 10205
+    },
+    {
+      "epoch": 0.4835820895522388,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010523235627522659,
+      "loss": 0.9412,
+      "step": 10206
+    },
+    {
+      "epoch": 0.4836294716891732,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010521748402360825,
+      "loss": 0.9764,
+      "step": 10207
+    },
+    {
+      "epoch": 0.48367685382610753,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010520261165627168,
+      "loss": 0.1322,
+      "step": 10208
+    },
+    {
+      "epoch": 0.4837242359630419,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010518773917354673,
+      "loss": 0.034,
+      "step": 10209
+    },
+    {
+      "epoch": 0.4837716180999763,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010517286657576324,
+      "loss": 0.8706,
+      "step": 10210
+    },
+    {
+      "epoch": 0.4838190002369107,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010515799386325107,
+      "loss": 0.4869,
+      "step": 10211
+    },
+    {
+      "epoch": 0.48386638237384505,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010514312103634012,
+      "loss": 0.9291,
+      "step": 10212
+    },
+    {
+      "epoch": 0.48391376451077944,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010512824809536019,
+      "loss": 1.3326,
+      "step": 10213
+    },
+    {
+      "epoch": 0.48396114664771384,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010511337504064118,
+      "loss": 0.8072,
+      "step": 10214
+    },
+    {
+      "epoch": 0.4840085287846482,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00010509850187251298,
+      "loss": 0.2092,
+      "step": 10215
+    },
+    {
+      "epoch": 0.48405591092158257,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010508362859130546,
+      "loss": 0.8456,
+      "step": 10216
+    },
+    {
+      "epoch": 0.48410329305851696,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010506875519734843,
+      "loss": 0.2392,
+      "step": 10217
+    },
+    {
+      "epoch": 0.4841506751954513,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010505388169097182,
+      "loss": 0.1449,
+      "step": 10218
+    },
+    {
+      "epoch": 0.4841980573323857,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00010503900807250548,
+      "loss": 0.128,
+      "step": 10219
+    },
+    {
+      "epoch": 0.4842454394693201,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00010502413434227933,
+      "loss": 1.0062,
+      "step": 10220
+    },
+    {
+      "epoch": 0.4842928216062544,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010500926050062323,
+      "loss": 0.8574,
+      "step": 10221
+    },
+    {
+      "epoch": 0.4843402037431888,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010499438654786706,
+      "loss": 0.9706,
+      "step": 10222
+    },
+    {
+      "epoch": 0.4843875858801232,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00010497951248434073,
+      "loss": 0.1635,
+      "step": 10223
+    },
+    {
+      "epoch": 0.48443496801705754,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010496463831037413,
+      "loss": 0.1218,
+      "step": 10224
+    },
+    {
+      "epoch": 0.48448235015399194,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001049497640262971,
+      "loss": 0.0917,
+      "step": 10225
+    },
+    {
+      "epoch": 0.48452973229092633,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010493488963243958,
+      "loss": 0.0567,
+      "step": 10226
+    },
+    {
+      "epoch": 0.4845771144278607,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001049200151291315,
+      "loss": 1.2936,
+      "step": 10227
+    },
+    {
+      "epoch": 0.48462449656479506,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010490514051670271,
+      "loss": 0.1276,
+      "step": 10228
+    },
+    {
+      "epoch": 0.48467187870172945,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00010489026579548311,
+      "loss": 1.432,
+      "step": 10229
+    },
+    {
+      "epoch": 0.48471926083866385,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010487539096580263,
+      "loss": 0.1523,
+      "step": 10230
+    },
+    {
+      "epoch": 0.4847666429755982,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010486051602799118,
+      "loss": 1.1239,
+      "step": 10231
+    },
+    {
+      "epoch": 0.4848140251125326,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00010484564098237868,
+      "loss": 0.0408,
+      "step": 10232
+    },
+    {
+      "epoch": 0.48486140724946697,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010483076582929501,
+      "loss": 0.1257,
+      "step": 10233
+    },
+    {
+      "epoch": 0.4849087893864013,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010481589056907006,
+      "loss": 0.8312,
+      "step": 10234
+    },
+    {
+      "epoch": 0.4849561715233357,
+      "grad_norm": 0.02490234375,
+      "learning_rate": 0.00010480101520203385,
+      "loss": 0.0009,
+      "step": 10235
+    },
+    {
+      "epoch": 0.4850035536602701,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010478613972851619,
+      "loss": 0.7291,
+      "step": 10236
+    },
+    {
+      "epoch": 0.48505093579720443,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00010477126414884706,
+      "loss": 1.1389,
+      "step": 10237
+    },
+    {
+      "epoch": 0.4850983179341388,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001047563884633564,
+      "loss": 0.4756,
+      "step": 10238
+    },
+    {
+      "epoch": 0.4851457000710732,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010474151267237408,
+      "loss": 0.8407,
+      "step": 10239
+    },
+    {
+      "epoch": 0.4851930822080076,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010472663677623009,
+      "loss": 0.8124,
+      "step": 10240
+    },
+    {
+      "epoch": 0.48524046434494195,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00010471176077525428,
+      "loss": 0.1401,
+      "step": 10241
+    },
+    {
+      "epoch": 0.48528784648187634,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010469688466977667,
+      "loss": 0.8301,
+      "step": 10242
+    },
+    {
+      "epoch": 0.48533522861881073,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010468200846012717,
+      "loss": 0.8308,
+      "step": 10243
+    },
+    {
+      "epoch": 0.48538261075574507,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010466713214663568,
+      "loss": 1.3149,
+      "step": 10244
+    },
+    {
+      "epoch": 0.48542999289267946,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010465225572963217,
+      "loss": 0.7159,
+      "step": 10245
+    },
+    {
+      "epoch": 0.48547737502961386,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010463737920944663,
+      "loss": 1.0653,
+      "step": 10246
+    },
+    {
+      "epoch": 0.4855247571665482,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010462250258640891,
+      "loss": 0.9131,
+      "step": 10247
+    },
+    {
+      "epoch": 0.4855721393034826,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010460762586084905,
+      "loss": 0.8464,
+      "step": 10248
+    },
+    {
+      "epoch": 0.485619521440417,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001045927490330969,
