diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12503 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9991005172026086,
+  "eval_steps": 500,
+  "global_step": 1778,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001124353496739375,
+      "grad_norm": 53.2423355941896,
+      "learning_rate": 3.7453183520599254e-08,
+      "loss": 1.8864,
+      "step": 1
+    },
+    {
+      "epoch": 0.00224870699347875,
+      "grad_norm": 65.08392945079504,
+      "learning_rate": 7.490636704119851e-08,
+      "loss": 1.9475,
+      "step": 2
+    },
+    {
+      "epoch": 0.0033730604902181245,
+      "grad_norm": 72.44650682698767,
+      "learning_rate": 1.1235955056179776e-07,
+      "loss": 1.9625,
+      "step": 3
+    },
+    {
+      "epoch": 0.0044974139869575,
+      "grad_norm": 65.3800202843672,
+      "learning_rate": 1.4981273408239702e-07,
+      "loss": 1.9919,
+      "step": 4
+    },
+    {
+      "epoch": 0.005621767483696874,
+      "grad_norm": 59.64270650349274,
+      "learning_rate": 1.8726591760299626e-07,
+      "loss": 2.0113,
+      "step": 5
+    },
+    {
+      "epoch": 0.006746120980436249,
+      "grad_norm": 67.71118897590156,
+      "learning_rate": 2.247191011235955e-07,
+      "loss": 2.0824,
+      "step": 6
+    },
+    {
+      "epoch": 0.007870474477175624,
+      "grad_norm": 74.03465149487954,
+      "learning_rate": 2.621722846441948e-07,
+      "loss": 2.1524,
+      "step": 7
+    },
+    {
+      "epoch": 0.008994827973915,
+      "grad_norm": 74.28964831472538,
+      "learning_rate": 2.9962546816479403e-07,
+      "loss": 2.0717,
+      "step": 8
+    },
+    {
+      "epoch": 0.010119181470654375,
+      "grad_norm": 57.838350349644685,
+      "learning_rate": 3.3707865168539325e-07,
+      "loss": 1.8721,
+      "step": 9
+    },
+    {
+      "epoch": 0.011243534967393748,
+      "grad_norm": 57.71146890677253,
+      "learning_rate": 3.7453183520599253e-07,
+      "loss": 2.027,
+      "step": 10
+    },
+    {
+      "epoch": 0.012367888464133123,
+      "grad_norm": 69.87473269969144,
+      "learning_rate": 4.1198501872659175e-07,
+      "loss": 1.9844,
+      "step": 11
+    },
+    {
+      "epoch": 0.013492241960872498,
+      "grad_norm": 50.26660919893619,
+      "learning_rate": 4.49438202247191e-07,
+      "loss": 1.9979,
+      "step": 12
+    },
+    {
+      "epoch": 0.014616595457611873,
+      "grad_norm": 49.11304098237808,
+      "learning_rate": 4.868913857677903e-07,
+      "loss": 1.8724,
+      "step": 13
+    },
+    {
+      "epoch": 0.01574094895435125,
+      "grad_norm": 39.82309581617159,
+      "learning_rate": 5.243445692883896e-07,
+      "loss": 1.8431,
+      "step": 14
+    },
+    {
+      "epoch": 0.016865302451090622,
+      "grad_norm": 42.79235546997139,
+      "learning_rate": 5.617977528089888e-07,
+      "loss": 1.7794,
+      "step": 15
+    },
+    {
+      "epoch": 0.01798965594783,
+      "grad_norm": 25.80050345626229,
+      "learning_rate": 5.992509363295881e-07,
+      "loss": 1.5616,
+      "step": 16
+    },
+    {
+      "epoch": 0.019114009444569372,
+      "grad_norm": 26.203485882933155,
+      "learning_rate": 6.367041198501874e-07,
+      "loss": 1.4985,
+      "step": 17
+    },
+    {
+      "epoch": 0.02023836294130875,
+      "grad_norm": 24.04197210537073,
+      "learning_rate": 6.741573033707865e-07,
+      "loss": 1.4796,
+      "step": 18
+    },
+    {
+      "epoch": 0.021362716438048122,
+      "grad_norm": 24.52841009725287,
+      "learning_rate": 7.116104868913857e-07,
+      "loss": 1.5855,
+      "step": 19
+    },
+    {
+      "epoch": 0.022487069934787496,
+      "grad_norm": 20.15888184882102,
+      "learning_rate": 7.490636704119851e-07,
+      "loss": 1.3599,
+      "step": 20
+    },
+    {
+      "epoch": 0.023611423431526873,
+      "grad_norm": 21.314015322730018,
+      "learning_rate": 7.865168539325843e-07,
+      "loss": 1.4994,
+      "step": 21
+    },
+    {
+      "epoch": 0.024735776928266246,
+      "grad_norm": 13.526833351659915,
+      "learning_rate": 8.239700374531835e-07,
+      "loss": 1.1872,
+      "step": 22
+    },
+    {
+      "epoch": 0.025860130425005623,
+      "grad_norm": 11.701799629186937,
+      "learning_rate": 8.614232209737828e-07,
+      "loss": 1.1049,
+      "step": 23
+    },
+    {
+      "epoch": 0.026984483921744996,
+      "grad_norm": 11.434118929225738,
+      "learning_rate": 8.98876404494382e-07,
+      "loss": 1.1204,
+      "step": 24
+    },
+    {
+      "epoch": 0.02810883741848437,
+      "grad_norm": 11.472564461292894,
+      "learning_rate": 9.363295880149814e-07,
+      "loss": 1.1627,
+      "step": 25
+    },
+    {
+      "epoch": 0.029233190915223747,
+      "grad_norm": 11.850637822554562,
+      "learning_rate": 9.737827715355806e-07,
+      "loss": 1.1754,
+      "step": 26
+    },
+    {
+      "epoch": 0.03035754441196312,
+      "grad_norm": 10.656990075900731,
+      "learning_rate": 1.01123595505618e-06,
+      "loss": 1.0052,
+      "step": 27
+    },
+    {
+      "epoch": 0.0314818979087025,
+      "grad_norm": 11.031616603632587,
+      "learning_rate": 1.0486891385767792e-06,
+      "loss": 0.9034,
+      "step": 28
+    },
+    {
+      "epoch": 0.03260625140544187,
+      "grad_norm": 10.318457932150741,
+      "learning_rate": 1.0861423220973784e-06,
+      "loss": 1.02,
+      "step": 29
+    },
+    {
+      "epoch": 0.033730604902181244,
+      "grad_norm": 8.797502255120557,
+      "learning_rate": 1.1235955056179777e-06,
+      "loss": 1.1181,
+      "step": 30
+    },
+    {
+      "epoch": 0.034854958398920624,
+      "grad_norm": 9.187711762416285,
+      "learning_rate": 1.161048689138577e-06,
+      "loss": 0.8618,
+      "step": 31
+    },
+    {
+      "epoch": 0.03597931189566,
+      "grad_norm": 8.7327554704921,
+      "learning_rate": 1.1985018726591761e-06,
+      "loss": 0.8726,
+      "step": 32
+    },
+    {
+      "epoch": 0.03710366539239937,
+      "grad_norm": 9.670047861557682,
+      "learning_rate": 1.2359550561797752e-06,
+      "loss": 0.9227,
+      "step": 33
+    },
+    {
+      "epoch": 0.038228018889138744,
+      "grad_norm": 8.261045384008728,
+      "learning_rate": 1.2734082397003748e-06,
+      "loss": 0.8633,
+      "step": 34
+    },
+    {
+      "epoch": 0.03935237238587812,
+      "grad_norm": 8.845813755680256,
+      "learning_rate": 1.3108614232209737e-06,
+      "loss": 0.8651,
+      "step": 35
+    },
+    {
+      "epoch": 0.0404767258826175,
+      "grad_norm": 7.329128177442364,
+      "learning_rate": 1.348314606741573e-06,
+      "loss": 0.7379,
+      "step": 36
+    },
+    {
+      "epoch": 0.04160107937935687,
+      "grad_norm": 7.343014245842069,
+      "learning_rate": 1.3857677902621726e-06,
+      "loss": 0.7474,
+      "step": 37
+    },
+    {
+      "epoch": 0.042725432876096245,
+      "grad_norm": 7.278527070452587,
+      "learning_rate": 1.4232209737827715e-06,
+      "loss": 0.7821,
+      "step": 38
+    },
+    {
+      "epoch": 0.04384978637283562,
+      "grad_norm": 8.091875033335016,
+      "learning_rate": 1.4606741573033708e-06,
+      "loss": 0.6743,
+      "step": 39
+    },
+    {
+      "epoch": 0.04497413986957499,
+      "grad_norm": 7.88923740246497,
+      "learning_rate": 1.4981273408239701e-06,
+      "loss": 0.7435,
+      "step": 40
+    },
+    {
+      "epoch": 0.04609849336631437,
+      "grad_norm": 7.2086651203714736,
+      "learning_rate": 1.5355805243445692e-06,
+      "loss": 0.7162,
+      "step": 41
+    },
+    {
+      "epoch": 0.047222846863053745,
+      "grad_norm": 6.271935469070975,
+      "learning_rate": 1.5730337078651686e-06,
+      "loss": 0.6928,
+      "step": 42
+    },
+    {
+      "epoch": 0.04834720035979312,
+      "grad_norm": 8.873664226547213,
+      "learning_rate": 1.6104868913857679e-06,
+      "loss": 0.8006,
+      "step": 43
+    },
+    {
+      "epoch": 0.04947155385653249,
+      "grad_norm": 6.687649050407341,
+      "learning_rate": 1.647940074906367e-06,
+      "loss": 0.7871,
+      "step": 44
+    },
+    {
+      "epoch": 0.050595907353271866,
+      "grad_norm": 8.12909996850433,
+      "learning_rate": 1.6853932584269663e-06,
+      "loss": 0.7537,
+      "step": 45
+    },
+    {
+      "epoch": 0.051720260850011246,
+      "grad_norm": 7.8705575562746795,
+      "learning_rate": 1.7228464419475657e-06,
+      "loss": 0.6455,
+      "step": 46
+    },
+    {
+      "epoch": 0.05284461434675062,
+      "grad_norm": 6.840958076332606,
+      "learning_rate": 1.760299625468165e-06,
+      "loss": 0.7425,
+      "step": 47
+    },
+    {
+      "epoch": 0.05396896784348999,
+      "grad_norm": 7.405240136379491,
+      "learning_rate": 1.797752808988764e-06,
+      "loss": 0.7467,
+      "step": 48
+    },
+    {
+      "epoch": 0.055093321340229366,
+      "grad_norm": 6.544486388927425,
+      "learning_rate": 1.8352059925093634e-06,
+      "loss": 0.8065,
+      "step": 49
+    },
+    {
+      "epoch": 0.05621767483696874,
+      "grad_norm": 6.934595247488222,
+      "learning_rate": 1.8726591760299627e-06,
+      "loss": 0.6468,
+      "step": 50
+    },
+    {
+      "epoch": 0.05734202833370812,
+      "grad_norm": 6.710081574173267,
+      "learning_rate": 1.910112359550562e-06,
+      "loss": 0.7033,
+      "step": 51
+    },
+    {
+      "epoch": 0.05846638183044749,
+      "grad_norm": 7.165513882280082,
+      "learning_rate": 1.947565543071161e-06,
+      "loss": 0.6365,
+      "step": 52
+    },
+    {
+      "epoch": 0.05959073532718687,
+      "grad_norm": 7.099843175881044,
+      "learning_rate": 1.9850187265917605e-06,
+      "loss": 0.6835,
+      "step": 53
+    },
+    {
+      "epoch": 0.06071508882392624,
+      "grad_norm": 6.934261573057071,
+      "learning_rate": 2.02247191011236e-06,
+      "loss": 0.6676,
+      "step": 54
+    },
+    {
+      "epoch": 0.06183944232066562,
+      "grad_norm": 7.2843254803203585,
+      "learning_rate": 2.059925093632959e-06,
+      "loss": 0.6602,
+      "step": 55
+    },
+    {
+      "epoch": 0.062963795817405,
+      "grad_norm": 7.970702303394735,
+      "learning_rate": 2.0973782771535585e-06,
+      "loss": 0.7774,
+      "step": 56
+    },
+    {
+      "epoch": 0.06408814931414436,
+      "grad_norm": 6.864546024216213,
+      "learning_rate": 2.1348314606741574e-06,
+      "loss": 0.664,
+      "step": 57
+    },
+    {
+      "epoch": 0.06521250281088374,
+      "grad_norm": 6.566997545065713,
+      "learning_rate": 2.1722846441947567e-06,
+      "loss": 0.6582,
+      "step": 58
+    },
+    {
+      "epoch": 0.06633685630762312,
+      "grad_norm": 6.608498688172469,
+      "learning_rate": 2.209737827715356e-06,
+      "loss": 0.6127,
+      "step": 59
+    },
+    {
+      "epoch": 0.06746120980436249,
+      "grad_norm": 5.939446915980758,
+      "learning_rate": 2.2471910112359554e-06,
+      "loss": 0.6815,
+      "step": 60
+    },
+    {
+      "epoch": 0.06858556330110187,
+      "grad_norm": 5.853150773421264,
+      "learning_rate": 2.2846441947565547e-06,
+      "loss": 0.6235,
+      "step": 61
+    },
+    {
+      "epoch": 0.06970991679784125,
+      "grad_norm": 5.694449038555023,
+      "learning_rate": 2.322097378277154e-06,
+      "loss": 0.7204,
+      "step": 62
+    },
+    {
+      "epoch": 0.07083427029458061,
+      "grad_norm": 5.463176166232917,
+      "learning_rate": 2.359550561797753e-06,
+      "loss": 0.6382,
+      "step": 63
+    },
+    {
+      "epoch": 0.07195862379132,
+      "grad_norm": 5.293766938534352,
+      "learning_rate": 2.3970037453183523e-06,
+      "loss": 0.5587,
+      "step": 64
+    },
+    {
+      "epoch": 0.07308297728805936,
+      "grad_norm": 5.579637334370734,
+      "learning_rate": 2.4344569288389516e-06,
+      "loss": 0.646,
+      "step": 65
+    },
+    {
+      "epoch": 0.07420733078479874,
+      "grad_norm": 6.322004920119401,
+      "learning_rate": 2.4719101123595505e-06,
+      "loss": 0.6823,
+      "step": 66
+    },
+    {
+      "epoch": 0.07533168428153812,
+      "grad_norm": 5.551514753024309,
+      "learning_rate": 2.5093632958801502e-06,
+      "loss": 0.6108,
+      "step": 67
+    },
+    {
+      "epoch": 0.07645603777827749,
+      "grad_norm": 4.976280117772823,
+      "learning_rate": 2.5468164794007496e-06,
+      "loss": 0.6966,
+      "step": 68
+    },
+    {
+      "epoch": 0.07758039127501687,
+      "grad_norm": 5.846087373838301,
+      "learning_rate": 2.584269662921349e-06,
+      "loss": 0.7005,
+      "step": 69
+    },
+    {
+      "epoch": 0.07870474477175624,
+      "grad_norm": 6.157949281610723,
+      "learning_rate": 2.6217228464419474e-06,
+      "loss": 0.6004,
+      "step": 70
+    },
+    {
+      "epoch": 0.07982909826849562,
+      "grad_norm": 5.9060845607021655,
+      "learning_rate": 2.6591760299625467e-06,
+      "loss": 0.5526,
+      "step": 71
+    },
+    {
+      "epoch": 0.080953451765235,
+      "grad_norm": 5.679463318022605,
+      "learning_rate": 2.696629213483146e-06,
+      "loss": 0.7142,
+      "step": 72
+    },
+    {
+      "epoch": 0.08207780526197436,
+      "grad_norm": 6.1105333008818965,
+      "learning_rate": 2.7340823970037454e-06,
+      "loss": 0.6712,
+      "step": 73
+    },
+    {
+      "epoch": 0.08320215875871374,
+      "grad_norm": 6.238187135950297,
+      "learning_rate": 2.771535580524345e-06,
+      "loss": 0.5821,
+      "step": 74
+    },
+    {
+      "epoch": 0.08432651225545311,
+      "grad_norm": 5.445913353072334,
+      "learning_rate": 2.8089887640449444e-06,
+      "loss": 0.76,
+      "step": 75
+    },
+    {
+      "epoch": 0.08545086575219249,
+      "grad_norm": 5.299687468305153,
+      "learning_rate": 2.846441947565543e-06,
+      "loss": 0.6289,
+      "step": 76
+    },
+    {
+      "epoch": 0.08657521924893187,
+      "grad_norm": 5.553941365147732,
+      "learning_rate": 2.8838951310861422e-06,
+      "loss": 0.6584,
+      "step": 77
+    },
+    {
+      "epoch": 0.08769957274567124,
+      "grad_norm": 5.4331655411285205,
+      "learning_rate": 2.9213483146067416e-06,
+      "loss": 0.652,
+      "step": 78
+    },
+    {
+      "epoch": 0.08882392624241062,
+      "grad_norm": 5.161451719433795,
+      "learning_rate": 2.958801498127341e-06,
+      "loss": 0.634,
+      "step": 79
+    },
+    {
+      "epoch": 0.08994827973914998,
+      "grad_norm": 5.611741410478026,
+      "learning_rate": 2.9962546816479402e-06,
+      "loss": 0.6636,
+      "step": 80
+    },
+    {
+      "epoch": 0.09107263323588936,
+      "grad_norm": 5.644411971221104,
+      "learning_rate": 3.03370786516854e-06,
+      "loss": 0.5983,
+      "step": 81
+    },
+    {
+      "epoch": 0.09219698673262874,
+      "grad_norm": 5.014404661764506,
+      "learning_rate": 3.0711610486891385e-06,
+      "loss": 0.6765,
+      "step": 82
+    },
+    {
+      "epoch": 0.09332134022936811,
+      "grad_norm": 5.911837749965389,
+      "learning_rate": 3.1086142322097378e-06,
+      "loss": 0.6848,
+      "step": 83
+    },
+    {
+      "epoch": 0.09444569372610749,
+      "grad_norm": 5.708354128107416,
+      "learning_rate": 3.146067415730337e-06,
+      "loss": 0.6426,
+      "step": 84
+    },
+    {
+      "epoch": 0.09557004722284686,
+      "grad_norm": 4.893188287267555,
+      "learning_rate": 3.1835205992509364e-06,
+      "loss": 0.5243,
+      "step": 85
+    },
+    {
+      "epoch": 0.09669440071958624,
+      "grad_norm": 5.764913450962946,
+      "learning_rate": 3.2209737827715358e-06,
+      "loss": 0.5477,
+      "step": 86
+    },
+    {
+      "epoch": 0.09781875421632562,
+      "grad_norm": 6.039605612413023,
+      "learning_rate": 3.258426966292135e-06,
+      "loss": 0.5067,
+      "step": 87
+    },
+    {
+      "epoch": 0.09894310771306498,
+      "grad_norm": 5.354897770660215,
+      "learning_rate": 3.295880149812734e-06,
+      "loss": 0.5766,
+      "step": 88
+    },
+    {
+      "epoch": 0.10006746120980436,