+      "loss": 0.634,
+      "step": 10249
+    },
+    {
+      "epoch": 0.4856669035773513,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010457787210348251,
+      "loss": 1.1345,
+      "step": 10250
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010456299507233577,
+      "loss": 0.3389,
+      "step": 10251
+    },
+    {
+      "epoch": 0.4857616678512201,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010454811793998666,
+      "loss": 0.445,
+      "step": 10252
+    },
+    {
+      "epoch": 0.48580904998815444,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010453324070676516,
+      "loss": 0.9121,
+      "step": 10253
+    },
+    {
+      "epoch": 0.48585643212508883,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010451836337300121,
+      "loss": 0.838,
+      "step": 10254
+    },
+    {
+      "epoch": 0.4859038142620232,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00010450348593902475,
+      "loss": 0.995,
+      "step": 10255
+    },
+    {
+      "epoch": 0.4859511963989576,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010448860840516579,
+      "loss": 0.5999,
+      "step": 10256
+    },
+    {
+      "epoch": 0.48599857853589196,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00010447373077175427,
+      "loss": 0.3551,
+      "step": 10257
+    },
+    {
+      "epoch": 0.48604596067282635,
+      "grad_norm": 0.0269775390625,
+      "learning_rate": 0.00010445885303912017,
+      "loss": 0.0014,
+      "step": 10258
+    },
+    {
+      "epoch": 0.48609334280976074,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010444397520759343,
+      "loss": 0.9227,
+      "step": 10259
+    },
+    {
+      "epoch": 0.4861407249466951,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010442909727750407,
+      "loss": 1.0303,
+      "step": 10260
+    },
+    {
+      "epoch": 0.4861881070836295,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.00010441421924918209,
+      "loss": 0.0103,
+      "step": 10261
+    },
+    {
+      "epoch": 0.48623548922056387,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010439934112295737,
+      "loss": 0.0176,
+      "step": 10262
+    },
+    {
+      "epoch": 0.4862828713574982,
+      "grad_norm": 0.020263671875,
+      "learning_rate": 0.00010438446289916,
+      "loss": 0.0008,
+      "step": 10263
+    },
+    {
+      "epoch": 0.4863302534944326,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010436958457811987,
+      "loss": 0.5984,
+      "step": 10264
+    },
+    {
+      "epoch": 0.486377635631367,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00010435470616016703,
+      "loss": 1.3061,
+      "step": 10265
+    },
+    {
+      "epoch": 0.4864250177683013,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010433982764563146,
+      "loss": 0.8373,
+      "step": 10266
+    },
+    {
+      "epoch": 0.4864723999052357,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010432494903484312,
+      "loss": 0.8359,
+      "step": 10267
+    },
+    {
+      "epoch": 0.4865197820421701,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010431007032813199,
+      "loss": 0.0201,
+      "step": 10268
+    },
+    {
+      "epoch": 0.48656716417910445,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010429519152582812,
+      "loss": 0.7844,
+      "step": 10269
+    },
+    {
+      "epoch": 0.48661454631603884,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010428031262826148,
+      "loss": 1.022,
+      "step": 10270
+    },
+    {
+      "epoch": 0.48666192845297324,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010426543363576207,
+      "loss": 0.8896,
+      "step": 10271
+    },
+    {
+      "epoch": 0.48670931058990763,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010425055454865987,
+      "loss": 1.1148,
+      "step": 10272
+    },
+    {
+      "epoch": 0.48675669272684197,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00010423567536728489,
+      "loss": 0.8805,
+      "step": 10273
+    },
+    {
+      "epoch": 0.48680407486377636,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010422079609196716,
+      "loss": 0.0896,
+      "step": 10274
+    },
+    {
+      "epoch": 0.48685145700071075,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010420591672303666,
+      "loss": 1.0757,
+      "step": 10275
+    },
+    {
+      "epoch": 0.4868988391376451,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00010419103726082339,
+      "loss": 1.2818,
+      "step": 10276
+    },
+    {
+      "epoch": 0.4869462212745795,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010417615770565741,
+      "loss": 0.677,
+      "step": 10277
+    },
+    {
+      "epoch": 0.4869936034115139,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00010416127805786869,
+      "loss": 0.0657,
+      "step": 10278
+    },
+    {
+      "epoch": 0.4870409855484482,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010414639831778726,
+      "loss": 0.7278,
+      "step": 10279
+    },
+    {
+      "epoch": 0.4870883676853826,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010413151848574309,
+      "loss": 0.5555,
+      "step": 10280
+    },
+    {
+      "epoch": 0.487135749822317,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001041166385620663,
+      "loss": 1.2327,
+      "step": 10281
+    },
+    {
+      "epoch": 0.48718313195925134,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00010410175854708681,
+      "loss": 0.0819,
+      "step": 10282
+    },
+    {
+      "epoch": 0.48723051409618573,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001040868784411347,
+      "loss": 0.9043,
+      "step": 10283
+    },
+    {
+      "epoch": 0.4872778962331201,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010407199824454,
+      "loss": 0.2155,
+      "step": 10284
+    },
+    {
+      "epoch": 0.4873252783700545,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00010405711795763269,
+      "loss": 0.0108,
+      "step": 10285
+    },
+    {
+      "epoch": 0.48737266050698885,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001040422375807428,
+      "loss": 0.0133,
+      "step": 10286
+    },
+    {
+      "epoch": 0.48742004264392325,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010402735711420045,
+      "loss": 0.7902,
+      "step": 10287
+    },
+    {
+      "epoch": 0.48746742478085764,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010401247655833555,
+      "loss": 1.0957,
+      "step": 10288
+    },
+    {
+      "epoch": 0.487514806917792,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.00010399759591347825,
+      "loss": 0.0049,
+      "step": 10289
+    },
+    {
+      "epoch": 0.48756218905472637,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00010398271517995848,
+      "loss": 0.0964,
+      "step": 10290
+    },
+    {
+      "epoch": 0.48760957119166076,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00010396783435810636,
+      "loss": 0.9047,
+      "step": 10291
+    },
+    {
+      "epoch": 0.4876569533285951,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010395295344825189,
+      "loss": 0.0149,
+      "step": 10292
+    },
+    {
+      "epoch": 0.4877043354655295,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010393807245072511,
+      "loss": 1.1522,
+      "step": 10293
+    },
+    {
+      "epoch": 0.4877517176024639,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00010392319136585609,
+      "loss": 0.0133,
+      "step": 10294
+    },
+    {
+      "epoch": 0.4877990997393982,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00010390831019397487,
+      "loss": 0.0323,
+      "step": 10295
+    },
+    {
+      "epoch": 0.4878464818763326,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001038934289354115,
+      "loss": 0.9189,
+      "step": 10296
+    },
+    {
+      "epoch": 0.487893864013267,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.000103878547590496,
+      "loss": 1.0949,
+      "step": 10297
+    },
+    {
+      "epoch": 0.48794124615020135,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010386366615955844,
+      "loss": 1.1003,
+      "step": 10298
+    },
+    {
+      "epoch": 0.48798862828713574,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001038487846429289,
+      "loss": 0.0741,
+      "step": 10299
+    },
+    {
+      "epoch": 0.48803601042407013,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010383390304093743,
+      "loss": 1.0395,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4880833925610045,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00010381902135391404,
+      "loss": 0.0056,
+      "step": 10301
+    },
+    {
+      "epoch": 0.48813077469793886,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010380413958218884,
+      "loss": 1.0296,
+      "step": 10302
+    },
+    {
+      "epoch": 0.48817815683487326,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001037892577260919,
+      "loss": 0.4365,
+      "step": 10303
+    },
+    {
+      "epoch": 0.48822553897180765,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00010377437578595323,
+      "loss": 1.3547,
+      "step": 10304
+    },
+    {
+      "epoch": 0.488272921108742,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00010375949376210295,
+      "loss": 0.6985,
+      "step": 10305
+    },
+    {
+      "epoch": 0.4883203032456764,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010374461165487105,
+      "loss": 0.8957,
+      "step": 10306
+    },
+    {
+      "epoch": 0.48836768538261077,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010372972946458772,
+      "loss": 1.1189,
+      "step": 10307
+    },
+    {
+      "epoch": 0.4884150675195451,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010371484719158291,
+      "loss": 0.007,
+      "step": 10308
+    },
+    {
+      "epoch": 0.4884624496564795,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010369996483618676,
+      "loss": 0.8763,
+      "step": 10309
+    },
+    {
+      "epoch": 0.4885098317934139,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010368508239872934,
+      "loss": 0.9605,
+      "step": 10310
+    },
+    {
+      "epoch": 0.48855721393034823,
+      "grad_norm": 0.000507354736328125,
+      "learning_rate": 0.00010367019987954072,
+      "loss": 0.0,
+      "step": 10311
+    },
+    {
+      "epoch": 0.4886045960672826,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00010365531727895099,
+      "loss": 1.1541,
+      "step": 10312
+    },
+    {
+      "epoch": 0.488651978204217,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010364043459729018,
+      "loss": 0.5847,
+      "step": 10313
+    },
+    {
+      "epoch": 0.4886993603411514,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00010362555183488842,
+      "loss": 0.7224,
+      "step": 10314
+    },
+    {
+      "epoch": 0.48874674247808575,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010361066899207581,
+      "loss": 0.724,
+      "step": 10315
+    },
+    {
+      "epoch": 0.48879412461502014,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.00010359578606918236,
+      "loss": 0.0045,
+      "step": 10316
+    },
+    {
+      "epoch": 0.48884150675195454,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00010358090306653822,
+      "loss": 0.1146,
+      "step": 10317
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010356601998447351,
+      "loss": 1.2059,
+      "step": 10318
+    },
+    {
+      "epoch": 0.48893627102582327,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010355113682331824,
+      "loss": 0.433,
+      "step": 10319
+    },
+    {
+      "epoch": 0.48898365316275766,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010353625358340253,
+      "loss": 0.9884,
+      "step": 10320
+    },
+    {
+      "epoch": 0.489031035299692,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010352137026505649,
+      "loss": 0.9016,
+      "step": 10321
+    },
+    {
+      "epoch": 0.4890784174366264,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0001035064868686102,
+      "loss": 0.0052,
+      "step": 10322
+    },
+    {
+      "epoch": 0.4891257995735608,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00010349160339439379,
+      "loss": 1.0137,
+      "step": 10323
+    },
+    {
+      "epoch": 0.4891731817104951,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010347671984273731,
+      "loss": 1.1278,
+      "step": 10324
+    },
+    {
+      "epoch": 0.4892205638474295,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010346183621397087,
+      "loss": 0.0816,
+      "step": 10325
+    },
+    {
+      "epoch": 0.4892679459843639,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010344695250842464,
+      "loss": 0.5045,
+      "step": 10326
+    },
+    {
+      "epoch": 0.48931532812129824,
+      "grad_norm": 0.01312255859375,
+      "learning_rate": 0.00010343206872642865,
+      "loss": 0.0006,
+      "step": 10327
+    },
+    {
+      "epoch": 0.48936271025823264,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00010341718486831304,
+      "loss": 0.6119,
+      "step": 10328
+    },
+    {
+      "epoch": 0.48941009239516703,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001034023009344079,
+      "loss": 1.2241,
+      "step": 10329
+    },
+    {
+      "epoch": 0.4894574745321014,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010338741692504336,
+      "loss": 0.8122,
+      "step": 10330
+    },
+    {
+      "epoch": 0.48950485666903576,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010337253284054951,
+      "loss": 1.0274,
+      "step": 10331
+    },
+    {
+      "epoch": 0.48955223880597015,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010335764868125646,
+      "loss": 0.1195,
+      "step": 10332
+    },
+    {
+      "epoch": 0.48959962094290455,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010334276444749436,
+      "loss": 1.2966,
+      "step": 10333
+    },
+    {
+      "epoch": 0.4896470030798389,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010332788013959331,
+      "loss": 0.8606,
+      "step": 10334
+    },
+    {
+      "epoch": 0.4896943852167733,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00010331299575788342,
+      "loss": 0.0228,
+      "step": 10335
+    },
+    {
+      "epoch": 0.48974176735370767,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001032981113026948,
+      "loss": 0.0618,
+      "step": 10336
+    },
+    {
+      "epoch": 0.489789149490642,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010328322677435757,