+      "grad_norm": 5.83983058694163,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6313,
+      "step": 89
+    },
+    {
+      "epoch": 0.10119181470654373,
+      "grad_norm": 5.949494754823609,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 0.6605,
+      "step": 90
+    },
+    {
+      "epoch": 0.10231616820328311,
+      "grad_norm": 4.942395167784771,
+      "learning_rate": 3.408239700374532e-06,
+      "loss": 0.5966,
+      "step": 91
+    },
+    {
+      "epoch": 0.10344052170002249,
+      "grad_norm": 6.06929059729908,
+      "learning_rate": 3.4456928838951313e-06,
+      "loss": 0.6215,
+      "step": 92
+    },
+    {
+      "epoch": 0.10456487519676186,
+      "grad_norm": 5.244106286188616,
+      "learning_rate": 3.4831460674157306e-06,
+      "loss": 0.6825,
+      "step": 93
+    },
+    {
+      "epoch": 0.10568922869350124,
+      "grad_norm": 5.094869573072037,
+      "learning_rate": 3.52059925093633e-06,
+      "loss": 0.5012,
+      "step": 94
+    },
+    {
+      "epoch": 0.1068135821902406,
+      "grad_norm": 5.138572066730571,
+      "learning_rate": 3.558052434456929e-06,
+      "loss": 0.6167,
+      "step": 95
+    },
+    {
+      "epoch": 0.10793793568697999,
+      "grad_norm": 4.877752958803828,
+      "learning_rate": 3.595505617977528e-06,
+      "loss": 0.5689,
+      "step": 96
+    },
+    {
+      "epoch": 0.10906228918371937,
+      "grad_norm": 5.888860237783254,
+      "learning_rate": 3.6329588014981275e-06,
+      "loss": 0.6549,
+      "step": 97
+    },
+    {
+      "epoch": 0.11018664268045873,
+      "grad_norm": 5.429514792045115,
+      "learning_rate": 3.670411985018727e-06,
+      "loss": 0.6668,
+      "step": 98
+    },
+    {
+      "epoch": 0.11131099617719811,
+      "grad_norm": 5.106904521603865,
+      "learning_rate": 3.707865168539326e-06,
+      "loss": 0.5734,
+      "step": 99
+    },
+    {
+      "epoch": 0.11243534967393748,
+      "grad_norm": 5.5268677009641864,
+      "learning_rate": 3.7453183520599255e-06,
+      "loss": 0.5639,
+      "step": 100
+    },
+    {
+      "epoch": 0.11355970317067686,
+      "grad_norm": 4.984436470976369,
+      "learning_rate": 3.7827715355805244e-06,
+      "loss": 0.564,
+      "step": 101
+    },
+    {
+      "epoch": 0.11468405666741624,
+      "grad_norm": 5.082541352752938,
+      "learning_rate": 3.820224719101124e-06,
+      "loss": 0.5994,
+      "step": 102
+    },
+    {
+      "epoch": 0.1158084101641556,
+      "grad_norm": 5.047459689379145,
+      "learning_rate": 3.857677902621723e-06,
+      "loss": 0.5351,
+      "step": 103
+    },
+    {
+      "epoch": 0.11693276366089499,
+      "grad_norm": 7.99051557319798,
+      "learning_rate": 3.895131086142322e-06,
+      "loss": 0.6144,
+      "step": 104
+    },
+    {
+      "epoch": 0.11805711715763437,
+      "grad_norm": 4.684329531241078,
+      "learning_rate": 3.932584269662922e-06,
+      "loss": 0.5728,
+      "step": 105
+    },
+    {
+      "epoch": 0.11918147065437373,
+      "grad_norm": 6.025957710262137,
+      "learning_rate": 3.970037453183521e-06,
+      "loss": 0.565,
+      "step": 106
+    },
+    {
+      "epoch": 0.12030582415111311,
+      "grad_norm": 5.132311248248364,
+      "learning_rate": 4.00749063670412e-06,
+      "loss": 0.5464,
+      "step": 107
+    },
+    {
+      "epoch": 0.12143017764785248,
+      "grad_norm": 4.806664335260601,
+      "learning_rate": 4.04494382022472e-06,
+      "loss": 0.6467,
+      "step": 108
+    },
+    {
+      "epoch": 0.12255453114459186,
+      "grad_norm": 5.7394880996427435,
+      "learning_rate": 4.082397003745319e-06,
+      "loss": 0.6527,
+      "step": 109
+    },
+    {
+      "epoch": 0.12367888464133124,
+      "grad_norm": 5.377294272836532,
+      "learning_rate": 4.119850187265918e-06,
+      "loss": 0.6041,
+      "step": 110
+    },
+    {
+      "epoch": 0.12480323813807061,
+      "grad_norm": 5.186073895525204,
+      "learning_rate": 4.157303370786518e-06,
+      "loss": 0.7023,
+      "step": 111
+    },
+    {
+      "epoch": 0.12592759163481,
+      "grad_norm": 5.624740693261092,
+      "learning_rate": 4.194756554307117e-06,
+      "loss": 0.6066,
+      "step": 112
+    },
+    {
+      "epoch": 0.12705194513154935,
+      "grad_norm": 5.086317181863122,
+      "learning_rate": 4.2322097378277155e-06,
+      "loss": 0.5791,
+      "step": 113
+    },
+    {
+      "epoch": 0.12817629862828872,
+      "grad_norm": 5.039978381517061,
+      "learning_rate": 4.269662921348315e-06,
+      "loss": 0.621,
+      "step": 114
+    },
+    {
+      "epoch": 0.12930065212502811,
+      "grad_norm": 4.818095221041318,
+      "learning_rate": 4.307116104868914e-06,
+      "loss": 0.6249,
+      "step": 115
+    },
+    {
+      "epoch": 0.13042500562176748,
+      "grad_norm": 5.462516915720135,
+      "learning_rate": 4.3445692883895135e-06,
+      "loss": 0.6256,
+      "step": 116
+    },
+    {
+      "epoch": 0.13154935911850685,
+      "grad_norm": 5.108735065599773,
+      "learning_rate": 4.382022471910113e-06,
+      "loss": 0.6705,
+      "step": 117
+    },
+    {
+      "epoch": 0.13267371261524624,
+      "grad_norm": 5.171753125741451,
+      "learning_rate": 4.419475655430712e-06,
+      "loss": 0.6198,
+      "step": 118
+    },
+    {
+      "epoch": 0.1337980661119856,
+      "grad_norm": 4.54818066892063,
+      "learning_rate": 4.456928838951311e-06,
+      "loss": 0.5518,
+      "step": 119
+    },
+    {
+      "epoch": 0.13492241960872497,
+      "grad_norm": 5.849926479469034,
+      "learning_rate": 4.494382022471911e-06,
+      "loss": 0.625,
+      "step": 120
+    },
+    {
+      "epoch": 0.13604677310546437,
+      "grad_norm": 5.3015940275376865,
+      "learning_rate": 4.53183520599251e-06,
+      "loss": 0.5265,
+      "step": 121
+    },
+    {
+      "epoch": 0.13717112660220374,
+      "grad_norm": 5.59979109980021,
+      "learning_rate": 4.569288389513109e-06,
+      "loss": 0.5965,
+      "step": 122
+    },
+    {
+      "epoch": 0.1382954800989431,
+      "grad_norm": 5.154111435063985,
+      "learning_rate": 4.606741573033709e-06,
+      "loss": 0.5961,
+      "step": 123
+    },
+    {
+      "epoch": 0.1394198335956825,
+      "grad_norm": 4.948647748915688,
+      "learning_rate": 4.644194756554308e-06,
+      "loss": 0.5678,
+      "step": 124
+    },
+    {
+      "epoch": 0.14054418709242186,
+      "grad_norm": 4.544922794800689,
+      "learning_rate": 4.6816479400749066e-06,
+      "loss": 0.4738,
+      "step": 125
+    },
+    {
+      "epoch": 0.14166854058916123,
+      "grad_norm": 5.236067341309945,
+      "learning_rate": 4.719101123595506e-06,
+      "loss": 0.5947,
+      "step": 126
+    },
+    {
+      "epoch": 0.1427928940859006,
+      "grad_norm": 4.826783118036744,
+      "learning_rate": 4.756554307116105e-06,
+      "loss": 0.5347,
+      "step": 127
+    },
+    {
+      "epoch": 0.14391724758264,
+      "grad_norm": 4.704180120015617,
+      "learning_rate": 4.7940074906367045e-06,
+      "loss": 0.6399,
+      "step": 128
+    },
+    {
+      "epoch": 0.14504160107937936,
+      "grad_norm": 5.490297304272613,
+      "learning_rate": 4.831460674157304e-06,
+      "loss": 0.6433,
+      "step": 129
+    },
+    {
+      "epoch": 0.14616595457611872,
+      "grad_norm": 5.597276756765826,
+      "learning_rate": 4.868913857677903e-06,
+      "loss": 0.5079,
+      "step": 130
+    },
+    {
+      "epoch": 0.14729030807285812,
+      "grad_norm": 6.142430081038773,
+      "learning_rate": 4.906367041198502e-06,
+      "loss": 0.6011,
+      "step": 131
+    },
+    {
+      "epoch": 0.14841466156959748,
+      "grad_norm": 4.861985353801883,
+      "learning_rate": 4.943820224719101e-06,
+      "loss": 0.5919,
+      "step": 132
+    },
+    {
+      "epoch": 0.14953901506633685,
+      "grad_norm": 4.952392826442031,
+      "learning_rate": 4.9812734082397e-06,
+      "loss": 0.6396,
+      "step": 133
+    },
+    {
+      "epoch": 0.15066336856307624,
+      "grad_norm": 4.5098501209174575,
+      "learning_rate": 5.0187265917603005e-06,
+      "loss": 0.5804,
+      "step": 134
+    },
+    {
+      "epoch": 0.1517877220598156,
+      "grad_norm": 5.088465269073771,
+      "learning_rate": 5.0561797752809e-06,
+      "loss": 0.461,
+      "step": 135
+    },
+    {
+      "epoch": 0.15291207555655498,
+      "grad_norm": 4.757052247782732,
+      "learning_rate": 5.093632958801499e-06,
+      "loss": 0.5137,
+      "step": 136
+    },
+    {
+      "epoch": 0.15403642905329434,
+      "grad_norm": 4.7362045140752365,
+      "learning_rate": 5.1310861423220985e-06,
+      "loss": 0.6035,
+      "step": 137
+    },
+    {
+      "epoch": 0.15516078255003374,
+      "grad_norm": 5.079378333364324,
+      "learning_rate": 5.168539325842698e-06,
+      "loss": 0.4788,
+      "step": 138
+    },
+    {
+      "epoch": 0.1562851360467731,
+      "grad_norm": 4.630122042780841,
+      "learning_rate": 5.205992509363297e-06,
+      "loss": 0.5402,
+      "step": 139
+    },
+    {
+      "epoch": 0.15740948954351247,
+      "grad_norm": 4.169387084600574,
+      "learning_rate": 5.243445692883895e-06,
+      "loss": 0.6173,
+      "step": 140
+    },
+    {
+      "epoch": 0.15853384304025187,
+      "grad_norm": 5.074286982165006,
+      "learning_rate": 5.280898876404494e-06,
+      "loss": 0.4692,
+      "step": 141
+    },
+    {
+      "epoch": 0.15965819653699123,
+      "grad_norm": 4.76739190242339,
+      "learning_rate": 5.318352059925093e-06,
+      "loss": 0.5484,
+      "step": 142
+    },
+    {
+      "epoch": 0.1607825500337306,
+      "grad_norm": 4.785934252885911,
+      "learning_rate": 5.355805243445693e-06,
+      "loss": 0.5434,
+      "step": 143
+    },
+    {
+      "epoch": 0.16190690353047,
+      "grad_norm": 4.642988496442961,
+      "learning_rate": 5.393258426966292e-06,
+      "loss": 0.5554,
+      "step": 144
+    },
+    {
+      "epoch": 0.16303125702720936,
+      "grad_norm": 4.655576985657423,
+      "learning_rate": 5.430711610486891e-06,
+      "loss": 0.5102,
+      "step": 145
+    },
+    {
+      "epoch": 0.16415561052394873,
+      "grad_norm": 6.086203389161473,
+      "learning_rate": 5.468164794007491e-06,
+      "loss": 0.6202,
+      "step": 146
+    },
+    {
+      "epoch": 0.1652799640206881,
+      "grad_norm": 5.0402882247319125,
+      "learning_rate": 5.50561797752809e-06,
+      "loss": 0.5939,
+      "step": 147
+    },
+    {
+      "epoch": 0.16640431751742749,
+      "grad_norm": 4.497111778655106,
+      "learning_rate": 5.54307116104869e-06,
+      "loss": 0.5565,
+      "step": 148
+    },
+    {
+      "epoch": 0.16752867101416685,
+      "grad_norm": 5.235237283270832,
+      "learning_rate": 5.5805243445692896e-06,
+      "loss": 0.5727,
+      "step": 149
+    },
+    {
+      "epoch": 0.16865302451090622,
+      "grad_norm": 5.3088096720025995,
+      "learning_rate": 5.617977528089889e-06,
+      "loss": 0.6494,
+      "step": 150
+    },
+    {
+      "epoch": 0.1697773780076456,
+      "grad_norm": 4.864768665027932,
+      "learning_rate": 5.655430711610488e-06,
+      "loss": 0.6014,
+      "step": 151
+    },
+    {
+      "epoch": 0.17090173150438498,
+      "grad_norm": 4.6225900435975005,
+      "learning_rate": 5.692883895131086e-06,
+      "loss": 0.5816,
+      "step": 152
+    },
+    {
+      "epoch": 0.17202608500112435,
+      "grad_norm": 5.343904927661094,
+      "learning_rate": 5.730337078651685e-06,
+      "loss": 0.4638,
+      "step": 153
+    },
+    {
+      "epoch": 0.17315043849786374,
+      "grad_norm": 4.318318961028793,
+      "learning_rate": 5.7677902621722845e-06,
+      "loss": 0.494,
+      "step": 154
+    },
+    {
+      "epoch": 0.1742747919946031,
+      "grad_norm": 5.021623359651704,
+      "learning_rate": 5.805243445692884e-06,
+      "loss": 0.57,
+      "step": 155
+    },
+    {
+      "epoch": 0.17539914549134247,
+      "grad_norm": 4.050028398746291,
+      "learning_rate": 5.842696629213483e-06,
+      "loss": 0.4865,
+      "step": 156
+    },
+    {
+      "epoch": 0.17652349898808184,
+      "grad_norm": 4.611781482767637,
+      "learning_rate": 5.8801498127340825e-06,
+      "loss": 0.5352,
+      "step": 157
+    },
+    {
+      "epoch": 0.17764785248482123,
+      "grad_norm": 5.003220329441464,
+      "learning_rate": 5.917602996254682e-06,
+      "loss": 0.6152,
+      "step": 158
+    },
+    {
+      "epoch": 0.1787722059815606,
+      "grad_norm": 5.098547077319575,
+      "learning_rate": 5.955056179775281e-06,
+      "loss": 0.5488,
+      "step": 159
+    },
+    {
+      "epoch": 0.17989655947829997,
+      "grad_norm": 5.334665493475377,
+      "learning_rate": 5.9925093632958805e-06,
+      "loss": 0.5059,
+      "step": 160
+    },
+    {
+      "epoch": 0.18102091297503936,
+      "grad_norm": 4.684206508603612,
+      "learning_rate": 6.02996254681648e-06,
+      "loss": 0.7165,
+      "step": 161
+    },
+    {
+      "epoch": 0.18214526647177873,
+      "grad_norm": 4.908121516097129,
+      "learning_rate": 6.06741573033708e-06,
+      "loss": 0.5892,
+      "step": 162
+    },
+    {
+      "epoch": 0.1832696199685181,
+      "grad_norm": 4.958238729962288,
+      "learning_rate": 6.104868913857679e-06,
+      "loss": 0.6456,
+      "step": 163
+    },
+    {
+      "epoch": 0.1843939734652575,
+      "grad_norm": 4.779466296718993,
+      "learning_rate": 6.142322097378277e-06,
+      "loss": 0.588,
+      "step": 164
+    },
+    {
+      "epoch": 0.18551832696199685,
+      "grad_norm": 4.292763620456942,
+      "learning_rate": 6.179775280898876e-06,
+      "loss": 0.5423,
+      "step": 165
+    },
+    {
+      "epoch": 0.18664268045873622,
+      "grad_norm": 4.264077571351202,
+      "learning_rate": 6.2172284644194756e-06,
+      "loss": 0.4763,
+      "step": 166
+    },
+    {
+      "epoch": 0.18776703395547562,
+      "grad_norm": 4.823174865538302,
+      "learning_rate": 6.254681647940075e-06,
+      "loss": 0.6491,
+      "step": 167
+    },
+    {
+      "epoch": 0.18889138745221498,
+      "grad_norm": 4.628880906485794,
+      "learning_rate": 6.292134831460674e-06,
+      "loss": 0.617,
+      "step": 168
+    },
+    {
+      "epoch": 0.19001574094895435,
+      "grad_norm": 4.319639394942016,
+      "learning_rate": 6.3295880149812736e-06,
+      "loss": 0.4859,
+      "step": 169
+    },
+    {
+      "epoch": 0.19114009444569371,
+      "grad_norm": 4.11891918198944,
+      "learning_rate": 6.367041198501873e-06,
+      "loss": 0.4923,
+      "step": 170
+    },
+    {
+      "epoch": 0.1922644479424331,
+      "grad_norm": 5.018793729344863,
+      "learning_rate": 6.404494382022472e-06,
+      "loss": 0.6702,
+      "step": 171
+    },
+    {
+      "epoch": 0.19338880143917248,
+      "grad_norm": 4.1987673124188465,
+      "learning_rate": 6.4419475655430715e-06,
+      "loss": 0.5599,
+      "step": 172
+    },
+    {
+      "epoch": 0.19451315493591184,
+      "grad_norm": 5.13953209472435,
+      "learning_rate": 6.479400749063671e-06,
+      "loss": 0.57,
+      "step": 173
+    },
+    {
+      "epoch": 0.19563750843265124,
+      "grad_norm": 4.107317974226564,
+      "learning_rate": 6.51685393258427e-06,
+      "loss": 0.5367,
+      "step": 174
+    },
+    {
+      "epoch": 0.1967618619293906,
+      "grad_norm": 4.185320718459167,
+      "learning_rate": 6.5543071161048695e-06,
+      "loss": 0.5542,
+      "step": 175
+    },
+    {
+      "epoch": 0.19788621542612997,
+      "grad_norm": 5.281181959074794,
+      "learning_rate": 6.591760299625468e-06,
+      "loss": 0.594,
+      "step": 176
+    },
+    {
+      "epoch": 0.19901056892286936,
+      "grad_norm": 5.320778776430486,
+      "learning_rate": 6.629213483146067e-06,
+      "loss": 0.5439,
+      "step": 177
+    },
+    {
+      "epoch": 0.20013492241960873,
+      "grad_norm": 5.214596075525528,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5463,