+      "loss": 0.024,
+      "step": 10337
+    },
+    {
+      "epoch": 0.4898365316275764,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010326834217320191,
+      "loss": 1.2858,
+      "step": 10338
+    },
+    {
+      "epoch": 0.4898839137645108,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010325345749955789,
+      "loss": 0.1277,
+      "step": 10339
+    },
+    {
+      "epoch": 0.48993129590144513,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010323857275375561,
+      "loss": 0.1292,
+      "step": 10340
+    },
+    {
+      "epoch": 0.4899786780383795,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010322368793612529,
+      "loss": 0.6946,
+      "step": 10341
+    },
+    {
+      "epoch": 0.4900260601753139,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00010320880304699699,
+      "loss": 0.8295,
+      "step": 10342
+    },
+    {
+      "epoch": 0.4900734423122483,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010319391808670082,
+      "loss": 0.8974,
+      "step": 10343
+    },
+    {
+      "epoch": 0.49012082444918265,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00010317903305556697,
+      "loss": 0.0449,
+      "step": 10344
+    },
+    {
+      "epoch": 0.49016820658611704,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010316414795392555,
+      "loss": 0.8658,
+      "step": 10345
+    },
+    {
+      "epoch": 0.49021558872305143,
+      "grad_norm": 0.00701904296875,
+      "learning_rate": 0.00010314926278210673,
+      "loss": 0.0005,
+      "step": 10346
+    },
+    {
+      "epoch": 0.49026297085998577,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010313437754044058,
+      "loss": 0.6892,
+      "step": 10347
+    },
+    {
+      "epoch": 0.49031035299692016,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010311949222925727,
+      "loss": 0.3972,
+      "step": 10348
+    },
+    {
+      "epoch": 0.49035773513385456,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010310460684888699,
+      "loss": 1.0022,
+      "step": 10349
+    },
+    {
+      "epoch": 0.4904051172707889,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010308972139965982,
+      "loss": 1.0487,
+      "step": 10350
+    },
+    {
+      "epoch": 0.4904524994077233,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001030748358819059,
+      "loss": 0.9667,
+      "step": 10351
+    },
+    {
+      "epoch": 0.4904998815446577,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00010305995029595539,
+      "loss": 0.0107,
+      "step": 10352
+    },
+    {
+      "epoch": 0.490547263681592,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010304506464213848,
+      "loss": 1.1267,
+      "step": 10353
+    },
+    {
+      "epoch": 0.4905946458185264,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00010303017892078523,
+      "loss": 0.0282,
+      "step": 10354
+    },
+    {
+      "epoch": 0.4906420279554608,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010301529313222587,
+      "loss": 0.6471,
+      "step": 10355
+    },
+    {
+      "epoch": 0.49068941009239514,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001030004072767905,
+      "loss": 0.9969,
+      "step": 10356
+    },
+    {
+      "epoch": 0.49073679222932953,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.00010298552135480932,
+      "loss": 0.0024,
+      "step": 10357
+    },
+    {
+      "epoch": 0.4907841743662639,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010297063536661247,
+      "loss": 0.7482,
+      "step": 10358
+    },
+    {
+      "epoch": 0.4908315565031983,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010295574931253005,
+      "loss": 0.8796,
+      "step": 10359
+    },
+    {
+      "epoch": 0.49087893864013266,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010294086319289227,
+      "loss": 0.4574,
+      "step": 10360
+    },
+    {
+      "epoch": 0.49092632077706705,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010292597700802928,
+      "loss": 1.0798,
+      "step": 10361
+    },
+    {
+      "epoch": 0.49097370291400144,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00010291109075827124,
+      "loss": 1.0982,
+      "step": 10362
+    },
+    {
+      "epoch": 0.4910210850509358,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001028962044439483,
+      "loss": 0.7837,
+      "step": 10363
+    },
+    {
+      "epoch": 0.4910684671878702,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00010288131806539063,
+      "loss": 0.7936,
+      "step": 10364
+    },
+    {
+      "epoch": 0.49111584932480457,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00010286643162292841,
+      "loss": 0.0107,
+      "step": 10365
+    },
+    {
+      "epoch": 0.4911632314617389,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010285154511689179,
+      "loss": 0.0806,
+      "step": 10366
+    },
+    {
+      "epoch": 0.4912106135986733,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010283665854761091,
+      "loss": 1.3351,
+      "step": 10367
+    },
+    {
+      "epoch": 0.4912579957356077,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00010282177191541597,
+      "loss": 0.9226,
+      "step": 10368
+    },
+    {
+      "epoch": 0.491305377872542,
+      "grad_norm": 0.0238037109375,
+      "learning_rate": 0.00010280688522063717,
+      "loss": 0.001,
+      "step": 10369
+    },
+    {
+      "epoch": 0.4913527600094764,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010279199846360461,
+      "loss": 1.0816,
+      "step": 10370
+    },
+    {
+      "epoch": 0.4914001421464108,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010277711164464852,
+      "loss": 1.2541,
+      "step": 10371
+    },
+    {
+      "epoch": 0.4914475242833452,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00010276222476409905,
+      "loss": 0.6041,
+      "step": 10372
+    },
+    {
+      "epoch": 0.49149490642027954,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010274733782228638,
+      "loss": 1.2356,
+      "step": 10373
+    },
+    {
+      "epoch": 0.49154228855721394,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010273245081954067,
+      "loss": 1.1031,
+      "step": 10374
+    },
+    {
+      "epoch": 0.49158967069414833,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010271756375619208,
+      "loss": 0.7845,
+      "step": 10375
+    },
+    {
+      "epoch": 0.49163705283108267,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010270267663257088,
+      "loss": 0.7275,
+      "step": 10376
+    },
+    {
+      "epoch": 0.49168443496801706,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00010268778944900714,
+      "loss": 0.8636,
+      "step": 10377
+    },
+    {
+      "epoch": 0.49173181710495145,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010267290220583112,
+      "loss": 0.8971,
+      "step": 10378
+    },
+    {
+      "epoch": 0.4917791992418858,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.00010265801490337296,
+      "loss": 0.0027,
+      "step": 10379
+    },
+    {
+      "epoch": 0.4918265813788202,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001026431275419629,
+      "loss": 0.0894,
+      "step": 10380
+    },
+    {
+      "epoch": 0.4918739635157546,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010262824012193107,
+      "loss": 0.9239,
+      "step": 10381
+    },
+    {
+      "epoch": 0.4919213456526889,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00010261335264360765,
+      "loss": 1.1283,
+      "step": 10382
+    },
+    {
+      "epoch": 0.4919687277896233,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00010259846510732288,
+      "loss": 1.5949,
+      "step": 10383
+    },
+    {
+      "epoch": 0.4920161099265577,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001025835775134069,
+      "loss": 1.0749,
+      "step": 10384
+    },
+    {
+      "epoch": 0.49206349206349204,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00010256868986218993,
+      "loss": 0.1287,
+      "step": 10385
+    },
+    {
+      "epoch": 0.49211087420042643,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00010255380215400215,
+      "loss": 0.9982,
+      "step": 10386
+    },
+    {
+      "epoch": 0.4921582563373608,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010253891438917375,
+      "loss": 1.1091,
+      "step": 10387
+    },
+    {
+      "epoch": 0.4922056384742952,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00010252402656803497,
+      "loss": 0.9044,
+      "step": 10388
+    },
+    {
+      "epoch": 0.49225302061122955,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010250913869091593,
+      "loss": 1.4099,
+      "step": 10389
+    },
+    {
+      "epoch": 0.49230040274816395,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001024942507581469,
+      "loss": 1.1162,
+      "step": 10390
+    },
+    {
+      "epoch": 0.49234778488509834,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000102479362770058,
+      "loss": 0.086,
+      "step": 10391
+    },
+    {
+      "epoch": 0.4923951670220327,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.00010246447472697953,
+      "loss": 0.0024,
+      "step": 10392
+    },
+    {
+      "epoch": 0.49244254915896707,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001024495866292416,
+      "loss": 1.1615,
+      "step": 10393
+    },
+    {
+      "epoch": 0.49248993129590146,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010243469847717444,
+      "loss": 0.6211,
+      "step": 10394
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001024198102711083,
+      "loss": 1.0741,
+      "step": 10395
+    },
+    {
+      "epoch": 0.4925846955697702,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001024049220113733,
+      "loss": 1.2978,
+      "step": 10396
+    },
+    {
+      "epoch": 0.4926320777067046,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.00010239003369829974,
+      "loss": 0.0049,
+      "step": 10397
+    },
+    {
+      "epoch": 0.4926794598436389,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010237514533221774,
+      "loss": 0.6888,
+      "step": 10398
+    },
+    {
+      "epoch": 0.4927268419805733,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00010236025691345757,
+      "loss": 0.1765,
+      "step": 10399
+    },
+    {
+      "epoch": 0.4927742241175077,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010234536844234941,
+      "loss": 1.2865,
+      "step": 10400
+    },
+    {
+      "epoch": 0.4928216062544421,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00010233047991922344,
+      "loss": 0.7992,
+      "step": 10401
+    },
+    {
+      "epoch": 0.49286898839137644,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00010231559134440993,
+      "loss": 1.3444,
+      "step": 10402
+    },
+    {
+      "epoch": 0.49291637052831083,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001023007027182391,
+      "loss": 1.0896,
+      "step": 10403
+    },
+    {
+      "epoch": 0.4929637526652452,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00010228581404104112,
+      "loss": 0.0345,
+      "step": 10404
+    },
+    {
+      "epoch": 0.49301113480217956,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00010227092531314621,
+      "loss": 0.0355,
+      "step": 10405
+    },
+    {
+      "epoch": 0.49305851693911396,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001022560365348846,
+      "loss": 1.055,
+      "step": 10406
+    },
+    {
+      "epoch": 0.49310589907604835,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001022411477065865,
+      "loss": 1.2288,
+      "step": 10407
+    },
+    {
+      "epoch": 0.4931532812129827,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00010222625882858212,
+      "loss": 0.9305,
+      "step": 10408
+    },
+    {
+      "epoch": 0.4932006633499171,
+      "grad_norm": 0.00274658203125,
+      "learning_rate": 0.0001022113699012017,
+      "loss": 0.0002,
+      "step": 10409
+    },
+    {
+      "epoch": 0.49324804548685147,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010219648092477545,
+      "loss": 0.8266,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4932954276237858,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010218159189963361,
+      "loss": 0.9723,
+      "step": 10411
+    },
+    {
+      "epoch": 0.4933428097607202,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00010216670282610637,
+      "loss": 0.2313,
+      "step": 10412
+    },
+    {
+      "epoch": 0.4933901918976546,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010215181370452399,
+      "loss": 0.5898,
+      "step": 10413
+    },
+    {
+      "epoch": 0.49343757403458893,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010213692453521661,
+      "loss": 0.8182,
+      "step": 10414
+    },
+    {
+      "epoch": 0.4934849561715233,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010212203531851459,
+      "loss": 0.0812,
+      "step": 10415
+    },
+    {
+      "epoch": 0.4935323383084577,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010210714605474805,
+      "loss": 0.2392,
+      "step": 10416
+    },
+    {
+      "epoch": 0.4935797204453921,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010209225674424727,
+      "loss": 0.1132,
+      "step": 10417
+    },
+    {
+      "epoch": 0.49362710258232645,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00010207736738734246,
+      "loss": 0.7996,
+      "step": 10418
+    },
+    {
+      "epoch": 0.49367448471926084,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010206247798436385,
+      "loss": 0.0155,
+      "step": 10419
+    },
+    {
+      "epoch": 0.49372186685619524,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00010204758853564167,
+      "loss": 0.3139,
+      "step": 10420
+    },
+    {
+      "epoch": 0.4937692489931296,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010203269904150619,
+      "loss": 0.546,
+      "step": 10421
+    },
+    {
+      "epoch": 0.49381663113006397,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00010201780950228759,
+      "loss": 1.3112,
+      "step": 10422
+    },
+    {
+      "epoch": 0.49386401326699836,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010200291991831611,
+      "loss": 0.8465,
+      "step": 10423
+    },
+    {
+      "epoch": 0.4939113954039327,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000101988030289922,
+      "loss": 0.4812,
+      "step": 10424
+    },
+    {