+      "step": 178
+    },
+    {
+      "epoch": 0.2012592759163481,
+      "grad_norm": 4.652440619846462,
+      "learning_rate": 6.704119850187266e-06,
+      "loss": 0.5755,
+      "step": 179
+    },
+    {
+      "epoch": 0.20238362941308746,
+      "grad_norm": 4.810779739675451,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 0.5009,
+      "step": 180
+    },
+    {
+      "epoch": 0.20350798290982686,
+      "grad_norm": 4.773031954142573,
+      "learning_rate": 6.779026217228465e-06,
+      "loss": 0.6315,
+      "step": 181
+    },
+    {
+      "epoch": 0.20463233640656622,
+      "grad_norm": 5.241968899268721,
+      "learning_rate": 6.816479400749064e-06,
+      "loss": 0.6664,
+      "step": 182
+    },
+    {
+      "epoch": 0.2057566899033056,
+      "grad_norm": 4.848623432014871,
+      "learning_rate": 6.853932584269663e-06,
+      "loss": 0.6505,
+      "step": 183
+    },
+    {
+      "epoch": 0.20688104340004498,
+      "grad_norm": 4.971156396411526,
+      "learning_rate": 6.891385767790263e-06,
+      "loss": 0.4748,
+      "step": 184
+    },
+    {
+      "epoch": 0.20800539689678435,
+      "grad_norm": 4.952015573513049,
+      "learning_rate": 6.928838951310862e-06,
+      "loss": 0.5848,
+      "step": 185
+    },
+    {
+      "epoch": 0.20912975039352372,
+      "grad_norm": 4.429808036239594,
+      "learning_rate": 6.966292134831461e-06,
+      "loss": 0.4813,
+      "step": 186
+    },
+    {
+      "epoch": 0.2102541038902631,
+      "grad_norm": 4.391513166719795,
+      "learning_rate": 7.003745318352061e-06,
+      "loss": 0.5475,
+      "step": 187
+    },
+    {
+      "epoch": 0.21137845738700248,
+      "grad_norm": 4.7139563769461885,
+      "learning_rate": 7.04119850187266e-06,
+      "loss": 0.4408,
+      "step": 188
+    },
+    {
+      "epoch": 0.21250281088374184,
+      "grad_norm": 5.02283673988439,
+      "learning_rate": 7.078651685393258e-06,
+      "loss": 0.5411,
+      "step": 189
+    },
+    {
+      "epoch": 0.2136271643804812,
+      "grad_norm": 4.380825767226325,
+      "learning_rate": 7.116104868913858e-06,
+      "loss": 0.548,
+      "step": 190
+    },
+    {
+      "epoch": 0.2147515178772206,
+      "grad_norm": 3.70383460256294,
+      "learning_rate": 7.153558052434457e-06,
+      "loss": 0.5468,
+      "step": 191
+    },
+    {
+      "epoch": 0.21587587137395997,
+      "grad_norm": 4.556248869958154,
+      "learning_rate": 7.191011235955056e-06,
+      "loss": 0.5326,
+      "step": 192
+    },
+    {
+      "epoch": 0.21700022487069934,
+      "grad_norm": 3.893671687894249,
+      "learning_rate": 7.228464419475656e-06,
+      "loss": 0.4541,
+      "step": 193
+    },
+    {
+      "epoch": 0.21812457836743873,
+      "grad_norm": 4.303457964169383,
+      "learning_rate": 7.265917602996255e-06,
+      "loss": 0.524,
+      "step": 194
+    },
+    {
+      "epoch": 0.2192489318641781,
+      "grad_norm": 4.972085430933498,
+      "learning_rate": 7.303370786516854e-06,
+      "loss": 0.6246,
+      "step": 195
+    },
+    {
+      "epoch": 0.22037328536091746,
+      "grad_norm": 4.571440416133472,
+      "learning_rate": 7.340823970037454e-06,
+      "loss": 0.4889,
+      "step": 196
+    },
+    {
+      "epoch": 0.22149763885765686,
+      "grad_norm": 4.417206989640032,
+      "learning_rate": 7.378277153558053e-06,
+      "loss": 0.5764,
+      "step": 197
+    },
+    {
+      "epoch": 0.22262199235439623,
+      "grad_norm": 4.250669060781474,
+      "learning_rate": 7.415730337078652e-06,
+      "loss": 0.4301,
+      "step": 198
+    },
+    {
+      "epoch": 0.2237463458511356,
+      "grad_norm": 4.625856949352494,
+      "learning_rate": 7.453183520599252e-06,
+      "loss": 0.5542,
+      "step": 199
+    },
+    {
+      "epoch": 0.22487069934787496,
+      "grad_norm": 4.788012617861282,
+      "learning_rate": 7.490636704119851e-06,
+      "loss": 0.5446,
+      "step": 200
+    },
+    {
+      "epoch": 0.22599505284461435,
+      "grad_norm": 4.530600955368931,
+      "learning_rate": 7.5280898876404495e-06,
+      "loss": 0.5434,
+      "step": 201
+    },
+    {
+      "epoch": 0.22711940634135372,
+      "grad_norm": 4.7895409778915425,
+      "learning_rate": 7.565543071161049e-06,
+      "loss": 0.6097,
+      "step": 202
+    },
+    {
+      "epoch": 0.22824375983809309,
+      "grad_norm": 4.906915635791381,
+      "learning_rate": 7.602996254681648e-06,
+      "loss": 0.5349,
+      "step": 203
+    },
+    {
+      "epoch": 0.22936811333483248,
+      "grad_norm": 4.3461611477847795,
+      "learning_rate": 7.640449438202247e-06,
+      "loss": 0.6184,
+      "step": 204
+    },
+    {
+      "epoch": 0.23049246683157185,
+      "grad_norm": 4.465360589820918,
+      "learning_rate": 7.677902621722848e-06,
+      "loss": 0.6002,
+      "step": 205
+    },
+    {
+      "epoch": 0.2316168203283112,
+      "grad_norm": 4.340565386065813,
+      "learning_rate": 7.715355805243446e-06,
+      "loss": 0.4929,
+      "step": 206
+    },
+    {
+      "epoch": 0.2327411738250506,
+      "grad_norm": 4.501670803112346,
+      "learning_rate": 7.752808988764046e-06,
+      "loss": 0.5644,
+      "step": 207
+    },
+    {
+      "epoch": 0.23386552732178997,
+      "grad_norm": 4.833038514989987,
+      "learning_rate": 7.790262172284645e-06,
+      "loss": 0.5538,
+      "step": 208
+    },
+    {
+      "epoch": 0.23498988081852934,
+      "grad_norm": 4.2705827108077115,
+      "learning_rate": 7.827715355805245e-06,
+      "loss": 0.612,
+      "step": 209
+    },
+    {
+      "epoch": 0.23611423431526873,
+      "grad_norm": 4.110911438083281,
+      "learning_rate": 7.865168539325843e-06,
+      "loss": 0.4774,
+      "step": 210
+    },
+    {
+      "epoch": 0.2372385878120081,
+      "grad_norm": 4.0313057036917215,
+      "learning_rate": 7.902621722846444e-06,
+      "loss": 0.5598,
+      "step": 211
+    },
+    {
+      "epoch": 0.23836294130874747,
+      "grad_norm": 4.606277096635832,
+      "learning_rate": 7.940074906367042e-06,
+      "loss": 0.6216,
+      "step": 212
+    },
+    {
+      "epoch": 0.23948729480548683,
+      "grad_norm": 5.281558987314689,
+      "learning_rate": 7.97752808988764e-06,
+      "loss": 0.5605,
+      "step": 213
+    },
+    {
+      "epoch": 0.24061164830222623,
+      "grad_norm": 4.299946862106653,
+      "learning_rate": 8.01498127340824e-06,
+      "loss": 0.456,
+      "step": 214
+    },
+    {
+      "epoch": 0.2417360017989656,
+      "grad_norm": 3.8955773478913405,
+      "learning_rate": 8.05243445692884e-06,
+      "loss": 0.4956,
+      "step": 215
+    },
+    {
+      "epoch": 0.24286035529570496,
+      "grad_norm": 4.015042470825777,
+      "learning_rate": 8.08988764044944e-06,
+      "loss": 0.5535,
+      "step": 216
+    },
+    {
+      "epoch": 0.24398470879244435,
+      "grad_norm": 4.503948900930498,
+      "learning_rate": 8.127340823970038e-06,
+      "loss": 0.5186,
+      "step": 217
+    },
+    {
+      "epoch": 0.24510906228918372,
+      "grad_norm": 3.7921975002398476,
+      "learning_rate": 8.164794007490638e-06,
+      "loss": 0.5,
+      "step": 218
+    },
+    {
+      "epoch": 0.2462334157859231,
+      "grad_norm": 4.70338864020324,
+      "learning_rate": 8.202247191011237e-06,
+      "loss": 0.6157,
+      "step": 219
+    },
+    {
+      "epoch": 0.24735776928266248,
+      "grad_norm": 4.013256202087387,
+      "learning_rate": 8.239700374531837e-06,
+      "loss": 0.4968,
+      "step": 220
+    },
+    {
+      "epoch": 0.24848212277940185,
+      "grad_norm": 4.3105739557884695,
+      "learning_rate": 8.277153558052435e-06,
+      "loss": 0.5341,
+      "step": 221
+    },
+    {
+      "epoch": 0.24960647627614121,
+      "grad_norm": 4.578931297370858,
+      "learning_rate": 8.314606741573035e-06,
+      "loss": 0.5683,
+      "step": 222
+    },
+    {
+      "epoch": 0.2507308297728806,
+      "grad_norm": 4.748627500659751,
+      "learning_rate": 8.352059925093634e-06,
+      "loss": 0.5695,
+      "step": 223
+    },
+    {
+      "epoch": 0.25185518326962,
+      "grad_norm": 4.446605841681016,
+      "learning_rate": 8.389513108614234e-06,
+      "loss": 0.5799,
+      "step": 224
+    },
+    {
+      "epoch": 0.25297953676635937,
+      "grad_norm": 4.593866761040076,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 0.5941,
+      "step": 225
+    },
+    {
+      "epoch": 0.2541038902630987,
+      "grad_norm": 3.8711733100955414,
+      "learning_rate": 8.464419475655431e-06,
+      "loss": 0.5386,
+      "step": 226
+    },
+    {
+      "epoch": 0.2552282437598381,
+      "grad_norm": 3.890310791082027,
+      "learning_rate": 8.501872659176031e-06,
+      "loss": 0.4871,
+      "step": 227
+    },
+    {
+      "epoch": 0.25635259725657744,
+      "grad_norm": 4.296114345272867,
+      "learning_rate": 8.53932584269663e-06,
+      "loss": 0.5798,
+      "step": 228
+    },
+    {
+      "epoch": 0.25747695075331684,
+      "grad_norm": 4.586223892652054,
+      "learning_rate": 8.57677902621723e-06,
+      "loss": 0.6279,
+      "step": 229
+    },
+    {
+      "epoch": 0.25860130425005623,
+      "grad_norm": 4.148198350500805,
+      "learning_rate": 8.614232209737828e-06,
+      "loss": 0.6231,
+      "step": 230
+    },
+    {
+      "epoch": 0.25972565774679557,
+      "grad_norm": 4.573292033027529,
+      "learning_rate": 8.651685393258428e-06,
+      "loss": 0.5551,
+      "step": 231
+    },
+    {
+      "epoch": 0.26085001124353496,
+      "grad_norm": 4.638393482042473,
+      "learning_rate": 8.689138576779027e-06,
+      "loss": 0.508,
+      "step": 232
+    },
+    {
+      "epoch": 0.26197436474027436,
+      "grad_norm": 4.409712202788947,
+      "learning_rate": 8.726591760299627e-06,
+      "loss": 0.6135,
+      "step": 233
+    },
+    {
+      "epoch": 0.2630987182370137,
+      "grad_norm": 3.9650338316034817,
+      "learning_rate": 8.764044943820226e-06,
+      "loss": 0.4562,
+      "step": 234
+    },
+    {
+      "epoch": 0.2642230717337531,
+      "grad_norm": 4.312805070319012,
+      "learning_rate": 8.801498127340826e-06,
+      "loss": 0.4772,
+      "step": 235
+    },
+    {
+      "epoch": 0.2653474252304925,
+      "grad_norm": 4.758192695674082,
+      "learning_rate": 8.838951310861424e-06,
+      "loss": 0.5495,
+      "step": 236
+    },
+    {
+      "epoch": 0.2664717787272318,
+      "grad_norm": 4.708116443238244,
+      "learning_rate": 8.876404494382023e-06,
+      "loss": 0.53,
+      "step": 237
+    },
+    {
+      "epoch": 0.2675961322239712,
+      "grad_norm": 3.9644987312032662,
+      "learning_rate": 8.913857677902621e-06,
+      "loss": 0.4995,
+      "step": 238
+    },
+    {
+      "epoch": 0.2687204857207106,
+      "grad_norm": 4.548883672169617,
+      "learning_rate": 8.951310861423221e-06,
+      "loss": 0.5302,
+      "step": 239
+    },
+    {
+      "epoch": 0.26984483921744995,
+      "grad_norm": 4.149058402359693,
+      "learning_rate": 8.988764044943822e-06,
+      "loss": 0.509,
+      "step": 240
+    },
+    {
+      "epoch": 0.27096919271418934,
+      "grad_norm": 4.461515283523925,
+      "learning_rate": 9.02621722846442e-06,
+      "loss": 0.5404,
+      "step": 241
+    },
+    {
+      "epoch": 0.27209354621092874,
+      "grad_norm": 4.381615736495519,
+      "learning_rate": 9.06367041198502e-06,
+      "loss": 0.5426,
+      "step": 242
+    },
+    {
+      "epoch": 0.2732178997076681,
+      "grad_norm": 4.462072731487522,
+      "learning_rate": 9.101123595505619e-06,
+      "loss": 0.651,
+      "step": 243
+    },
+    {
+      "epoch": 0.27434225320440747,
+      "grad_norm": 4.0161521024079505,
+      "learning_rate": 9.138576779026219e-06,
+      "loss": 0.5761,
+      "step": 244
+    },
+    {
+      "epoch": 0.27546660670114687,
+      "grad_norm": 4.462397980487328,
+      "learning_rate": 9.176029962546817e-06,
+      "loss": 0.421,
+      "step": 245
+    },
+    {
+      "epoch": 0.2765909601978862,
+      "grad_norm": 4.305452167701929,
+      "learning_rate": 9.213483146067417e-06,
+      "loss": 0.5739,
+      "step": 246
+    },
+    {
+      "epoch": 0.2777153136946256,
+      "grad_norm": 5.864377723641452,
+      "learning_rate": 9.250936329588016e-06,
+      "loss": 0.4908,
+      "step": 247
+    },
+    {
+      "epoch": 0.278839667191365,
+      "grad_norm": 4.282672417699527,
+      "learning_rate": 9.288389513108616e-06,
+      "loss": 0.6605,
+      "step": 248
+    },
+    {
+      "epoch": 0.27996402068810433,
+      "grad_norm": 4.53264055867988,
+      "learning_rate": 9.325842696629213e-06,
+      "loss": 0.4628,
+      "step": 249
+    },
+    {
+      "epoch": 0.2810883741848437,
+      "grad_norm": 3.6072680040744185,
+      "learning_rate": 9.363295880149813e-06,
+      "loss": 0.4527,
+      "step": 250
+    },
+    {
+      "epoch": 0.28221272768158306,
+      "grad_norm": 4.538679049369346,
+      "learning_rate": 9.400749063670412e-06,
+      "loss": 0.7116,
+      "step": 251
+    },
+    {
+      "epoch": 0.28333708117832246,
+      "grad_norm": 5.8944801497385635,
+      "learning_rate": 9.438202247191012e-06,
+      "loss": 0.5627,
+      "step": 252
+    },
+    {
+      "epoch": 0.28446143467506185,
+      "grad_norm": 4.3732547599106475,
+      "learning_rate": 9.475655430711612e-06,
+      "loss": 0.6226,
+      "step": 253
+    },
+    {
+      "epoch": 0.2855857881718012,
+      "grad_norm": 4.454244032390872,
+      "learning_rate": 9.51310861423221e-06,
+      "loss": 0.527,
+      "step": 254
+    },
+    {
+      "epoch": 0.2867101416685406,
+      "grad_norm": 3.8981780370466517,
+      "learning_rate": 9.55056179775281e-06,
+      "loss": 0.5651,
+      "step": 255
+    },
+    {
+      "epoch": 0.28783449516528,
+      "grad_norm": 4.709113966619958,
+      "learning_rate": 9.588014981273409e-06,
+      "loss": 0.6613,
+      "step": 256
+    },
+    {
+      "epoch": 0.2889588486620193,
+      "grad_norm": 4.4855187143305955,
+      "learning_rate": 9.62546816479401e-06,
+      "loss": 0.5508,
+      "step": 257
+    },
+    {
+      "epoch": 0.2900832021587587,
+      "grad_norm": 4.160325214820284,
+      "learning_rate": 9.662921348314608e-06,
+      "loss": 0.6448,
+      "step": 258
+    },
+    {
+      "epoch": 0.2912075556554981,
+      "grad_norm": 4.229291532322749,
+      "learning_rate": 9.700374531835208e-06,
+      "loss": 0.5802,
+      "step": 259
+    },
+    {
+      "epoch": 0.29233190915223745,
+      "grad_norm": 4.337243810374759,
+      "learning_rate": 9.737827715355806e-06,
+      "loss": 0.6201,
+      "step": 260
+    },
+    {
+      "epoch": 0.29345626264897684,
+      "grad_norm": 4.043229721393927,
+      "learning_rate": 9.775280898876405e-06,
+      "loss": 0.6032,
+      "step": 261
+    },
+    {
+      "epoch": 0.29458061614571623,
+      "grad_norm": 4.295722654791335,
+      "learning_rate": 9.812734082397003e-06,
+      "loss": 0.613,
+      "step": 262
+    },
+    {
+      "epoch": 0.2957049696424556,
+      "grad_norm": 5.09163797590291,
+      "learning_rate": 9.850187265917604e-06,
+      "loss": 0.5175,
+      "step": 263
+    },
+    {
+      "epoch": 0.29682932313919497,
+      "grad_norm": 4.163263816361626,
+      "learning_rate": 9.887640449438202e-06,
+      "loss": 0.5927,
+      "step": 264
+    },
+    {
+      "epoch": 0.29795367663593436,
+      "grad_norm": 3.7129368090640993,
+      "learning_rate": 9.925093632958802e-06,
+      "loss": 0.4892,
+      "step": 265
+    },
+    {
+      "epoch": 0.2990780301326737,
+      "grad_norm": 3.92253047208784,
+      "learning_rate": 9.9625468164794e-06,
+      "loss": 0.5668,
+      "step": 266
+    },
+    {
+      "epoch": 0.3002023836294131,
+      "grad_norm": 4.381521742652477,
+      "learning_rate": 1e-05,
+      "loss": 0.5978,
+      "step": 267
+    },
+    {
+      "epoch": 0.3013267371261525,
+      "grad_norm": 4.168103255852118,