+      "epoch": 0.4939587775408671,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010197314061743551,
+      "loss": 1.0264,
+      "step": 10425
+    },
+    {
+      "epoch": 0.4940061596778015,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00010195825090118687,
+      "loss": 0.8086,
+      "step": 10426
+    },
+    {
+      "epoch": 0.4940535418147358,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00010194336114150629,
+      "loss": 2.0573,
+      "step": 10427
+    },
+    {
+      "epoch": 0.4941009239516702,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.00010192847133872405,
+      "loss": 0.0087,
+      "step": 10428
+    },
+    {
+      "epoch": 0.4941483060886046,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010191358149317036,
+      "loss": 0.6567,
+      "step": 10429
+    },
+    {
+      "epoch": 0.494195688225539,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010189869160517549,
+      "loss": 0.5986,
+      "step": 10430
+    },
+    {
+      "epoch": 0.49424307036247334,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010188380167506963,
+      "loss": 0.0313,
+      "step": 10431
+    },
+    {
+      "epoch": 0.49429045249940773,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010186891170318306,
+      "loss": 0.4554,
+      "step": 10432
+    },
+    {
+      "epoch": 0.4943378346363421,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 0.00010185402168984601,
+      "loss": 0.0037,
+      "step": 10433
+    },
+    {
+      "epoch": 0.49438521677327646,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00010183913163538876,
+      "loss": 0.8794,
+      "step": 10434
+    },
+    {
+      "epoch": 0.49443259891021085,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00010182424154014152,
+      "loss": 1.1408,
+      "step": 10435
+    },
+    {
+      "epoch": 0.49447998104714525,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00010180935140443458,
+      "loss": 0.1155,
+      "step": 10436
+    },
+    {
+      "epoch": 0.4945273631840796,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001017944612285981,
+      "loss": 1.2385,
+      "step": 10437
+    },
+    {
+      "epoch": 0.494574745321014,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010177957101296241,
+      "loss": 0.7973,
+      "step": 10438
+    },
+    {
+      "epoch": 0.49462212745794837,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010176468075785773,
+      "loss": 1.2734,
+      "step": 10439
+    },
+    {
+      "epoch": 0.4946695095948827,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00010174979046361429,
+      "loss": 1.0335,
+      "step": 10440
+    },
+    {
+      "epoch": 0.4947168917318171,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010173490013056237,
+      "loss": 0.7154,
+      "step": 10441
+    },
+    {
+      "epoch": 0.4947642738687515,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00010172000975903221,
+      "loss": 1.1064,
+      "step": 10442
+    },
+    {
+      "epoch": 0.49481165600568583,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00010170511934935408,
+      "loss": 0.0372,
+      "step": 10443
+    },
+    {
+      "epoch": 0.4948590381426202,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001016902289018582,
+      "loss": 0.1204,
+      "step": 10444
+    },
+    {
+      "epoch": 0.4949064202795546,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010167533841687486,
+      "loss": 0.984,
+      "step": 10445
+    },
+    {
+      "epoch": 0.494953802416489,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00010166044789473431,
+      "loss": 0.0056,
+      "step": 10446
+    },
+    {
+      "epoch": 0.49500118455342335,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010164555733576675,
+      "loss": 0.6282,
+      "step": 10447
+    },
+    {
+      "epoch": 0.49504856669035774,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010163066674030251,
+      "loss": 1.0109,
+      "step": 10448
+    },
+    {
+      "epoch": 0.49509594882729213,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010161577610867183,
+      "loss": 1.0961,
+      "step": 10449
+    },
+    {
+      "epoch": 0.49514333096422647,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00010160088544120494,
+      "loss": 0.9837,
+      "step": 10450
+    },
+    {
+      "epoch": 0.49519071310116086,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00010158599473823209,
+      "loss": 1.0436,
+      "step": 10451
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00010157110400008363,
+      "loss": 0.9952,
+      "step": 10452
+    },
+    {
+      "epoch": 0.4952854773750296,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010155621322708973,
+      "loss": 1.1686,
+      "step": 10453
+    },
+    {
+      "epoch": 0.495332859511964,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010154132241958065,
+      "loss": 1.1098,
+      "step": 10454
+    },
+    {
+      "epoch": 0.4953802416488984,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001015264315778867,
+      "loss": 0.4043,
+      "step": 10455
+    },
+    {
+      "epoch": 0.4954276237858327,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00010151154070233813,
+      "loss": 1.1451,
+      "step": 10456
+    },
+    {
+      "epoch": 0.4954750059227671,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010149664979326518,
+      "loss": 0.969,
+      "step": 10457
+    },
+    {
+      "epoch": 0.4955223880597015,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010148175885099815,
+      "loss": 0.9108,
+      "step": 10458
+    },
+    {
+      "epoch": 0.4955697701966359,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010146686787586728,
+      "loss": 0.5957,
+      "step": 10459
+    },
+    {
+      "epoch": 0.49561715233357023,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010145197686820285,
+      "loss": 1.1212,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4956645344705046,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010143708582833513,
+      "loss": 0.6518,
+      "step": 10461
+    },
+    {
+      "epoch": 0.495711916607439,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010142219475659436,
+      "loss": 0.4941,
+      "step": 10462
+    },
+    {
+      "epoch": 0.49575929874437336,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00010140730365331082,
+      "loss": 1.5096,
+      "step": 10463
+    },
+    {
+      "epoch": 0.49580668088130775,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001013924125188148,
+      "loss": 0.0465,
+      "step": 10464
+    },
+    {
+      "epoch": 0.49585406301824214,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010137752135343653,
+      "loss": 0.0188,
+      "step": 10465
+    },
+    {
+      "epoch": 0.4959014451551765,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001013626301575063,
+      "loss": 1.0475,
+      "step": 10466
+    },
+    {
+      "epoch": 0.49594882729211087,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010134773893135438,
+      "loss": 0.5929,
+      "step": 10467
+    },
+    {
+      "epoch": 0.49599620942904526,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001013328476753111,
+      "loss": 0.4285,
+      "step": 10468
+    },
+    {
+      "epoch": 0.4960435915659796,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00010131795638970665,
+      "loss": 0.1665,
+      "step": 10469
+    },
+    {
+      "epoch": 0.496090973702914,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00010130306507487133,
+      "loss": 0.8403,
+      "step": 10470
+    },
+    {
+      "epoch": 0.4961383558398484,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010128817373113539,
+      "loss": 0.9496,
+      "step": 10471
+    },
+    {
+      "epoch": 0.4961857379767827,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010127328235882915,
+      "loss": 0.7994,
+      "step": 10472
+    },
+    {
+      "epoch": 0.4962331201137171,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010125839095828286,
+      "loss": 0.0093,
+      "step": 10473
+    },
+    {
+      "epoch": 0.4962805022506515,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010124349952982681,
+      "loss": 0.5962,
+      "step": 10474
+    },
+    {
+      "epoch": 0.4963278843875859,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00010122860807379127,
+      "loss": 0.0048,
+      "step": 10475
+    },
+    {
+      "epoch": 0.49637526652452024,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00010121371659050652,
+      "loss": 0.0295,
+      "step": 10476
+    },
+    {
+      "epoch": 0.49642264866145464,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00010119882508030279,
+      "loss": 0.7112,
+      "step": 10477
+    },
+    {
+      "epoch": 0.49647003079838903,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001011839335435104,
+      "loss": 0.5813,
+      "step": 10478
+    },
+    {
+      "epoch": 0.49651741293532337,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00010116904198045963,
+      "loss": 0.0494,
+      "step": 10479
+    },
+    {
+      "epoch": 0.49656479507225776,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.00010115415039148078,
+      "loss": 0.0034,
+      "step": 10480
+    },
+    {
+      "epoch": 0.49661217720919215,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010113925877690406,
+      "loss": 0.2992,
+      "step": 10481
+    },
+    {
+      "epoch": 0.4966595593461265,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010112436713705984,
+      "loss": 0.7335,
+      "step": 10482
+    },
+    {
+      "epoch": 0.4967069414830609,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00010110947547227834,
+      "loss": 1.1745,
+      "step": 10483
+    },
+    {
+      "epoch": 0.4967543236199953,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010109458378288987,
+      "loss": 0.8676,
+      "step": 10484
+    },
+    {
+      "epoch": 0.4968017057569296,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00010107969206922465,
+      "loss": 0.0138,
+      "step": 10485
+    },
+    {
+      "epoch": 0.496849087893864,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.00010106480033161306,
+      "loss": 0.0081,
+      "step": 10486
+    },
+    {
+      "epoch": 0.4968964700307984,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001010499085703853,
+      "loss": 1.1629,
+      "step": 10487
+    },
+    {
+      "epoch": 0.4969438521677328,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010103501678587171,
+      "loss": 0.7232,
+      "step": 10488
+    },
+    {
+      "epoch": 0.49699123430466713,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00010102012497840252,
+      "loss": 1.1733,
+      "step": 10489
+    },
+    {
+      "epoch": 0.4970386164416015,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010100523314830806,
+      "loss": 0.9117,
+      "step": 10490
+    },
+    {
+      "epoch": 0.4970859985785359,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010099034129591862,
+      "loss": 0.9889,
+      "step": 10491
+    },
+    {
+      "epoch": 0.49713338071547025,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010097544942156445,
+      "loss": 1.1874,
+      "step": 10492
+    },
+    {
+      "epoch": 0.49718076285240465,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010096055752557583,
+      "loss": 0.947,
+      "step": 10493
+    },
+    {
+      "epoch": 0.49722814498933904,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0001009456656082831,
+      "loss": 0.0447,
+      "step": 10494
+    },
+    {
+      "epoch": 0.4972755271262734,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00010093077367001649,
+      "loss": 0.0465,
+      "step": 10495
+    },
+    {
+      "epoch": 0.49732290926320777,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010091588171110631,
+      "loss": 1.3243,
+      "step": 10496
+    },
+    {
+      "epoch": 0.49737029140014216,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010090098973188286,
+      "loss": 0.1363,
+      "step": 10497
+    },
+    {
+      "epoch": 0.4974176735370765,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00010088609773267643,
+      "loss": 0.1587,
+      "step": 10498
+    },
+    {
+      "epoch": 0.4974650556740109,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001008712057138173,
+      "loss": 0.0312,
+      "step": 10499
+    },
+    {
+      "epoch": 0.4975124378109453,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010085631367563573,
+      "loss": 0.3897,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4975598199478796,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00010084142161846202,
+      "loss": 0.0112,
+      "step": 10501
+    },
+    {
+      "epoch": 0.497607202084814,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001008265295426265,
+      "loss": 0.7306,
+      "step": 10502
+    },
+    {
+      "epoch": 0.4976545842217484,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010081163744845945,
+      "loss": 0.805,
+      "step": 10503
+    },
+    {
+      "epoch": 0.4977019663586828,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010079674533629111,
+      "loss": 0.584,
+      "step": 10504
+    },
+    {
+      "epoch": 0.49774934849561714,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00010078185320645182,
+      "loss": 0.773,
+      "step": 10505
+    },
+    {
+      "epoch": 0.49779673063255153,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 0.00010076696105927188,
+      "loss": 0.0028,
+      "step": 10506
+    },
+    {
+      "epoch": 0.4978441127694859,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00010075206889508155,
+      "loss": 0.0104,
+      "step": 10507
+    },
+    {
+      "epoch": 0.49789149490642026,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010073717671421113,
+      "loss": 0.1281,
+      "step": 10508
+    },
+    {
+      "epoch": 0.49793887704335466,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001007222845169909,
+      "loss": 0.1277,
+      "step": 10509
+    },
+    {
+      "epoch": 0.49798625918028905,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001007073923037512,
+      "loss": 1.2249,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4980336413172234,
+      "grad_norm": 0.0257568359375,
+      "learning_rate": 0.00010069250007482227,
+      "loss": 0.002,
+      "step": 10511
+    },
+    {
+      "epoch": 0.4980810234541578,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010067760783053441,
+      "loss": 1.2454,
+      "step": 10512
+    },
+    {