+      "learning_rate": 9.999995716318147e-06,
+      "loss": 0.4697,
+      "step": 268
+    },
+    {
+      "epoch": 0.3024510906228918,
+      "grad_norm": 4.269319207820486,
+      "learning_rate": 9.999982865279924e-06,
+      "loss": 0.5609,
+      "step": 269
+    },
+    {
+      "epoch": 0.3035754441196312,
+      "grad_norm": 4.100861732076085,
+      "learning_rate": 9.999961446907354e-06,
+      "loss": 0.5133,
+      "step": 270
+    },
+    {
+      "epoch": 0.30469979761637056,
+      "grad_norm": 4.401781654297728,
+      "learning_rate": 9.999931461237135e-06,
+      "loss": 0.5499,
+      "step": 271
+    },
+    {
+      "epoch": 0.30582415111310995,
+      "grad_norm": 3.887935987304003,
+      "learning_rate": 9.999892908320647e-06,
+      "loss": 0.5343,
+      "step": 272
+    },
+    {
+      "epoch": 0.30694850460984935,
+      "grad_norm": 3.9527647533512873,
+      "learning_rate": 9.99984578822395e-06,
+      "loss": 0.5699,
+      "step": 273
+    },
+    {
+      "epoch": 0.3080728581065887,
+      "grad_norm": 3.9517664742123007,
+      "learning_rate": 9.999790101027783e-06,
+      "loss": 0.5517,
+      "step": 274
+    },
+    {
+      "epoch": 0.3091972116033281,
+      "grad_norm": 3.2833604131239063,
+      "learning_rate": 9.999725846827562e-06,
+      "loss": 0.4553,
+      "step": 275
+    },
+    {
+      "epoch": 0.3103215651000675,
+      "grad_norm": 3.5160816796635603,
+      "learning_rate": 9.999653025733386e-06,
+      "loss": 0.4529,
+      "step": 276
+    },
+    {
+      "epoch": 0.3114459185968068,
+      "grad_norm": 4.2649696884035375,
+      "learning_rate": 9.999571637870035e-06,
+      "loss": 0.5979,
+      "step": 277
+    },
+    {
+      "epoch": 0.3125702720935462,
+      "grad_norm": 3.7733548623494877,
+      "learning_rate": 9.999481683376964e-06,
+      "loss": 0.4808,
+      "step": 278
+    },
+    {
+      "epoch": 0.3136946255902856,
+      "grad_norm": 3.8144030621853893,
+      "learning_rate": 9.999383162408303e-06,
+      "loss": 0.56,
+      "step": 279
+    },
+    {
+      "epoch": 0.31481897908702494,
+      "grad_norm": 4.154018976346022,
+      "learning_rate": 9.999276075132871e-06,
+      "loss": 0.5166,
+      "step": 280
+    },
+    {
+      "epoch": 0.31594333258376434,
+      "grad_norm": 3.8644675961587094,
+      "learning_rate": 9.999160421734155e-06,
+      "loss": 0.5775,
+      "step": 281
+    },
+    {
+      "epoch": 0.31706768608050373,
+      "grad_norm": 4.129419048503998,
+      "learning_rate": 9.999036202410324e-06,
+      "loss": 0.5611,
+      "step": 282
+    },
+    {
+      "epoch": 0.31819203957724307,
+      "grad_norm": 4.195616281547412,
+      "learning_rate": 9.998903417374228e-06,
+      "loss": 0.5572,
+      "step": 283
+    },
+    {
+      "epoch": 0.31931639307398246,
+      "grad_norm": 4.103799433610849,
+      "learning_rate": 9.998762066853388e-06,
+      "loss": 0.5842,
+      "step": 284
+    },
+    {
+      "epoch": 0.32044074657072186,
+      "grad_norm": 4.125720692215004,
+      "learning_rate": 9.998612151090004e-06,
+      "loss": 0.5977,
+      "step": 285
+    },
+    {
+      "epoch": 0.3215651000674612,
+      "grad_norm": 3.4548373480968135,
+      "learning_rate": 9.998453670340954e-06,
+      "loss": 0.5509,
+      "step": 286
+    },
+    {
+      "epoch": 0.3226894535642006,
+      "grad_norm": 3.551176331900102,
+      "learning_rate": 9.998286624877786e-06,
+      "loss": 0.5107,
+      "step": 287
+    },
+    {
+      "epoch": 0.32381380706094,
+      "grad_norm": 3.722204797034346,
+      "learning_rate": 9.998111014986735e-06,
+      "loss": 0.5731,
+      "step": 288
+    },
+    {
+      "epoch": 0.3249381605576793,
+      "grad_norm": 3.3248058919372814,
+      "learning_rate": 9.997926840968699e-06,
+      "loss": 0.5014,
+      "step": 289
+    },
+    {
+      "epoch": 0.3260625140544187,
+      "grad_norm": 4.064075270856846,
+      "learning_rate": 9.997734103139255e-06,
+      "loss": 0.5369,
+      "step": 290
+    },
+    {
+      "epoch": 0.3271868675511581,
+      "grad_norm": 4.023863796412585,
+      "learning_rate": 9.997532801828659e-06,
+      "loss": 0.545,
+      "step": 291
+    },
+    {
+      "epoch": 0.32831122104789745,
+      "grad_norm": 3.6502357159379053,
+      "learning_rate": 9.997322937381829e-06,
+      "loss": 0.5749,
+      "step": 292
+    },
+    {
+      "epoch": 0.32943557454463684,
+      "grad_norm": 3.6306275490467157,
+      "learning_rate": 9.997104510158365e-06,
+      "loss": 0.4948,
+      "step": 293
+    },
+    {
+      "epoch": 0.3305599280413762,
+      "grad_norm": 4.1281061898392215,
+      "learning_rate": 9.996877520532535e-06,
+      "loss": 0.4772,
+      "step": 294
+    },
+    {
+      "epoch": 0.3316842815381156,
+      "grad_norm": 3.882215960549001,
+      "learning_rate": 9.996641968893281e-06,
+      "loss": 0.5888,
+      "step": 295
+    },
+    {
+      "epoch": 0.33280863503485497,
+      "grad_norm": 3.625025877961329,
+      "learning_rate": 9.996397855644216e-06,
+      "loss": 0.5191,
+      "step": 296
+    },
+    {
+      "epoch": 0.3339329885315943,
+      "grad_norm": 4.712046788087188,
+      "learning_rate": 9.996145181203616e-06,
+      "loss": 0.5009,
+      "step": 297
+    },
+    {
+      "epoch": 0.3350573420283337,
+      "grad_norm": 4.250491181665793,
+      "learning_rate": 9.995883946004435e-06,
+      "loss": 0.5777,
+      "step": 298
+    },
+    {
+      "epoch": 0.3361816955250731,
+      "grad_norm": 4.141833172111204,
+      "learning_rate": 9.995614150494293e-06,
+      "loss": 0.5029,
+      "step": 299
+    },
+    {
+      "epoch": 0.33730604902181244,
+      "grad_norm": 3.830269568275683,
+      "learning_rate": 9.995335795135475e-06,
+      "loss": 0.5517,
+      "step": 300
+    },
+    {
+      "epoch": 0.33843040251855183,
+      "grad_norm": 3.9730846814318266,
+      "learning_rate": 9.995048880404939e-06,
+      "loss": 0.587,
+      "step": 301
+    },
+    {
+      "epoch": 0.3395547560152912,
+      "grad_norm": 4.3861735521839975,
+      "learning_rate": 9.994753406794303e-06,
+      "loss": 0.5252,
+      "step": 302
+    },
+    {
+      "epoch": 0.34067910951203056,
+      "grad_norm": 3.7096309944051007,
+      "learning_rate": 9.994449374809851e-06,
+      "loss": 0.6588,
+      "step": 303
+    },
+    {
+      "epoch": 0.34180346300876996,
+      "grad_norm": 4.535682021740793,
+      "learning_rate": 9.994136784972537e-06,
+      "loss": 0.5218,
+      "step": 304
+    },
+    {
+      "epoch": 0.34292781650550935,
+      "grad_norm": 3.9618390001667474,
+      "learning_rate": 9.993815637817974e-06,
+      "loss": 0.5505,
+      "step": 305
+    },
+    {
+      "epoch": 0.3440521700022487,
+      "grad_norm": 3.758687848142667,
+      "learning_rate": 9.99348593389644e-06,
+      "loss": 0.6028,
+      "step": 306
+    },
+    {
+      "epoch": 0.3451765234989881,
+      "grad_norm": 3.6963054520270706,
+      "learning_rate": 9.993147673772869e-06,
+      "loss": 0.5644,
+      "step": 307
+    },
+    {
+      "epoch": 0.3463008769957275,
+      "grad_norm": 4.410785253033696,
+      "learning_rate": 9.992800858026868e-06,
+      "loss": 0.6078,
+      "step": 308
+    },
+    {
+      "epoch": 0.3474252304924668,
+      "grad_norm": 3.812746638369813,
+      "learning_rate": 9.992445487252692e-06,
+      "loss": 0.5363,
+      "step": 309
+    },
+    {
+      "epoch": 0.3485495839892062,
+      "grad_norm": 3.8496149382193234,
+      "learning_rate": 9.992081562059258e-06,
+      "loss": 0.5709,
+      "step": 310
+    },
+    {
+      "epoch": 0.3496739374859456,
+      "grad_norm": 4.077436394406757,
+      "learning_rate": 9.991709083070143e-06,
+      "loss": 0.5563,
+      "step": 311
+    },
+    {
+      "epoch": 0.35079829098268495,
+      "grad_norm": 3.882045840443277,
+      "learning_rate": 9.99132805092358e-06,
+      "loss": 0.5496,
+      "step": 312
+    },
+    {
+      "epoch": 0.35192264447942434,
+      "grad_norm": 3.6846457424561145,
+      "learning_rate": 9.990938466272459e-06,
+      "loss": 0.5416,
+      "step": 313
+    },
+    {
+      "epoch": 0.3530469979761637,
+      "grad_norm": 3.789141540609254,
+      "learning_rate": 9.990540329784319e-06,
+      "loss": 0.4853,
+      "step": 314
+    },
+    {
+      "epoch": 0.3541713514729031,
+      "grad_norm": 3.3924673232632907,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 0.5558,
+      "step": 315
+    },
+    {
+      "epoch": 0.35529570496964247,
+      "grad_norm": 3.6900511728274066,
+      "learning_rate": 9.989718404040424e-06,
+      "loss": 0.6133,
+      "step": 316
+    },
+    {
+      "epoch": 0.3564200584663818,
+      "grad_norm": 4.326899139132089,
+      "learning_rate": 9.989294616193018e-06,
+      "loss": 0.5994,
+      "step": 317
+    },
+    {
+      "epoch": 0.3575444119631212,
+      "grad_norm": 3.8921338166057953,
+      "learning_rate": 9.988862279325287e-06,
+      "loss": 0.5304,
+      "step": 318
+    },
+    {
+      "epoch": 0.3586687654598606,
+      "grad_norm": 3.6709997128549143,
+      "learning_rate": 9.988421394178027e-06,
+      "loss": 0.6469,
+      "step": 319
+    },
+    {
+      "epoch": 0.35979311895659993,
+      "grad_norm": 4.245741967123779,
+      "learning_rate": 9.987971961506686e-06,
+      "loss": 0.5131,
+      "step": 320
+    },
+    {
+      "epoch": 0.3609174724533393,
+      "grad_norm": 4.392473015688166,
+      "learning_rate": 9.987513982081352e-06,
+      "loss": 0.5687,
+      "step": 321
+    },
+    {
+      "epoch": 0.3620418259500787,
+      "grad_norm": 3.53917359913007,
+      "learning_rate": 9.98704745668676e-06,
+      "loss": 0.4878,
+      "step": 322
+    },
+    {
+      "epoch": 0.36316617944681806,
+      "grad_norm": 3.9137510788759573,
+      "learning_rate": 9.98657238612229e-06,
+      "loss": 0.6372,
+      "step": 323
+    },
+    {
+      "epoch": 0.36429053294355745,
+      "grad_norm": 3.6599264673530287,
+      "learning_rate": 9.986088771201965e-06,
+      "loss": 0.5888,
+      "step": 324
+    },
+    {
+      "epoch": 0.36541488644029685,
+      "grad_norm": 3.829850616844456,
+      "learning_rate": 9.985596612754441e-06,
+      "loss": 0.547,
+      "step": 325
+    },
+    {
+      "epoch": 0.3665392399370362,
+      "grad_norm": 4.177899074196137,
+      "learning_rate": 9.98509591162302e-06,
+      "loss": 0.6349,
+      "step": 326
+    },
+    {
+      "epoch": 0.3676635934337756,
+      "grad_norm": 4.243955645751383,
+      "learning_rate": 9.984586668665641e-06,
+      "loss": 0.4762,
+      "step": 327
+    },
+    {
+      "epoch": 0.368787946930515,
+      "grad_norm": 3.7715994101788395,
+      "learning_rate": 9.984068884754876e-06,
+      "loss": 0.502,
+      "step": 328
+    },
+    {
+      "epoch": 0.3699123004272543,
+      "grad_norm": 4.931078709848011,
+      "learning_rate": 9.983542560777934e-06,
+      "loss": 0.5828,
+      "step": 329
+    },
+    {
+      "epoch": 0.3710366539239937,
+      "grad_norm": 3.9100675908015536,
+      "learning_rate": 9.983007697636659e-06,
+      "loss": 0.5447,
+      "step": 330
+    },
+    {
+      "epoch": 0.3721610074207331,
+      "grad_norm": 3.582028364093252,
+      "learning_rate": 9.982464296247523e-06,
+      "loss": 0.4565,
+      "step": 331
+    },
+    {
+      "epoch": 0.37328536091747244,
+      "grad_norm": 3.7227669278863904,
+      "learning_rate": 9.981912357541628e-06,
+      "loss": 0.5839,
+      "step": 332
+    },
+    {
+      "epoch": 0.37440971441421184,
+      "grad_norm": 3.6880893089035514,
+      "learning_rate": 9.981351882464706e-06,
+      "loss": 0.5087,
+      "step": 333
+    },
+    {
+      "epoch": 0.37553406791095123,
+      "grad_norm": 3.7293969753596565,
+      "learning_rate": 9.980782871977119e-06,
+      "loss": 0.6077,
+      "step": 334
+    },
+    {
+      "epoch": 0.37665842140769057,
+      "grad_norm": 3.763493533693305,
+      "learning_rate": 9.98020532705385e-06,
+      "loss": 0.5455,
+      "step": 335
+    },
+    {
+      "epoch": 0.37778277490442996,
+      "grad_norm": 3.8929584401119466,
+      "learning_rate": 9.979619248684503e-06,
+      "loss": 0.4603,
+      "step": 336
+    },
+    {
+      "epoch": 0.3789071284011693,
+      "grad_norm": 3.2210854385887817,
+      "learning_rate": 9.979024637873309e-06,
+      "loss": 0.443,
+      "step": 337
+    },
+    {
+      "epoch": 0.3800314818979087,
+      "grad_norm": 3.9062399313984826,
+      "learning_rate": 9.978421495639118e-06,
+      "loss": 0.5869,
+      "step": 338
+    },
+    {
+      "epoch": 0.3811558353946481,
+      "grad_norm": 4.019064942307236,
+      "learning_rate": 9.9778098230154e-06,
+      "loss": 0.5048,
+      "step": 339
+    },
+    {
+      "epoch": 0.38228018889138743,
+      "grad_norm": 3.7027619784974015,
+      "learning_rate": 9.977189621050237e-06,
+      "loss": 0.5524,
+      "step": 340
+    },
+    {
+      "epoch": 0.3834045423881268,
+      "grad_norm": 5.020373379711682,
+      "learning_rate": 9.976560890806328e-06,
+      "loss": 0.5512,
+      "step": 341
+    },
+    {
+      "epoch": 0.3845288958848662,
+      "grad_norm": 3.5340879536302645,
+      "learning_rate": 9.975923633360985e-06,
+      "loss": 0.4677,
+      "step": 342
+    },
+    {
+      "epoch": 0.38565324938160556,
+      "grad_norm": 3.493858419914277,
+      "learning_rate": 9.975277849806133e-06,
+      "loss": 0.5844,
+      "step": 343
+    },
+    {
+      "epoch": 0.38677760287834495,
+      "grad_norm": 3.945608256398267,
+      "learning_rate": 9.974623541248301e-06,
+      "loss": 0.5283,
+      "step": 344
+    },
+    {
+      "epoch": 0.38790195637508434,
+      "grad_norm": 3.991869817244121,
+      "learning_rate": 9.973960708808633e-06,
+      "loss": 0.4815,
+      "step": 345
+    },
+    {
+      "epoch": 0.3890263098718237,
+      "grad_norm": 4.139152021048733,
+      "learning_rate": 9.973289353622872e-06,
+      "loss": 0.6218,
+      "step": 346
+    },
+    {
+      "epoch": 0.3901506633685631,
+      "grad_norm": 3.195650030270729,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.4467,
+      "step": 347
+    },
+    {
+      "epoch": 0.39127501686530247,
+      "grad_norm": 4.006728125141407,
+      "learning_rate": 9.97192107962907e-06,
+      "loss": 0.5406,
+      "step": 348
+    },
+    {
+      "epoch": 0.3923993703620418,
+      "grad_norm": 4.907651964375414,
+      "learning_rate": 9.971224163165526e-06,
+      "loss": 0.5761,
+      "step": 349
+    },
+    {
+      "epoch": 0.3935237238587812,
+      "grad_norm": 4.639820623110264,
+      "learning_rate": 9.97051872864489e-06,
+      "loss": 0.5938,
+      "step": 350
+    },
+    {
+      "epoch": 0.3946480773555206,
+      "grad_norm": 3.7866762934125866,
+      "learning_rate": 9.9698047772759e-06,
+      "loss": 0.5199,
+      "step": 351
+    },
+    {
+      "epoch": 0.39577243085225994,
+      "grad_norm": 3.7243238078419805,
+      "learning_rate": 9.96908231028189e-06,
+      "loss": 0.4898,
+      "step": 352
+    },
+    {
+      "epoch": 0.39689678434899933,
+      "grad_norm": 4.315183422293352,
+      "learning_rate": 9.968351328900793e-06,
+      "loss": 0.5937,
+      "step": 353
+    },
+    {
+      "epoch": 0.3980211378457387,
+      "grad_norm": 3.930467892457228,
+      "learning_rate": 9.967611834385122e-06,
+      "loss": 0.5291,
+      "step": 354
+    },
+    {
+      "epoch": 0.39914549134247806,
+      "grad_norm": 4.7535233936012595,
+      "learning_rate": 9.966863828001982e-06,
+      "loss": 0.5183,
+      "step": 355
+    },
+    {
+      "epoch": 0.40026984483921746,
+      "grad_norm": 3.9214452529381987,
+      "learning_rate": 9.966107311033062e-06,
+      "loss": 0.4627,
+      "step": 356
+    },
+    {
+      "epoch": 0.4013941983359568,
+      "grad_norm": 3.5093090880923112,
+      "learning_rate": 9.965342284774633e-06,
+      "loss": 0.5681,
+      "step": 357
+    },
+    {
+      "epoch": 0.4025185518326962,
+      "grad_norm": 3.808123247767938,
+      "learning_rate": 9.964568750537545e-06,
+      "loss": 0.5437,
+      "step": 358
+    },
+    {
+      "epoch": 0.4036429053294356,
+      "grad_norm": 3.7355076524933892,
+      "learning_rate": 9.963786709647228e-06,
+      "loss": 0.639,
+      "step": 359
+    },
+    {
+      "epoch": 0.4047672588261749,
+      "grad_norm": 3.9337716270770295,
+      "learning_rate": 9.96299616344369e-06,
+      "loss": 0.5964,
+      "step": 360
+    },
+    {
+      "epoch": 0.4058916123229143,
+      "grad_norm": 4.629181101221476,
+      "learning_rate": 9.962197113281508e-06,
+      "loss": 0.5248,
+      "step": 361
+    },
+    {
+      "epoch": 0.4070159658196537,
+      "grad_norm": 3.6151623494971883,
+      "learning_rate": 9.961389560529835e-06,
+      "loss": 0.5258,
+      "step": 362
+    },
+    {
+      "epoch": 0.40814031931639305,
+      "grad_norm": 3.8752677967523463,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 0.5832,
+      "step": 363
+    },
+    {
+      "epoch": 0.40926467281313245,
+      "grad_norm": 4.004451874569043,
+      "learning_rate": 9.959748952807457e-06,
+      "loss": 0.5077,
+      "step": 364
+    },
+    {
+      "epoch": 0.41038902630987184,
+      "grad_norm": 3.888377753121046,
+      "learning_rate": 9.95891590064789e-06,
+      "loss": 0.5136,
+      "step": 365
+    },
+    {
+      "epoch": 0.4115133798066112,
+      "grad_norm": 3.3722057839165878,
+      "learning_rate": 9.958074351521097e-06,
+      "loss": 0.4776,
+      "step": 366
+    },
+    {
+      "epoch": 0.4126377333033506,
+      "grad_norm": 3.3144711839014755,
+      "learning_rate": 9.957224306869053e-06,
+      "loss": 0.4338,
+      "step": 367
+    },
+    {
+      "epoch": 0.41376208680008997,
+      "grad_norm": 3.769138773902626,
+      "learning_rate": 9.956365768148284e-06,
+      "loss": 0.5278,
+      "step": 368
+    },
+    {
+      "epoch": 0.4148864402968293,
+      "grad_norm": 3.9638656127063316,
+      "learning_rate": 9.955498736829876e-06,
+      "loss": 0.6808,
+      "step": 369
+    },
+    {
+      "epoch": 0.4160107937935687,
+      "grad_norm": 3.653095683955353,
+      "learning_rate": 9.954623214399457e-06,
+      "loss": 0.5802,
+      "step": 370
+    },
+    {
+      "epoch": 0.4171351472903081,
+      "grad_norm": 3.606045449814684,
+      "learning_rate": 9.953739202357219e-06,
+      "loss": 0.5237,
+      "step": 371
+    },
+    {
+      "epoch": 0.41825950078704743,
+      "grad_norm": 3.951197063499177,
+      "learning_rate": 9.952846702217886e-06,
+      "loss": 0.4824,
+      "step": 372
+    },
+    {
+      "epoch": 0.4193838542837868,
+      "grad_norm": 4.459953116414348,
+      "learning_rate": 9.951945715510738e-06,
+      "loss": 0.6024,
+      "step": 373
+    },
+    {
+      "epoch": 0.4205082077805262,
+      "grad_norm": 3.439589417940646,
+      "learning_rate": 9.951036243779586e-06,
+      "loss": 0.6106,
+      "step": 374
+    },
+    {
+      "epoch": 0.42163256127726556,
+      "grad_norm": 3.3238678144843643,
+      "learning_rate": 9.95011828858279e-06,
+      "loss": 0.4381,
+      "step": 375
+    },
+    {
+      "epoch": 0.42275691477400495,
+      "grad_norm": 3.5108420650584753,
+      "learning_rate": 9.949191851493236e-06,
+      "loss": 0.574,
+      "step": 376
+    },
+    {
+      "epoch": 0.42388126827074435,
+      "grad_norm": 3.287558077530219,
+      "learning_rate": 9.948256934098353e-06,
+      "loss": 0.5563,
+      "step": 377
+    },
+    {
+      "epoch": 0.4250056217674837,
+      "grad_norm": 3.8227718544297793,
+      "learning_rate": 9.947313538000093e-06,
+      "loss": 0.5728,
+      "step": 378
+    },
+    {
+      "epoch": 0.4261299752642231,
+      "grad_norm": 3.2917471456928062,
+      "learning_rate": 9.946361664814942e-06,
+      "loss": 0.5121,
+      "step": 379
+    },
+    {
+      "epoch": 0.4272543287609624,
+      "grad_norm": 3.5331970675121815,
+      "learning_rate": 9.945401316173908e-06,
+      "loss": 0.5844,
+      "step": 380
+    },
+    {
+      "epoch": 0.4283786822577018,
+      "grad_norm": 4.118712922624492,
+      "learning_rate": 9.944432493722525e-06,
+      "loss": 0.6904,
+      "step": 381
+    },
+    {
+      "epoch": 0.4295030357544412,
+      "grad_norm": 3.6617279020737614,
+      "learning_rate": 9.943455199120836e-06,
+      "loss": 0.6066,
+      "step": 382
+    },
+    {
+      "epoch": 0.43062738925118055,
+      "grad_norm": 3.4957295732309204,
+      "learning_rate": 9.942469434043418e-06,
+      "loss": 0.5735,
+      "step": 383
+    },
+    {
+      "epoch": 0.43175174274791994,
+      "grad_norm": 3.6017609565344695,
+      "learning_rate": 9.941475200179347e-06,
+      "loss": 0.6433,
+      "step": 384
+    },
+    {
+      "epoch": 0.43287609624465934,
+      "grad_norm": 4.1352865526064475,
+      "learning_rate": 9.940472499232218e-06,
+      "loss": 0.487,
+      "step": 385
+    },
+    {
+      "epoch": 0.4340004497413987,
+      "grad_norm": 3.3213760467298594,
+      "learning_rate": 9.93946133292013e-06,
+      "loss": 0.5198,
+      "step": 386
+    },
+    {
+      "epoch": 0.43512480323813807,
+      "grad_norm": 5.515434395849467,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.5712,
+      "step": 387
+    },
+    {
+      "epoch": 0.43624915673487746,
+      "grad_norm": 3.283970848186277,
+      "learning_rate": 9.937413611146006e-06,
+      "loss": 0.4786,
+      "step": 388
+    },
+    {
+      "epoch": 0.4373735102316168,
+      "grad_norm": 3.3337174276131294,
+      "learning_rate": 9.936377059192684e-06,
+      "loss": 0.4422,
+      "step": 389
+    },
+    {
+      "epoch": 0.4384978637283562,
+      "grad_norm": 3.8648427120880022,
+      "learning_rate": 9.935332048891828e-06,
+      "loss": 0.5823,
+      "step": 390
+    },
+    {
+      "epoch": 0.4396222172250956,
+      "grad_norm": 4.003650515793954,
+      "learning_rate": 9.934278582034037e-06,
+      "loss": 0.5526,
+      "step": 391
+    },
+    {
+      "epoch": 0.44074657072183493,
+      "grad_norm": 3.58595130257847,
+      "learning_rate": 9.933216660424396e-06,
+      "loss": 0.689,
+      "step": 392
+    },
+    {
+      "epoch": 0.4418709242185743,
+      "grad_norm": 3.7011150439390352,
+      "learning_rate": 9.932146285882478e-06,
+      "loss": 0.4557,
+      "step": 393
+    },
+    {
+      "epoch": 0.4429952777153137,
+      "grad_norm": 3.5278772451294476,
+      "learning_rate": 9.931067460242342e-06,
+      "loss": 0.5411,
+      "step": 394
+    },
+    {
+      "epoch": 0.44411963121205306,
+      "grad_norm": 4.4307941131699975,
+      "learning_rate": 9.929980185352525e-06,
+      "loss": 0.6207,
+      "step": 395
+    },
+    {
+      "epoch": 0.44524398470879245,
+      "grad_norm": 3.583422397228172,
+      "learning_rate": 9.928884463076045e-06,
+      "loss": 0.5635,
+      "step": 396
+    },
+    {
+      "epoch": 0.44636833820553184,
+      "grad_norm": 3.9357750151840114,
+      "learning_rate": 9.92778029529039e-06,
+      "loss": 0.5685,
+      "step": 397
+    },
+    {
+      "epoch": 0.4474926917022712,
+      "grad_norm": 3.9360833159583746,
+      "learning_rate": 9.926667683887521e-06,
+      "loss": 0.6205,
+      "step": 398
+    },
+    {
+      "epoch": 0.4486170451990106,
+      "grad_norm": 4.157022034711035,
+      "learning_rate": 9.92554663077387e-06,
+      "loss": 0.6177,
+      "step": 399
+    },
+    {
+      "epoch": 0.4497413986957499,
+      "grad_norm": 4.093485292287712,
+      "learning_rate": 9.92441713787033e-06,
+      "loss": 0.5299,
+      "step": 400
+    },
+    {
+      "epoch": 0.4508657521924893,
+      "grad_norm": 3.703591594644585,
+      "learning_rate": 9.923279207112256e-06,
+      "loss": 0.6651,
+      "step": 401
+    },
+    {
+      "epoch": 0.4519901056892287,
+      "grad_norm": 3.366467862644327,
+      "learning_rate": 9.922132840449459e-06,
+      "loss": 0.4353,
+      "step": 402
+    },
+    {
+      "epoch": 0.45311445918596804,
+      "grad_norm": 3.4773365760132036,
+      "learning_rate": 9.92097803984621e-06,
+      "loss": 0.5343,
+      "step": 403
+    },
+    {
+      "epoch": 0.45423881268270744,
+      "grad_norm": 3.1312504082670714,
+      "learning_rate": 9.919814807281228e-06,
+      "loss": 0.5574,
+      "step": 404
+    },
+    {
+      "epoch": 0.45536316617944683,
+      "grad_norm": 3.4919543036699023,
+      "learning_rate": 9.918643144747681e-06,
+      "loss": 0.5768,
+      "step": 405
+    },
+    {
+      "epoch": 0.45648751967618617,
+      "grad_norm": 3.4077696137529547,
+      "learning_rate": 9.917463054253179e-06,
+      "loss": 0.5465,
+      "step": 406
+    },
+    {
+      "epoch": 0.45761187317292557,
+      "grad_norm": 3.7331539593099476,
+      "learning_rate": 9.916274537819774e-06,
+      "loss": 0.4461,
+      "step": 407
+    },
+    {
+      "epoch": 0.45873622666966496,
+      "grad_norm": 3.9088882747127753,
+      "learning_rate": 9.915077597483959e-06,
+      "loss": 0.5705,
+      "step": 408
+    },
+    {
+      "epoch": 0.4598605801664043,
+      "grad_norm": 3.450279277169046,
+      "learning_rate": 9.913872235296657e-06,
+      "loss": 0.4681,
+      "step": 409
+    },
+    {
+      "epoch": 0.4609849336631437,
+      "grad_norm": 3.360001720476299,
+      "learning_rate": 9.912658453323225e-06,
+      "loss": 0.5125,
+      "step": 410
+    },
+    {
+      "epoch": 0.4621092871598831,
+      "grad_norm": 4.020533505812009,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 0.6391,
+      "step": 411
+    },
+    {
+      "epoch": 0.4632336406566224,
+      "grad_norm": 3.7258187955998077,
+      "learning_rate": 9.91020563835152e-06,
+      "loss": 0.6104,
+      "step": 412
+    },
+    {
+      "epoch": 0.4643579941533618,
+      "grad_norm": 3.7436440191724953,
+      "learning_rate": 9.90896660955608e-06,
+      "loss": 0.473,
+      "step": 413
+    },
+    {
+      "epoch": 0.4654823476501012,
+      "grad_norm": 3.6904207067503934,
+      "learning_rate": 9.907719169380164e-06,
+      "loss": 0.7211,
+      "step": 414
+    },
+    {
+      "epoch": 0.46660670114684055,
+      "grad_norm": 3.588363134945268,
+      "learning_rate": 9.906463319961225e-06,
+      "loss": 0.5571,
+      "step": 415
+    },
+    {
+      "epoch": 0.46773105464357995,
+      "grad_norm": 3.0258523284766494,
+      "learning_rate": 9.905199063451135e-06,
+      "loss": 0.4625,
+      "step": 416
+    },
+    {
+      "epoch": 0.46885540814031934,
+      "grad_norm": 3.5665343425128606,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 0.5143,
+      "step": 417
+    },
+    {
+      "epoch": 0.4699797616370587,
+      "grad_norm": 4.06007995893093,
+      "learning_rate": 9.902645337836955e-06,
+      "loss": 0.5767,
+      "step": 418
+    },
+    {
+      "epoch": 0.4711041151337981,
+      "grad_norm": 3.583327220888298,
+      "learning_rate": 9.901355873108611e-06,
+      "loss": 0.5423,
+      "step": 419
+    },
+    {
+      "epoch": 0.47222846863053747,
+      "grad_norm": 3.4572657088614847,
+      "learning_rate": 9.900058010040578e-06,
+      "loss": 0.5179,
+      "step": 420
+    },
+    {
+      "epoch": 0.4733528221272768,
+      "grad_norm": 4.621100332205935,
+      "learning_rate": 9.898751750856715e-06,
+      "loss": 0.592,
+      "step": 421
+    },
+    {
+      "epoch": 0.4744771756240162,
+      "grad_norm": 3.597218512109592,
+      "learning_rate": 9.897437097795257e-06,
+      "loss": 0.497,
+      "step": 422
+    },
+    {
+      "epoch": 0.47560152912075554,
+      "grad_norm": 3.4229379409554035,
+      "learning_rate": 9.89611405310883e-06,
+      "loss": 0.588,
+      "step": 423
+    },
+    {
+      "epoch": 0.47672588261749493,
+      "grad_norm": 3.175869164294685,
+      "learning_rate": 9.89478261906443e-06,
+      "loss": 0.5994,
+      "step": 424
+    },
+    {
+      "epoch": 0.47785023611423433,
+      "grad_norm": 3.5446901244571714,
+      "learning_rate": 9.89344279794344e-06,
+      "loss": 0.607,
+      "step": 425
+    },
+    {
+      "epoch": 0.47897458961097367,
+      "grad_norm": 3.1751488739859965,
+      "learning_rate": 9.892094592041602e-06,
+      "loss": 0.5533,
+      "step": 426
+    },
+    {
+      "epoch": 0.48009894310771306,
+      "grad_norm": 3.296322370927872,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.4542,
+      "step": 427
+    },
+    {
+      "epoch": 0.48122329660445246,
+      "grad_norm": 3.4112487659632142,
+      "learning_rate": 9.889373035150202e-06,
+      "loss": 0.4573,
+      "step": 428
+    },
+    {
+      "epoch": 0.4823476501011918,
+      "grad_norm": 3.3106549112329247,
+      "learning_rate": 9.887999688823955e-06,
+      "loss": 0.554,
+      "step": 429
+    },
+    {
+      "epoch": 0.4834720035979312,
+      "grad_norm": 3.109251606943226,
+      "learning_rate": 9.88661796704348e-06,
+      "loss": 0.5194,
+      "step": 430
+    },
+    {
+      "epoch": 0.4845963570946706,
+      "grad_norm": 3.0634109803954694,
+      "learning_rate": 9.88522787217632e-06,
+      "loss": 0.4414,
+      "step": 431
+    },
+    {
+      "epoch": 0.4857207105914099,
+      "grad_norm": 3.47565939839024,
+      "learning_rate": 9.883829406604363e-06,
+      "loss": 0.5754,
+      "step": 432
+    },
+    {
+      "epoch": 0.4868450640881493,
+      "grad_norm": 3.6799763076192997,
+      "learning_rate": 9.882422572723844e-06,
+      "loss": 0.637,
+      "step": 433
+    },
+    {
+      "epoch": 0.4879694175848887,
+      "grad_norm": 3.6031778742943583,
+      "learning_rate": 9.881007372945334e-06,
+      "loss": 0.5872,
+      "step": 434
+    },
+    {
+      "epoch": 0.48909377108162805,
+      "grad_norm": 3.2902084914563314,
+      "learning_rate": 9.879583809693737e-06,
+      "loss": 0.4471,
+      "step": 435
+    },
+    {
+      "epoch": 0.49021812457836744,
+      "grad_norm": 3.698046147306681,
+      "learning_rate": 9.878151885408294e-06,
+      "loss": 0.4845,
+      "step": 436
+    },
+    {
+      "epoch": 0.49134247807510684,
+      "grad_norm": 3.235767538466883,
+      "learning_rate": 9.876711602542564e-06,
+      "loss": 0.4824,
+      "step": 437
+    },
+    {
+      "epoch": 0.4924668315718462,
+      "grad_norm": 3.4387314834978198,
+      "learning_rate": 9.875262963564436e-06,
+      "loss": 0.5486,
+      "step": 438
+    },
+    {
+      "epoch": 0.49359118506858557,
+      "grad_norm": 3.7587078960626474,
+      "learning_rate": 9.87380597095611e-06,
+      "loss": 0.5167,
+      "step": 439
+    },
+    {
+      "epoch": 0.49471553856532496,
+      "grad_norm": 4.293504740627422,
+      "learning_rate": 9.872340627214106e-06,
+      "loss": 0.5843,
+      "step": 440
+    },
+    {
+      "epoch": 0.4958398920620643,
+      "grad_norm": 3.3964015812282367,
+      "learning_rate": 9.870866934849248e-06,
+      "loss": 0.5502,
+      "step": 441
+    },
+    {
+      "epoch": 0.4969642455588037,
+      "grad_norm": 3.1728704778597234,
+      "learning_rate": 9.869384896386669e-06,
+      "loss": 0.5332,
+      "step": 442
+    },
+    {
+      "epoch": 0.49808859905554304,
+      "grad_norm": 3.62682975355754,
+      "learning_rate": 9.867894514365802e-06,
+      "loss": 0.5532,
+      "step": 443
+    },
+    {
+      "epoch": 0.49921295255228243,
+      "grad_norm": 3.346240782483105,
+      "learning_rate": 9.866395791340376e-06,
+      "loss": 0.4657,
+      "step": 444
+    },
+    {
+      "epoch": 0.5003373060490218,
+      "grad_norm": 3.4552617975676023,
+      "learning_rate": 9.86488872987841e-06,
+      "loss": 0.4316,
+      "step": 445
+    },
+    {
+      "epoch": 0.5014616595457612,
+      "grad_norm": 3.4566195584536077,
+      "learning_rate": 9.863373332562215e-06,
+      "loss": 0.5601,
+      "step": 446
+    },
+    {
+      "epoch": 0.5025860130425006,
+      "grad_norm": 3.345038962395701,
+      "learning_rate": 9.861849601988384e-06,