+      "epoch": 0.49812840559109217,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010066271557121795,
+      "loss": 1.1108,
+      "step": 10513
+    },
+    {
+      "epoch": 0.4981757877280265,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010064782329720317,
+      "loss": 1.4221,
+      "step": 10514
+    },
+    {
+      "epoch": 0.4982231698649609,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010063293100882038,
+      "loss": 0.5729,
+      "step": 10515
+    },
+    {
+      "epoch": 0.4982705520018953,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00010061803870639981,
+      "loss": 0.0268,
+      "step": 10516
+    },
+    {
+      "epoch": 0.4983179341388297,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010060314639027182,
+      "loss": 0.2148,
+      "step": 10517
+    },
+    {
+      "epoch": 0.498365316275764,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00010058825406076671,
+      "loss": 0.1574,
+      "step": 10518
+    },
+    {
+      "epoch": 0.4984126984126984,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001005733617182147,
+      "loss": 0.1309,
+      "step": 10519
+    },
+    {
+      "epoch": 0.4984600805496328,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010055846936294616,
+      "loss": 0.2592,
+      "step": 10520
+    },
+    {
+      "epoch": 0.49850746268656715,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00010054357699529138,
+      "loss": 0.1505,
+      "step": 10521
+    },
+    {
+      "epoch": 0.49855484482350154,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00010052868461558063,
+      "loss": 0.7928,
+      "step": 10522
+    },
+    {
+      "epoch": 0.49860222696043593,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00010051379222414423,
+      "loss": 0.0357,
+      "step": 10523
+    },
+    {
+      "epoch": 0.49864960909737027,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00010049889982131244,
+      "loss": 1.3586,
+      "step": 10524
+    },
+    {
+      "epoch": 0.49869699123430467,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00010048400740741559,
+      "loss": 0.0079,
+      "step": 10525
+    },
+    {
+      "epoch": 0.49874437337123906,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000100469114982784,
+      "loss": 0.4986,
+      "step": 10526
+    },
+    {
+      "epoch": 0.4987917555081734,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001004542225477479,
+      "loss": 0.6466,
+      "step": 10527
+    },
+    {
+      "epoch": 0.4988391376451078,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.00010043933010263764,
+      "loss": 0.0078,
+      "step": 10528
+    },
+    {
+      "epoch": 0.4988865197820422,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001004244376477835,
+      "loss": 1.187,
+      "step": 10529
+    },
+    {
+      "epoch": 0.4989339019189765,
+      "grad_norm": 0.03515625,
+      "learning_rate": 0.0001004095451835158,
+      "loss": 0.0023,
+      "step": 10530
+    },
+    {
+      "epoch": 0.4989812840559109,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00010039465271016484,
+      "loss": 0.1877,
+      "step": 10531
+    },
+    {
+      "epoch": 0.4990286661928453,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00010037976022806088,
+      "loss": 1.3169,
+      "step": 10532
+    },
+    {
+      "epoch": 0.4990760483297797,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00010036486773753427,
+      "loss": 0.0134,
+      "step": 10533
+    },
+    {
+      "epoch": 0.49912343046671404,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010034997523891523,
+      "loss": 1.1287,
+      "step": 10534
+    },
+    {
+      "epoch": 0.49917081260364843,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00010033508273253414,
+      "loss": 1.1797,
+      "step": 10535
+    },
+    {
+      "epoch": 0.4992181947405828,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010032019021872126,
+      "loss": 0.028,
+      "step": 10536
+    },
+    {
+      "epoch": 0.49926557687751716,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010030529769780693,
+      "loss": 0.7847,
+      "step": 10537
+    },
+    {
+      "epoch": 0.49931295901445155,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010029040517012143,
+      "loss": 0.0235,
+      "step": 10538
+    },
+    {
+      "epoch": 0.49936034115138594,
+      "grad_norm": 0.022705078125,
+      "learning_rate": 0.00010027551263599501,
+      "loss": 0.0018,
+      "step": 10539
+    },
+    {
+      "epoch": 0.4994077232883203,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00010026062009575803,
+      "loss": 0.8512,
+      "step": 10540
+    },
+    {
+      "epoch": 0.4994551054252547,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010024572754974078,
+      "loss": 0.8281,
+      "step": 10541
+    },
+    {
+      "epoch": 0.49950248756218907,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00010023083499827355,
+      "loss": 1.0409,
+      "step": 10542
+    },
+    {
+      "epoch": 0.4995498696991234,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010021594244168664,
+      "loss": 0.1338,
+      "step": 10543
+    },
+    {
+      "epoch": 0.4995972518360578,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00010020104988031038,
+      "loss": 0.6374,
+      "step": 10544
+    },
+    {
+      "epoch": 0.4996446339729922,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010018615731447502,
+      "loss": 0.1551,
+      "step": 10545
+    },
+    {
+      "epoch": 0.4996920161099266,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001001712647445109,
+      "loss": 1.5011,
+      "step": 10546
+    },
+    {
+      "epoch": 0.4997393982468609,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00010015637217074832,
+      "loss": 1.4456,
+      "step": 10547
+    },
+    {
+      "epoch": 0.4997867803837953,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00010014147959351754,
+      "loss": 0.0076,
+      "step": 10548
+    },
+    {
+      "epoch": 0.4998341625207297,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00010012658701314894,
+      "loss": 0.3427,
+      "step": 10549
+    },
+    {
+      "epoch": 0.49988154465766405,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00010011169442997274,
+      "loss": 0.7347,
+      "step": 10550
+    },
+    {
+      "epoch": 0.49992892679459844,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00010009680184431927,
+      "loss": 0.1643,
+      "step": 10551
+    },
+    {
+      "epoch": 0.49997630893153283,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010008190925651888,
+      "loss": 0.1251,
+      "step": 10552
+    },
+    {
+      "epoch": 0.5000236910684672,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010006701666690179,
+      "loss": 1.2011,
+      "step": 10553
+    },
+    {
+      "epoch": 0.5000236910684672,
+      "eval_loss": 0.6681665182113647,
+      "eval_runtime": 1301.5041,
+      "eval_samples_per_second": 1.803,
+      "eval_steps_per_second": 1.803,
+      "step": 10553
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 21105,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10553,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.436552471964549e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}