+      "loss": 0.5691,
+      "step": 447
+    },
+    {
+      "epoch": 0.50371036653924,
+      "grad_norm": 3.5301109563736146,
+      "learning_rate": 9.860317540767784e-06,
+      "loss": 0.6567,
+      "step": 448
+    },
+    {
+      "epoch": 0.5048347200359793,
+      "grad_norm": 2.728216193517549,
+      "learning_rate": 9.858777151525563e-06,
+      "loss": 0.3743,
+      "step": 449
+    },
+    {
+      "epoch": 0.5059590735327187,
+      "grad_norm": 2.8259837183026457,
+      "learning_rate": 9.857228436901137e-06,
+      "loss": 0.4421,
+      "step": 450
+    },
+    {
+      "epoch": 0.5070834270294581,
+      "grad_norm": 3.469613007759425,
+      "learning_rate": 9.85567139954818e-06,
+      "loss": 0.4831,
+      "step": 451
+    },
+    {
+      "epoch": 0.5082077805261974,
+      "grad_norm": 3.7626537144691183,
+      "learning_rate": 9.854106042134642e-06,
+      "loss": 0.6583,
+      "step": 452
+    },
+    {
+      "epoch": 0.5093321340229369,
+      "grad_norm": 3.5621907574407214,
+      "learning_rate": 9.852532367342712e-06,
+      "loss": 0.5668,
+      "step": 453
+    },
+    {
+      "epoch": 0.5104564875196762,
+      "grad_norm": 3.0493305612412445,
+      "learning_rate": 9.850950377868846e-06,
+      "loss": 0.4545,
+      "step": 454
+    },
+    {
+      "epoch": 0.5115808410164155,
+      "grad_norm": 4.12362054504951,
+      "learning_rate": 9.849360076423736e-06,
+      "loss": 0.5296,
+      "step": 455
+    },
+    {
+      "epoch": 0.5127051945131549,
+      "grad_norm": 3.2809216133417305,
+      "learning_rate": 9.847761465732319e-06,
+      "loss": 0.5188,
+      "step": 456
+    },
+    {
+      "epoch": 0.5138295480098943,
+      "grad_norm": 2.9550748198809567,
+      "learning_rate": 9.846154548533773e-06,
+      "loss": 0.4487,
+      "step": 457
+    },
+    {
+      "epoch": 0.5149539015066337,
+      "grad_norm": 3.3634324875125934,
+      "learning_rate": 9.844539327581505e-06,
+      "loss": 0.4821,
+      "step": 458
+    },
+    {
+      "epoch": 0.516078255003373,
+      "grad_norm": 3.8964772998145834,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 0.5309,
+      "step": 459
+    },
+    {
+      "epoch": 0.5172026085001125,
+      "grad_norm": 3.5086741172462483,
+      "learning_rate": 9.841283985500583e-06,
+      "loss": 0.5712,
+      "step": 460
+    },
+    {
+      "epoch": 0.5183269619968518,
+      "grad_norm": 3.6056384752480897,
+      "learning_rate": 9.839643869949867e-06,
+      "loss": 0.5501,
+      "step": 461
+    },
+    {
+      "epoch": 0.5194513154935911,
+      "grad_norm": 3.278518746467835,
+      "learning_rate": 9.8379954618013e-06,
+      "loss": 0.5508,
+      "step": 462
+    },
+    {
+      "epoch": 0.5205756689903306,
+      "grad_norm": 3.690504779097785,
+      "learning_rate": 9.836338763879386e-06,
+      "loss": 0.6016,
+      "step": 463
+    },
+    {
+      "epoch": 0.5217000224870699,
+      "grad_norm": 3.427224236551698,
+      "learning_rate": 9.834673779022829e-06,
+      "loss": 0.5079,
+      "step": 464
+    },
+    {
+      "epoch": 0.5228243759838093,
+      "grad_norm": 3.3990357073596305,
+      "learning_rate": 9.833000510084537e-06,
+      "loss": 0.576,
+      "step": 465
+    },
+    {
+      "epoch": 0.5239487294805487,
+      "grad_norm": 3.2481979882196694,
+      "learning_rate": 9.831318959931612e-06,
+      "loss": 0.5226,
+      "step": 466
+    },
+    {
+      "epoch": 0.525073082977288,
+      "grad_norm": 3.671279308246881,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.5144,
+      "step": 467
+    },
+    {
+      "epoch": 0.5261974364740274,
+      "grad_norm": 3.2289616401186096,
+      "learning_rate": 9.827931027521204e-06,
+      "loss": 0.5617,
+      "step": 468
+    },
+    {
+      "epoch": 0.5273217899707668,
+      "grad_norm": 3.410257296244155,
+      "learning_rate": 9.826224651068853e-06,
+      "loss": 0.5903,
+      "step": 469
+    },
+    {
+      "epoch": 0.5284461434675062,
+      "grad_norm": 2.936066759242053,
+      "learning_rate": 9.824510005012114e-06,
+      "loss": 0.4716,
+      "step": 470
+    },
+    {
+      "epoch": 0.5295704969642455,
+      "grad_norm": 3.5114155551923014,
+      "learning_rate": 9.822787092288991e-06,
+      "loss": 0.4439,
+      "step": 471
+    },
+    {
+      "epoch": 0.530694850460985,
+      "grad_norm": 3.5757058845796985,
+      "learning_rate": 9.821055915851647e-06,
+      "loss": 0.5756,
+      "step": 472
+    },
+    {
+      "epoch": 0.5318192039577243,
+      "grad_norm": 3.4577317904288787,
+      "learning_rate": 9.819316478666406e-06,
+      "loss": 0.5541,
+      "step": 473
+    },
+    {
+      "epoch": 0.5329435574544636,
+      "grad_norm": 3.1635890061525345,
+      "learning_rate": 9.817568783713744e-06,
+      "loss": 0.4845,
+      "step": 474
+    },
+    {
+      "epoch": 0.5340679109512031,
+      "grad_norm": 3.3979694050871823,
+      "learning_rate": 9.815812833988292e-06,
+      "loss": 0.514,
+      "step": 475
+    },
+    {
+      "epoch": 0.5351922644479424,
+      "grad_norm": 4.1171550358774915,
+      "learning_rate": 9.814048632498818e-06,
+      "loss": 0.4921,
+      "step": 476
+    },
+    {
+      "epoch": 0.5363166179446818,
+      "grad_norm": 3.7273012215692596,
+      "learning_rate": 9.812276182268236e-06,
+      "loss": 0.5768,
+      "step": 477
+    },
+    {
+      "epoch": 0.5374409714414212,
+      "grad_norm": 3.2146685547860576,
+      "learning_rate": 9.810495486333592e-06,
+      "loss": 0.5706,
+      "step": 478
+    },
+    {
+      "epoch": 0.5385653249381606,
+      "grad_norm": 3.287421442791594,
+      "learning_rate": 9.808706547746057e-06,
+      "loss": 0.4678,
+      "step": 479
+    },
+    {
+      "epoch": 0.5396896784348999,
+      "grad_norm": 3.375796329669829,
+      "learning_rate": 9.806909369570931e-06,
+      "loss": 0.4763,
+      "step": 480
+    },
+    {
+      "epoch": 0.5408140319316393,
+      "grad_norm": 3.611966077762474,
+      "learning_rate": 9.805103954887626e-06,
+      "loss": 0.6274,
+      "step": 481
+    },
+    {
+      "epoch": 0.5419383854283787,
+      "grad_norm": 3.871583039160485,
+      "learning_rate": 9.803290306789676e-06,
+      "loss": 0.5593,
+      "step": 482
+    },
+    {
+      "epoch": 0.543062738925118,
+      "grad_norm": 3.5077539892944296,
+      "learning_rate": 9.801468428384716e-06,
+      "loss": 0.5176,
+      "step": 483
+    },
+    {
+      "epoch": 0.5441870924218575,
+      "grad_norm": 3.6183018557858544,
+      "learning_rate": 9.799638322794482e-06,
+      "loss": 0.654,
+      "step": 484
+    },
+    {
+      "epoch": 0.5453114459185968,
+      "grad_norm": 3.6968286018116903,
+      "learning_rate": 9.797799993154815e-06,
+      "loss": 0.5421,
+      "step": 485
+    },
+    {
+      "epoch": 0.5464357994153362,
+      "grad_norm": 3.308286983274441,
+      "learning_rate": 9.795953442615637e-06,
+      "loss": 0.565,
+      "step": 486
+    },
+    {
+      "epoch": 0.5475601529120756,
+      "grad_norm": 3.2919976187407243,
+      "learning_rate": 9.794098674340966e-06,
+      "loss": 0.4381,
+      "step": 487
+    },
+    {
+      "epoch": 0.5486845064088149,
+      "grad_norm": 102.86710250060017,
+      "learning_rate": 9.792235691508896e-06,
+      "loss": 0.782,
+      "step": 488
+    },
+    {
+      "epoch": 0.5498088599055543,
+      "grad_norm": 5.137624713746361,
+      "learning_rate": 9.790364497311597e-06,
+      "loss": 0.4872,
+      "step": 489
+    },
+    {
+      "epoch": 0.5509332134022937,
+      "grad_norm": 3.445119461900327,
+      "learning_rate": 9.788485094955309e-06,
+      "loss": 0.5642,
+      "step": 490
+    },
+    {
+      "epoch": 0.5520575668990331,
+      "grad_norm": 3.2666369358584038,
+      "learning_rate": 9.786597487660336e-06,
+      "loss": 0.4947,
+      "step": 491
+    },
+    {
+      "epoch": 0.5531819203957724,
+      "grad_norm": 3.6498309674852556,
+      "learning_rate": 9.784701678661045e-06,
+      "loss": 0.5565,
+      "step": 492
+    },
+    {
+      "epoch": 0.5543062738925119,
+      "grad_norm": 3.499277726079078,
+      "learning_rate": 9.78279767120585e-06,
+      "loss": 0.5057,
+      "step": 493
+    },
+    {
+      "epoch": 0.5554306273892512,
+      "grad_norm": 3.2892653894312276,
+      "learning_rate": 9.780885468557216e-06,
+      "loss": 0.5095,
+      "step": 494
+    },
+    {
+      "epoch": 0.5565549808859905,
+      "grad_norm": 3.2916772857989813,
+      "learning_rate": 9.778965073991652e-06,
+      "loss": 0.4533,
+      "step": 495
+    },
+    {
+      "epoch": 0.55767933438273,
+      "grad_norm": 2.9015468968368183,
+      "learning_rate": 9.777036490799699e-06,
+      "loss": 0.421,
+      "step": 496
+    },
+    {
+      "epoch": 0.5588036878794693,
+      "grad_norm": 3.1604035084383955,
+      "learning_rate": 9.775099722285934e-06,
+      "loss": 0.5664,
+      "step": 497
+    },
+    {
+      "epoch": 0.5599280413762087,
+      "grad_norm": 3.2884749405743845,
+      "learning_rate": 9.773154771768956e-06,
+      "loss": 0.5764,
+      "step": 498
+    },
+    {
+      "epoch": 0.561052394872948,
+      "grad_norm": 3.3754283664248126,
+      "learning_rate": 9.771201642581384e-06,
+      "loss": 0.4918,
+      "step": 499
+    },
+    {
+      "epoch": 0.5621767483696875,
+      "grad_norm": 3.509015505851844,
+      "learning_rate": 9.769240338069854e-06,
+      "loss": 0.5579,
+      "step": 500
+    },
+    {
+      "epoch": 0.5621767483696875,
+      "eval_loss": 0.2699817717075348,
+      "eval_runtime": 359.2723,
+      "eval_samples_per_second": 148.52,
+      "eval_steps_per_second": 4.643,
+      "step": 500
+    },
+    {
+      "epoch": 0.5633011018664268,
+      "grad_norm": 3.9765812033758206,
+      "learning_rate": 9.767270861595006e-06,
+      "loss": 0.6044,
+      "step": 501
+    },
+    {
+      "epoch": 0.5644254553631661,
+      "grad_norm": 3.6144367551383825,
+      "learning_rate": 9.765293216531486e-06,
+      "loss": 0.4939,
+      "step": 502
+    },
+    {
+      "epoch": 0.5655498088599056,
+      "grad_norm": 3.4156094608506877,
+      "learning_rate": 9.763307406267933e-06,
+      "loss": 0.5488,
+      "step": 503
+    },
+    {
+      "epoch": 0.5666741623566449,
+      "grad_norm": 3.633052444737866,
+      "learning_rate": 9.761313434206978e-06,
+      "loss": 0.6159,
+      "step": 504
+    },
+    {
+      "epoch": 0.5677985158533843,
+      "grad_norm": 3.1997356848744296,
+      "learning_rate": 9.759311303765239e-06,
+      "loss": 0.5585,
+      "step": 505
+    },
+    {
+      "epoch": 0.5689228693501237,
+      "grad_norm": 3.592037681887678,
+      "learning_rate": 9.757301018373314e-06,
+      "loss": 0.4884,
+      "step": 506
+    },
+    {
+      "epoch": 0.570047222846863,
+      "grad_norm": 2.9593690194809175,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.3916,
+      "step": 507
+    },
+    {
+      "epoch": 0.5711715763436024,
+      "grad_norm": 3.719284182003654,
+      "learning_rate": 9.75325599653114e-06,
+      "loss": 0.4805,
+      "step": 508
+    },
+    {
+      "epoch": 0.5722959298403418,
+      "grad_norm": 3.144652980277522,
+      "learning_rate": 9.75122126701193e-06,
+      "loss": 0.594,
+      "step": 509
+    },
+    {
+      "epoch": 0.5734202833370812,
+      "grad_norm": 2.9090227290313138,
+      "learning_rate": 9.749178396404588e-06,
+      "loss": 0.4922,
+      "step": 510
+    },
+    {
+      "epoch": 0.5745446368338205,
+      "grad_norm": 3.813205423966551,
+      "learning_rate": 9.74712738820952e-06,
+      "loss": 0.597,
+      "step": 511
+    },
+    {
+      "epoch": 0.57566899033056,
+      "grad_norm": 3.6778066945376175,
+      "learning_rate": 9.745068245941071e-06,
+      "loss": 0.5671,
+      "step": 512
+    },
+    {
+      "epoch": 0.5767933438272993,
+      "grad_norm": 3.1635295630956053,
+      "learning_rate": 9.743000973127523e-06,
+      "loss": 0.5359,
+      "step": 513
+    },
+    {
+      "epoch": 0.5779176973240386,
+      "grad_norm": 2.917650642385911,
+      "learning_rate": 9.740925573311098e-06,
+      "loss": 0.5154,
+      "step": 514
+    },
+    {
+      "epoch": 0.5790420508207781,
+      "grad_norm": 3.244020010237936,
+      "learning_rate": 9.73884205004793e-06,
+      "loss": 0.5289,
+      "step": 515
+    },
+    {
+      "epoch": 0.5801664043175174,
+      "grad_norm": 3.3971513719582163,
+      "learning_rate": 9.736750406908082e-06,
+      "loss": 0.5081,
+      "step": 516
+    },
+    {
+      "epoch": 0.5812907578142568,
+      "grad_norm": 3.28768431449784,
+      "learning_rate": 9.73465064747553e-06,
+      "loss": 0.447,
+      "step": 517
+    },
+    {
+      "epoch": 0.5824151113109962,
+      "grad_norm": 3.824414557877193,
+      "learning_rate": 9.732542775348151e-06,
+      "loss": 0.4709,
+      "step": 518
+    },
+    {
+      "epoch": 0.5835394648077356,
+      "grad_norm": 3.721520798762539,
+      "learning_rate": 9.730426794137727e-06,
+      "loss": 0.5086,
+      "step": 519
+    },
+    {
+      "epoch": 0.5846638183044749,
+      "grad_norm": 4.134250946781774,
+      "learning_rate": 9.728302707469936e-06,
+      "loss": 0.6072,
+      "step": 520
+    },
+    {
+      "epoch": 0.5857881718012143,
+      "grad_norm": 3.3082336006121515,
+      "learning_rate": 9.726170518984342e-06,
+      "loss": 0.5482,
+      "step": 521
+    },
+    {
+      "epoch": 0.5869125252979537,
+      "grad_norm": 4.3836531835061265,
+      "learning_rate": 9.72403023233439e-06,
+      "loss": 0.5475,
+      "step": 522
+    },
+    {
+      "epoch": 0.588036878794693,
+      "grad_norm": 3.5759644165261575,
+      "learning_rate": 9.721881851187406e-06,
+      "loss": 0.5824,
+      "step": 523
+    },
+    {
+      "epoch": 0.5891612322914325,
+      "grad_norm": 3.5136178063171712,
+      "learning_rate": 9.71972537922458e-06,
+      "loss": 0.5377,
+      "step": 524
+    },
+    {
+      "epoch": 0.5902855857881718,
+      "grad_norm": 3.256576708271542,
+      "learning_rate": 9.717560820140968e-06,
+      "loss": 0.5182,
+      "step": 525
+    },
+    {
+      "epoch": 0.5914099392849111,
+      "grad_norm": 3.3301868777398993,
+      "learning_rate": 9.715388177645485e-06,
+      "loss": 0.6156,
+      "step": 526
+    },
+    {
+      "epoch": 0.5925342927816506,
+      "grad_norm": 3.7318275561982603,
+      "learning_rate": 9.713207455460893e-06,
+      "loss": 0.5999,
+      "step": 527
+    },
+    {
+      "epoch": 0.5936586462783899,
+      "grad_norm": 3.643262745185435,
+      "learning_rate": 9.7110186573238e-06,
+      "loss": 0.688,
+      "step": 528
+    },
+    {
+      "epoch": 0.5947829997751293,
+      "grad_norm": 3.088379207674747,
+      "learning_rate": 9.708821786984652e-06,
+      "loss": 0.5086,
+      "step": 529
+    },
+    {
+      "epoch": 0.5959073532718687,
+      "grad_norm": 3.3205979356301123,
+      "learning_rate": 9.706616848207728e-06,
+      "loss": 0.6106,
+      "step": 530
+    },
+    {
+      "epoch": 0.5970317067686081,
+      "grad_norm": 2.9509510850288105,
+      "learning_rate": 9.704403844771128e-06,
+      "loss": 0.4976,
+      "step": 531
+    },
+    {
+      "epoch": 0.5981560602653474,
+      "grad_norm": 3.6151140725976565,
+      "learning_rate": 9.702182780466775e-06,
+      "loss": 0.5831,
+      "step": 532
+    },
+    {
+      "epoch": 0.5992804137620868,
+      "grad_norm": 3.266470305388926,
+      "learning_rate": 9.6999536591004e-06,
+      "loss": 0.5175,
+      "step": 533
+    },
+    {
+      "epoch": 0.6004047672588262,
+      "grad_norm": 3.1909128923950094,
+      "learning_rate": 9.697716484491545e-06,
+      "loss": 0.6097,
+      "step": 534
+    },
+    {
+      "epoch": 0.6015291207555655,
+      "grad_norm": 3.2779444944482177,
+      "learning_rate": 9.695471260473546e-06,
+      "loss": 0.6115,
+      "step": 535
+    },
+    {
+      "epoch": 0.602653474252305,
+      "grad_norm": 3.2362163927256065,
+      "learning_rate": 9.693217990893533e-06,
+      "loss": 0.464,
+      "step": 536
+    },
+    {
+      "epoch": 0.6037778277490443,
+      "grad_norm": 3.2059438371847264,
+      "learning_rate": 9.690956679612422e-06,
+      "loss": 0.5929,
+      "step": 537
+    },
+    {
+      "epoch": 0.6049021812457837,
+      "grad_norm": 3.1603232968362724,
+      "learning_rate": 9.688687330504908e-06,
+      "loss": 0.5167,
+      "step": 538
+    },
+    {
+      "epoch": 0.6060265347425231,
+      "grad_norm": 3.6793758000734598,
+      "learning_rate": 9.68640994745946e-06,
+      "loss": 0.5021,
+      "step": 539
+    },
+    {
+      "epoch": 0.6071508882392624,
+      "grad_norm": 3.06833748748244,
+      "learning_rate": 9.684124534378307e-06,
+      "loss": 0.581,
+      "step": 540
+    },
+    {
+      "epoch": 0.6082752417360018,
+      "grad_norm": 3.020732855443918,
+      "learning_rate": 9.68183109517745e-06,
+      "loss": 0.4826,
+      "step": 541
+    },
+    {
+      "epoch": 0.6093995952327411,
+      "grad_norm": 3.2566336035543855,
+      "learning_rate": 9.67952963378663e-06,
+      "loss": 0.5629,
+      "step": 542
+    },
+    {
+      "epoch": 0.6105239487294806,
+      "grad_norm": 2.905337884974537,
+      "learning_rate": 9.677220154149338e-06,
+      "loss": 0.4626,
+      "step": 543
+    },
+    {
+      "epoch": 0.6116483022262199,
+      "grad_norm": 3.777463219372923,
+      "learning_rate": 9.674902660222805e-06,
+      "loss": 0.4498,
+      "step": 544
+    },
+    {
+      "epoch": 0.6127726557229592,
+      "grad_norm": 2.7736194419409825,
+      "learning_rate": 9.672577155977993e-06,
+      "loss": 0.4393,
+      "step": 545
+    },
+    {
+      "epoch": 0.6138970092196987,
+      "grad_norm": 3.6300278751824187,
+      "learning_rate": 9.670243645399594e-06,
+      "loss": 0.5376,
+      "step": 546
+    },
+    {
+      "epoch": 0.615021362716438,
+      "grad_norm": 3.3731806907004924,
+      "learning_rate": 9.667902132486009e-06,
+      "loss": 0.5063,
+      "step": 547
+    },
+    {
+      "epoch": 0.6161457162131774,
+      "grad_norm": 3.2572388646275874,
+      "learning_rate": 9.66555262124936e-06,
+      "loss": 0.4621,
+      "step": 548
+    },
+    {
+      "epoch": 0.6172700697099168,
+      "grad_norm": 3.1752402869913974,
+      "learning_rate": 9.663195115715472e-06,
+      "loss": 0.4517,
+      "step": 549
+    },
+    {
+      "epoch": 0.6183944232066562,
+      "grad_norm": 3.705586950485253,
+      "learning_rate": 9.660829619923861e-06,
+      "loss": 0.5187,
+      "step": 550
+    },
+    {
+      "epoch": 0.6195187767033955,
+      "grad_norm": 3.156399873140764,
+      "learning_rate": 9.658456137927745e-06,
+      "loss": 0.471,
+      "step": 551
+    },
+    {
+      "epoch": 0.620643130200135,
+      "grad_norm": 3.094372391815433,
+      "learning_rate": 9.656074673794018e-06,
+      "loss": 0.4982,
+      "step": 552
+    },
+    {
+      "epoch": 0.6217674836968743,
+      "grad_norm": 3.70533757232998,
+      "learning_rate": 9.653685231603256e-06,
+      "loss": 0.5794,
+      "step": 553
+    },
+    {
+      "epoch": 0.6228918371936136,
+      "grad_norm": 3.57415231935082,
+      "learning_rate": 9.651287815449699e-06,
+      "loss": 0.5145,
+      "step": 554
+    },
+    {
+      "epoch": 0.6240161906903531,
+      "grad_norm": 3.62282787948232,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 0.5097,
+      "step": 555
+    },
+    {
+      "epoch": 0.6251405441870924,
+      "grad_norm": 3.427085007986106,
+      "learning_rate": 9.646469077699494e-06,
+      "loss": 0.5837,
+      "step": 556
+    },
+    {
+      "epoch": 0.6262648976838318,
+      "grad_norm": 3.0657890540520065,
+      "learning_rate": 9.644047764359623e-06,
+      "loss": 0.5309,
+      "step": 557
+    },
+    {
+      "epoch": 0.6273892511805712,
+      "grad_norm": 3.2250582766532294,
+      "learning_rate": 9.641618493570495e-06,
+      "loss": 0.5758,
+      "step": 558
+    },
+    {
+      "epoch": 0.6285136046773105,
+      "grad_norm": 2.725854318170067,
+      "learning_rate": 9.6391812694946e-06,
+      "loss": 0.4122,
+      "step": 559
+    },
+    {
+      "epoch": 0.6296379581740499,
+      "grad_norm": 3.70965609066183,
+      "learning_rate": 9.63673609630806e-06,
+      "loss": 0.5439,
+      "step": 560
+    },
+    {
+      "epoch": 0.6307623116707893,
+      "grad_norm": 3.3958009552579234,
+      "learning_rate": 9.634282978200605e-06,
+      "loss": 0.5865,
+      "step": 561
+    },
+    {
+      "epoch": 0.6318866651675287,
+      "grad_norm": 3.5325487998121905,
+      "learning_rate": 9.63182191937559e-06,
+      "loss": 0.5509,
+      "step": 562
+    },
+    {
+      "epoch": 0.633011018664268,
+      "grad_norm": 3.1621393334332684,
+      "learning_rate": 9.629352924049975e-06,
+      "loss": 0.5148,
+      "step": 563
+    },
+    {
+      "epoch": 0.6341353721610075,
+      "grad_norm": 3.1999402556438516,
+      "learning_rate": 9.626875996454312e-06,
+      "loss": 0.5308,
+      "step": 564
+    },
+    {
+      "epoch": 0.6352597256577468,
+      "grad_norm": 3.0272181982071826,
+      "learning_rate": 9.624391140832749e-06,
+      "loss": 0.4458,
+      "step": 565
+    },
+    {
+      "epoch": 0.6363840791544861,
+      "grad_norm": 2.9966522279659737,
+      "learning_rate": 9.62189836144302e-06,
+      "loss": 0.4668,
+      "step": 566
+    },
+    {
+      "epoch": 0.6375084326512256,
+      "grad_norm": 3.154865707914656,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.5359,
+      "step": 567
+    },
+    {
+      "epoch": 0.6386327861479649,
+      "grad_norm": 3.124177412232392,
+      "learning_rate": 9.616889048457872e-06,
+      "loss": 0.6103,
+      "step": 568
+    },
+    {
+      "epoch": 0.6397571396447043,
+      "grad_norm": 3.1363154865793543,
+      "learning_rate": 9.614372523445772e-06,
+      "loss": 0.5128,
+      "step": 569
+    },
+    {
+      "epoch": 0.6408814931414437,
+      "grad_norm": 2.9701858903763867,
+      "learning_rate": 9.611848091832134e-06,
+      "loss": 0.5117,
+      "step": 570
+    },
+    {
+      "epoch": 0.642005846638183,
+      "grad_norm": 3.2506643042659777,
+      "learning_rate": 9.609315757942504e-06,
+      "loss": 0.5616,
+      "step": 571
+    },
+    {
+      "epoch": 0.6431302001349224,
+      "grad_norm": 3.238999783275299,
+      "learning_rate": 9.606775526115963e-06,
+      "loss": 0.5583,
+      "step": 572
+    },
+    {
+      "epoch": 0.6442545536316618,
+      "grad_norm": 3.055524911345345,
+      "learning_rate": 9.604227400705134e-06,
+      "loss": 0.492,
+      "step": 573
+    },
+    {
+      "epoch": 0.6453789071284012,
+      "grad_norm": 3.912051968601764,
+      "learning_rate": 9.601671386076154e-06,
+      "loss": 0.4466,
+      "step": 574
+    },
+    {
+      "epoch": 0.6465032606251405,
+      "grad_norm": 3.0728120334464624,
+      "learning_rate": 9.59910748660869e-06,
+      "loss": 0.5387,
+      "step": 575
+    },
+    {
+      "epoch": 0.64762761412188,
+      "grad_norm": 2.7988916304268305,
+      "learning_rate": 9.596535706695911e-06,
+      "loss": 0.4289,
+      "step": 576
+    },
+    {
+      "epoch": 0.6487519676186193,
+      "grad_norm": 3.078486213393598,
+      "learning_rate": 9.593956050744493e-06,
+      "loss": 0.5263,
+      "step": 577
+    },
+    {
+      "epoch": 0.6498763211153586,
+      "grad_norm": 3.061799427069967,
+      "learning_rate": 9.591368523174604e-06,
+      "loss": 0.4529,
+      "step": 578
+    },
+    {
+      "epoch": 0.6510006746120981,
+      "grad_norm": 3.500536928365465,
+      "learning_rate": 9.588773128419907e-06,
+      "loss": 0.4923,
+      "step": 579
+    },
+    {
+      "epoch": 0.6521250281088374,
+      "grad_norm": 3.29573374605907,
+      "learning_rate": 9.586169870927535e-06,
+      "loss": 0.5548,
+      "step": 580
+    },
+    {
+      "epoch": 0.6532493816055768,
+      "grad_norm": 3.230538955259426,
+      "learning_rate": 9.583558755158102e-06,
+      "loss": 0.5072,
+      "step": 581
+    },
+    {
+      "epoch": 0.6543737351023162,
+      "grad_norm": 3.495966822630218,
+      "learning_rate": 9.58093978558568e-06,
+      "loss": 0.5884,
+      "step": 582
+    },
+    {
+      "epoch": 0.6554980885990556,
+      "grad_norm": 3.569045306855439,
+      "learning_rate": 9.578312966697807e-06,
+      "loss": 0.5638,
+      "step": 583
+    },
+    {
+      "epoch": 0.6566224420957949,
+      "grad_norm": 3.089452408736275,
+      "learning_rate": 9.575678302995461e-06,
+      "loss": 0.5001,
+      "step": 584
+    },
+    {
+      "epoch": 0.6577467955925342,
+      "grad_norm": 2.9903715803377486,
+      "learning_rate": 9.57303579899307e-06,
+      "loss": 0.5845,
+      "step": 585
+    },
+    {
+      "epoch": 0.6588711490892737,
+      "grad_norm": 3.0724033993033193,
+      "learning_rate": 9.570385459218489e-06,
+      "loss": 0.5963,
+      "step": 586
+    },
+    {
+      "epoch": 0.659995502586013,
+      "grad_norm": 3.0091737475742444,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.4564,
+      "step": 587
+    },
+    {
+      "epoch": 0.6611198560827524,
+      "grad_norm": 3.4441175382695564,
+      "learning_rate": 9.565061290531323e-06,
+      "loss": 0.5326,
+      "step": 588
+    },
+    {
+      "epoch": 0.6622442095794918,
+      "grad_norm": 3.1509506060671764,
+      "learning_rate": 9.562387470741555e-06,
+      "loss": 0.5866,
+      "step": 589
+    },
+    {
+      "epoch": 0.6633685630762312,
+      "grad_norm": 3.384713260109587,
+      "learning_rate": 9.559705833425219e-06,
+      "loss": 0.508,
+      "step": 590
+    },
+    {
+      "epoch": 0.6644929165729705,
+      "grad_norm": 3.0931853284685222,
+      "learning_rate": 9.557016383177226e-06,
+      "loss": 0.4706,
+      "step": 591
+    },
+    {
+      "epoch": 0.6656172700697099,
+      "grad_norm": 3.3403353498773765,
+      "learning_rate": 9.55431912460588e-06,
+      "loss": 0.556,
+      "step": 592
+    },
+    {
+      "epoch": 0.6667416235664493,
+      "grad_norm": 3.2579958749215776,
+      "learning_rate": 9.551614062332857e-06,
+      "loss": 0.4669,
+      "step": 593
+    },
+    {
+      "epoch": 0.6678659770631886,
+      "grad_norm": 3.2768489882512624,
+      "learning_rate": 9.548901200993206e-06,
+      "loss": 0.5309,
+      "step": 594
+    },
+    {
+      "epoch": 0.6689903305599281,
+      "grad_norm": 3.3361131717740924,
+      "learning_rate": 9.546180545235344e-06,
+      "loss": 0.4672,
+      "step": 595
+    },
+    {
+      "epoch": 0.6701146840566674,
+      "grad_norm": 5.521237961009089,
+      "learning_rate": 9.543452099721038e-06,
+      "loss": 0.5242,
+      "step": 596
+    },
+    {
+      "epoch": 0.6712390375534067,
+      "grad_norm": 3.2692997555429826,
+      "learning_rate": 9.540715869125407e-06,
+      "loss": 0.4688,
+      "step": 597
+    },
+    {
+      "epoch": 0.6723633910501462,
+      "grad_norm": 3.174132921671061,
+      "learning_rate": 9.537971858136907e-06,
+      "loss": 0.4974,
+      "step": 598
+    },
+    {
+      "epoch": 0.6734877445468855,
+      "grad_norm": 3.981156653832147,
+      "learning_rate": 9.535220071457325e-06,
+      "loss": 0.5407,
+      "step": 599
+    },
+    {
+      "epoch": 0.6746120980436249,
+      "grad_norm": 3.340009344251507,
+      "learning_rate": 9.532460513801774e-06,
+      "loss": 0.5243,
+      "step": 600
+    },
+    {
+      "epoch": 0.6757364515403643,
+      "grad_norm": 3.323476475129269,
+      "learning_rate": 9.52969318989868e-06,
+      "loss": 0.4266,
+      "step": 601
+    },
+    {
+      "epoch": 0.6768608050371037,
+      "grad_norm": 2.909601856092722,
+      "learning_rate": 9.526918104489777e-06,
+      "loss": 0.4618,
+      "step": 602
+    },
+    {
+      "epoch": 0.677985158533843,
+      "grad_norm": 3.1567078338364043,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 0.6196,
+      "step": 603
+    },
+    {
+      "epoch": 0.6791095120305825,
+      "grad_norm": 2.972676469384324,
+      "learning_rate": 9.521344668187968e-06,
+      "loss": 0.5218,
+      "step": 604
+    },
+    {
+      "epoch": 0.6802338655273218,
+      "grad_norm": 2.8707461180665788,
+      "learning_rate": 9.518546326844993e-06,
+      "loss": 0.4885,
+      "step": 605
+    },
+    {
+      "epoch": 0.6813582190240611,
+      "grad_norm": 3.3145598800663314,
+      "learning_rate": 9.515740243096056e-06,
+      "loss": 0.5094,
+      "step": 606
+    },
+    {
+      "epoch": 0.6824825725208006,
+      "grad_norm": 2.880430942430119,
+      "learning_rate": 9.512926421749305e-06,
+      "loss": 0.3924,
+      "step": 607
+    },
+    {
+      "epoch": 0.6836069260175399,
+      "grad_norm": 2.8686067279472702,
+      "learning_rate": 9.510104867626143e-06,
+      "loss": 0.4545,
+      "step": 608
+    },
+    {
+      "epoch": 0.6847312795142793,
+      "grad_norm": 3.6508624924191406,
+      "learning_rate": 9.507275585561228e-06,
+      "loss": 0.5686,
+      "step": 609
+    },
+    {
+      "epoch": 0.6858556330110187,
+      "grad_norm": 2.7905367591015215,
+      "learning_rate": 9.504438580402462e-06,
+      "loss": 0.4774,
+      "step": 610
+    },
+    {
+      "epoch": 0.686979986507758,
+      "grad_norm": 3.4876827220660807,
+      "learning_rate": 9.501593857010968e-06,
+      "loss": 0.5981,
+      "step": 611
+    },
+    {
+      "epoch": 0.6881043400044974,
+      "grad_norm": 3.1366049044226845,
+      "learning_rate": 9.498741420261109e-06,
+      "loss": 0.589,
+      "step": 612
+    },
+    {
+      "epoch": 0.6892286935012368,
+      "grad_norm": 3.874349598581286,
+      "learning_rate": 9.495881275040453e-06,
+      "loss": 0.541,
+      "step": 613
+    },
+    {
+      "epoch": 0.6903530469979762,
+      "grad_norm": 2.9106208406665597,
+      "learning_rate": 9.493013426249782e-06,
+      "loss": 0.4703,
+      "step": 614
+    },
+    {
+      "epoch": 0.6914774004947155,
+      "grad_norm": 3.279651651332076,
+      "learning_rate": 9.490137878803078e-06,
+      "loss": 0.4985,
+      "step": 615
+    },
+    {
+      "epoch": 0.692601753991455,
+      "grad_norm": 3.464764142500255,
+      "learning_rate": 9.487254637627514e-06,
+      "loss": 0.5369,
+      "step": 616
+    },
+    {
+      "epoch": 0.6937261074881943,
+      "grad_norm": 4.9597818319764055,
+      "learning_rate": 9.484363707663443e-06,
+      "loss": 0.5185,
+      "step": 617
+    },
+    {
+      "epoch": 0.6948504609849336,
+      "grad_norm": 3.3194447583988405,
+      "learning_rate": 9.481465093864395e-06,
+      "loss": 0.5895,
+      "step": 618
+    },
+    {
+      "epoch": 0.6959748144816731,
+      "grad_norm": 3.294819513128514,
+      "learning_rate": 9.478558801197065e-06,
+      "loss": 0.4657,
+      "step": 619
+    },
+    {
+      "epoch": 0.6970991679784124,
+      "grad_norm": 3.206252150504104,
+      "learning_rate": 9.475644834641309e-06,
+      "loss": 0.409,
+      "step": 620
+    },
+    {
+      "epoch": 0.6982235214751518,
+      "grad_norm": 3.4376778561720815,
+      "learning_rate": 9.472723199190126e-06,
+      "loss": 0.5535,
+      "step": 621
+    },
+    {
+      "epoch": 0.6993478749718912,
+      "grad_norm": 2.971481371866168,
+      "learning_rate": 9.469793899849663e-06,
+      "loss": 0.4658,
+      "step": 622
+    },
+    {
+      "epoch": 0.7004722284686306,
+      "grad_norm": 2.99441955684681,
+      "learning_rate": 9.46685694163919e-06,
+      "loss": 0.5011,
+      "step": 623
+    },
+    {
+      "epoch": 0.7015965819653699,
+      "grad_norm": 3.149987972101776,
+      "learning_rate": 9.463912329591105e-06,
+      "loss": 0.4652,
+      "step": 624
+    },
+    {
+      "epoch": 0.7027209354621093,
+      "grad_norm": 3.9136119895223636,
+      "learning_rate": 9.460960068750924e-06,
+      "loss": 0.6319,
+      "step": 625
+    },
+    {
+      "epoch": 0.7038452889588487,
+      "grad_norm": 3.357906854396289,
+      "learning_rate": 9.458000164177264e-06,
+      "loss": 0.6259,
+      "step": 626
+    },
+    {
+      "epoch": 0.704969642455588,
+      "grad_norm": 3.497239367898239,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.6181,
+      "step": 627
+    },
+    {
+      "epoch": 0.7060939959523274,
+      "grad_norm": 2.870362791044923,
+      "learning_rate": 9.452057444129457e-06,
+      "loss": 0.4556,
+      "step": 628
+    },
+    {
+      "epoch": 0.7072183494490668,
+      "grad_norm": 3.583600413658476,
+      "learning_rate": 9.449074638838e-06,
+      "loss": 0.5446,
+      "step": 629
+    },
+    {
+      "epoch": 0.7083427029458061,
+      "grad_norm": 3.2498911690541994,
+      "learning_rate": 9.446084210178423e-06,
+      "loss": 0.5125,
+      "step": 630
+    },
+    {
+      "epoch": 0.7094670564425455,
+      "grad_norm": 3.2135677191067415,
+      "learning_rate": 9.443086163274745e-06,
+      "loss": 0.4748,
+      "step": 631
+    },
+    {
+      "epoch": 0.7105914099392849,
+      "grad_norm": 3.2756026813494574,
+      "learning_rate": 9.440080503264038e-06,
+      "loss": 0.5361,
+      "step": 632
+    },
+    {
+      "epoch": 0.7117157634360243,
+      "grad_norm": 3.258049851286569,
+      "learning_rate": 9.437067235296418e-06,
+      "loss": 0.5052,
+      "step": 633
+    },
+    {
+      "epoch": 0.7128401169327636,
+      "grad_norm": 2.808387487795211,
+      "learning_rate": 9.434046364535036e-06,
+      "loss": 0.4323,
+      "step": 634
+    },
+    {
+      "epoch": 0.7139644704295031,
+      "grad_norm": 3.1179208777287535,
+      "learning_rate": 9.431017896156074e-06,
+      "loss": 0.4898,
+      "step": 635
+    },
+    {
+      "epoch": 0.7150888239262424,
+      "grad_norm": 3.017798454926226,
+      "learning_rate": 9.427981835348729e-06,
+      "loss": 0.5007,
+      "step": 636
+    },
+    {
+      "epoch": 0.7162131774229817,
+      "grad_norm": 3.1347831015055565,
+      "learning_rate": 9.42493818731521e-06,
+      "loss": 0.5233,
+      "step": 637
+    },
+    {
+      "epoch": 0.7173375309197212,
+      "grad_norm": 3.225922085263521,
+      "learning_rate": 9.421886957270723e-06,
+      "loss": 0.4949,
+      "step": 638
+    },
+    {
+      "epoch": 0.7184618844164605,
+      "grad_norm": 3.1040230115351215,
+      "learning_rate": 9.418828150443469e-06,
+      "loss": 0.4533,
+      "step": 639
+    },
+    {
+      "epoch": 0.7195862379131999,
+      "grad_norm": 3.280153962945605,
+      "learning_rate": 9.415761772074628e-06,
+      "loss": 0.4613,
+      "step": 640
+    },
+    {
+      "epoch": 0.7207105914099393,
+      "grad_norm": 3.159260679703447,
+      "learning_rate": 9.412687827418356e-06,
+      "loss": 0.5899,
+      "step": 641
+    },
+    {
+      "epoch": 0.7218349449066787,
+      "grad_norm": 3.043236122880194,
+      "learning_rate": 9.409606321741776e-06,
+      "loss": 0.5373,
+      "step": 642
+    },
+    {
+      "epoch": 0.722959298403418,
+      "grad_norm": 2.9551106216833722,
+      "learning_rate": 9.406517260324962e-06,
+      "loss": 0.4874,
+      "step": 643
+    },
+    {
+      "epoch": 0.7240836519001574,
+      "grad_norm": 2.9422653201665603,
+      "learning_rate": 9.403420648460937e-06,
+      "loss": 0.4644,
+      "step": 644
+    },
+    {
+      "epoch": 0.7252080053968968,
+      "grad_norm": 3.0425116035131463,
+      "learning_rate": 9.40031649145566e-06,
+      "loss": 0.4297,
+      "step": 645
+    },
+    {
+      "epoch": 0.7263323588936361,
+      "grad_norm": 3.2063984193153052,
+      "learning_rate": 9.397204794628021e-06,
+      "loss": 0.5293,
+      "step": 646
+    },
+    {
+      "epoch": 0.7274567123903756,
+      "grad_norm": 3.238755163149549,
+      "learning_rate": 9.394085563309827e-06,
+      "loss": 0.5724,
+      "step": 647
+    },
+    {
+      "epoch": 0.7285810658871149,
+      "grad_norm": 3.310424140875838,
+      "learning_rate": 9.390958802845797e-06,
+      "loss": 0.5557,
+      "step": 648
+    },
+    {
+      "epoch": 0.7297054193838542,
+      "grad_norm": 2.895307821179136,
+      "learning_rate": 9.387824518593546e-06,
+      "loss": 0.498,
+      "step": 649
+    },
+    {
+      "epoch": 0.7308297728805937,
+      "grad_norm": 3.39729474252247,
+      "learning_rate": 9.38468271592359e-06,
+      "loss": 0.5577,
+      "step": 650
+    },
+    {
+      "epoch": 0.731954126377333,
+      "grad_norm": 3.5351140651169146,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 0.4235,
+      "step": 651
+    },
+    {
+      "epoch": 0.7330784798740724,
+      "grad_norm": 3.1606505971693504,
+      "learning_rate": 9.378376576876999e-06,
+      "loss": 0.5814,
+      "step": 652
+    },
+    {
+      "epoch": 0.7342028333708118,
+      "grad_norm": 3.009129597872765,
+      "learning_rate": 9.375212251305764e-06,
+      "loss": 0.5398,
+      "step": 653
+    },
+    {
+      "epoch": 0.7353271868675512,
+      "grad_norm": 3.3071616023597112,
+      "learning_rate": 9.372040428927595e-06,
+      "loss": 0.5309,
+      "step": 654
+    },
+    {
+      "epoch": 0.7364515403642905,
+      "grad_norm": 3.057212569129712,
+      "learning_rate": 9.368861115177327e-06,
+      "loss": 0.4983,
+      "step": 655
+    },
+    {
+      "epoch": 0.73757589386103,
+      "grad_norm": 2.883737348381482,
+      "learning_rate": 9.365674315502627e-06,
+      "loss": 0.4658,
+      "step": 656
+    },
+    {
+      "epoch": 0.7387002473577693,
+      "grad_norm": 2.833822389128405,
+      "learning_rate": 9.362480035363987e-06,
+      "loss": 0.5436,
+      "step": 657
+    },
+    {
+      "epoch": 0.7398246008545086,
+      "grad_norm": 3.1294766965723544,
+      "learning_rate": 9.35927828023472e-06,
+      "loss": 0.5441,
+      "step": 658
+    },
+    {
+      "epoch": 0.7409489543512481,
+      "grad_norm": 3.1457541319107225,
+      "learning_rate": 9.356069055600949e-06,
+      "loss": 0.5643,
+      "step": 659
+    },
+    {
+      "epoch": 0.7420733078479874,
+      "grad_norm": 3.3577166244784022,
+      "learning_rate": 9.352852366961588e-06,
+      "loss": 0.5666,
+      "step": 660
+    },
+    {
+      "epoch": 0.7431976613447268,
+      "grad_norm": 4.265603236377326,
+      "learning_rate": 9.349628219828348e-06,
+      "loss": 0.6204,
+      "step": 661
+    },
+    {
+      "epoch": 0.7443220148414662,
+      "grad_norm": 3.2988876066876864,
+      "learning_rate": 9.34639661972572e-06,
+      "loss": 0.4692,
+      "step": 662
+    },
+    {
+      "epoch": 0.7454463683382055,
+      "grad_norm": 3.3960992232200136,
+      "learning_rate": 9.343157572190957e-06,
+      "loss": 0.5533,
+      "step": 663
+    },
+    {
+      "epoch": 0.7465707218349449,
+      "grad_norm": 3.105418053336733,
+      "learning_rate": 9.339911082774083e-06,
+      "loss": 0.568,
+      "step": 664
+    },
+    {
+      "epoch": 0.7476950753316843,
+      "grad_norm": 2.993203438220799,
+      "learning_rate": 9.336657157037866e-06,
+      "loss": 0.4506,
+      "step": 665
+    },
+    {
+      "epoch": 0.7488194288284237,
+      "grad_norm": 3.07609541931759,
+      "learning_rate": 9.33339580055782e-06,
+      "loss": 0.5509,
+      "step": 666
+    },
+    {
+      "epoch": 0.749943782325163,
+      "grad_norm": 3.2903163373150797,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.5986,
+      "step": 667
+    },
+    {
+      "epoch": 0.7510681358219025,
+      "grad_norm": 3.1196241431107588,
+      "learning_rate": 9.326850817731952e-06,
+      "loss": 0.4427,
+      "step": 668
+    },
+    {
+      "epoch": 0.7521924893186418,
+      "grad_norm": 3.203480619910118,
+      "learning_rate": 9.323567202600777e-06,
+      "loss": 0.46,
+      "step": 669
+    },
+    {
+      "epoch": 0.7533168428153811,
+      "grad_norm": 2.962246227427346,
+      "learning_rate": 9.320276179155052e-06,
+      "loss": 0.461,
+      "step": 670
+    },
+    {
+      "epoch": 0.7544411963121205,
+      "grad_norm": 3.165207094640078,
+      "learning_rate": 9.316977753033858e-06,
+      "loss": 0.5605,
+      "step": 671
+    },
+    {
+      "epoch": 0.7555655498088599,
+      "grad_norm": 2.9923367903832565,
+      "learning_rate": 9.31367192988896e-06,
+      "loss": 0.5556,
+      "step": 672
+    },
+    {
+      "epoch": 0.7566899033055993,
+      "grad_norm": 2.893901060934259,
+      "learning_rate": 9.310358715384793e-06,
+      "loss": 0.5822,
+      "step": 673
+    },
+    {
+      "epoch": 0.7578142568023386,
+      "grad_norm": 3.0874690721882487,
+      "learning_rate": 9.30703811519846e-06,
+      "loss": 0.5077,
+      "step": 674
+    },
+    {
+      "epoch": 0.758938610299078,
+      "grad_norm": 3.626553711010787,
+      "learning_rate": 9.30371013501972e-06,
+      "loss": 0.4189,
+      "step": 675
+    },
+    {
+      "epoch": 0.7600629637958174,
+      "grad_norm": 3.0826055160265042,
+      "learning_rate": 9.300374780550973e-06,
+      "loss": 0.4939,
+      "step": 676
+    },
+    {
+      "epoch": 0.7611873172925567,
+      "grad_norm": 3.119653091793988,
+      "learning_rate": 9.297032057507264e-06,
+      "loss": 0.4295,
+      "step": 677
+    },
+    {
+      "epoch": 0.7623116707892962,
+      "grad_norm": 3.1551912158421187,
+      "learning_rate": 9.293681971616252e-06,
+      "loss": 0.6171,
+      "step": 678
+    },
+    {
+      "epoch": 0.7634360242860355,
+      "grad_norm": 3.22868459145578,
+      "learning_rate": 9.290324528618225e-06,
+      "loss": 0.4734,
+      "step": 679
+    },
+    {
+      "epoch": 0.7645603777827749,
+      "grad_norm": 2.810070941138584,
+      "learning_rate": 9.28695973426606e-06,
+      "loss": 0.4087,
+      "step": 680
+    },
+    {
+      "epoch": 0.7656847312795143,
+      "grad_norm": 3.4911092465792666,
+      "learning_rate": 9.28358759432525e-06,
+      "loss": 0.4907,
+      "step": 681
+    },
+    {
+      "epoch": 0.7668090847762536,
+      "grad_norm": 2.9224805357616557,
+      "learning_rate": 9.280208114573859e-06,
+      "loss": 0.4954,
+      "step": 682
+    },
+    {
+      "epoch": 0.767933438272993,
+      "grad_norm": 3.482136684849554,
+      "learning_rate": 9.276821300802535e-06,
+      "loss": 0.5723,
+      "step": 683
+    },
+    {
+      "epoch": 0.7690577917697324,
+      "grad_norm": 3.031610889194102,
+      "learning_rate": 9.27342715881449e-06,
+      "loss": 0.5574,
+      "step": 684
+    },
+    {
+      "epoch": 0.7701821452664718,
+      "grad_norm": 3.353227791627315,
+      "learning_rate": 9.270025694425497e-06,
+      "loss": 0.5462,
+      "step": 685
+    },
+    {
+      "epoch": 0.7713064987632111,
+      "grad_norm": 3.1674340808676673,
+      "learning_rate": 9.26661691346387e-06,
+      "loss": 0.5473,
+      "step": 686
+    },
+    {
+      "epoch": 0.7724308522599506,
+      "grad_norm": 3.1473391534902637,
+      "learning_rate": 9.263200821770462e-06,
+      "loss": 0.5137,
+      "step": 687
+    },
+    {
+      "epoch": 0.7735552057566899,
+      "grad_norm": 3.040201149384232,
+      "learning_rate": 9.259777425198655e-06,
+      "loss": 0.4799,
+      "step": 688
+    },
+    {
+      "epoch": 0.7746795592534292,
+      "grad_norm": 2.756329726064213,
+      "learning_rate": 9.256346729614342e-06,
+      "loss": 0.5285,
+      "step": 689
+    },
+    {
+      "epoch": 0.7758039127501687,
+      "grad_norm": 2.9213494894731094,
+      "learning_rate": 9.252908740895932e-06,
+      "loss": 0.5166,
+      "step": 690
+    },
+    {
+      "epoch": 0.776928266246908,
+      "grad_norm": 3.240246668415497,
+      "learning_rate": 9.24946346493432e-06,
+      "loss": 0.4846,
+      "step": 691
+    },
+    {
+      "epoch": 0.7780526197436474,
+      "grad_norm": 3.117791188104041,
+      "learning_rate": 9.246010907632894e-06,
+      "loss": 0.4295,
+      "step": 692
+    },
+    {
+      "epoch": 0.7791769732403868,
+      "grad_norm": 3.135492197553572,
+      "learning_rate": 9.242551074907519e-06,
+      "loss": 0.4834,
+      "step": 693
+    },
+    {
+      "epoch": 0.7803013267371262,
+      "grad_norm": 3.0016140695751212,
+      "learning_rate": 9.239083972686522e-06,
+      "loss": 0.4974,
+      "step": 694
+    },
+    {
+      "epoch": 0.7814256802338655,
+      "grad_norm": 3.1897825195154,
+      "learning_rate": 9.235609606910687e-06,
+      "loss": 0.5507,
+      "step": 695
+    },
+    {
+      "epoch": 0.7825500337306049,
+      "grad_norm": 2.9668267502571513,
+      "learning_rate": 9.232127983533247e-06,
+      "loss": 0.4908,
+      "step": 696
+    },
+    {
+      "epoch": 0.7836743872273443,
+      "grad_norm": 2.8826284678840675,
+      "learning_rate": 9.228639108519867e-06,
+      "loss": 0.5505,
+      "step": 697
+    },
+    {
+      "epoch": 0.7847987407240836,
+      "grad_norm": 2.950306297103887,
+      "learning_rate": 9.225142987848642e-06,
+      "loss": 0.488,
+      "step": 698
+    },
+    {
+      "epoch": 0.7859230942208231,
+      "grad_norm": 3.1442864557617125,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 0.5311,
+      "step": 699
+    },
+    {
+      "epoch": 0.7870474477175624,
+      "grad_norm": 3.8338917128509884,
+      "learning_rate": 9.218129033507084e-06,
+      "loss": 0.6407,
+      "step": 700
+    },
+    {
+      "epoch": 0.7881718012143017,
+      "grad_norm": 3.394073458450031,
+      "learning_rate": 9.214611211854973e-06,
+      "loss": 0.6384,
+      "step": 701
+    },
+    {
+      "epoch": 0.7892961547110412,
+      "grad_norm": 2.839970874812004,
+      "learning_rate": 9.211086168581433e-06,
+      "loss": 0.479,
+      "step": 702
+    },
+    {
+      "epoch": 0.7904205082077805,
+      "grad_norm": 2.9157038136882343,
+      "learning_rate": 9.207553909726532e-06,
+      "loss": 0.4967,
+      "step": 703
+    },
+    {
+      "epoch": 0.7915448617045199,
+      "grad_norm": 3.1243384192518464,
+      "learning_rate": 9.204014441342697e-06,
+      "loss": 0.5016,
+      "step": 704
+    },
+    {
+      "epoch": 0.7926692152012593,
+      "grad_norm": 3.0173797241578706,
+      "learning_rate": 9.20046776949471e-06,
+      "loss": 0.5234,
+      "step": 705
+    },
+    {
+      "epoch": 0.7937935686979987,
+      "grad_norm": 3.1276574337775433,
+      "learning_rate": 9.196913900259699e-06,
+      "loss": 0.4837,
+      "step": 706
+    },
+    {
+      "epoch": 0.794917922194738,
+      "grad_norm": 3.126070031597478,
+      "learning_rate": 9.193352839727122e-06,
+      "loss": 0.5213,
+      "step": 707
+    },
+    {
+      "epoch": 0.7960422756914775,
+      "grad_norm": 2.91064895383391,
+      "learning_rate": 9.189784593998757e-06,
+      "loss": 0.5268,
+      "step": 708
+    },
+    {
+      "epoch": 0.7971666291882168,
+      "grad_norm": 3.0516337856414735,
+      "learning_rate": 9.186209169188695e-06,
+      "loss": 0.5097,
+      "step": 709
+    },
+    {
+      "epoch": 0.7982909826849561,
+      "grad_norm": 3.120801990742874,
+      "learning_rate": 9.182626571423333e-06,
+      "loss": 0.5988,
+      "step": 710
+    },
+    {
+      "epoch": 0.7994153361816956,
+      "grad_norm": 2.834292276874291,
+      "learning_rate": 9.179036806841352e-06,
+      "loss": 0.4913,
+      "step": 711
+    },
+    {
+      "epoch": 0.8005396896784349,
+      "grad_norm": 3.254528125211088,
+      "learning_rate": 9.175439881593716e-06,
+      "loss": 0.4448,
+      "step": 712
+    },
+    {
+      "epoch": 0.8016640431751743,
+      "grad_norm": 3.4409402045803335,
+      "learning_rate": 9.17183580184366e-06,
+      "loss": 0.5677,
+      "step": 713
+    },
+    {
+      "epoch": 0.8027883966719136,
+      "grad_norm": 3.182168045975382,
+      "learning_rate": 9.168224573766673e-06,
+      "loss": 0.4528,
+      "step": 714
+    },
+    {
+      "epoch": 0.803912750168653,
+      "grad_norm": 3.462489449460724,
+      "learning_rate": 9.164606203550498e-06,
+      "loss": 0.5856,
+      "step": 715
+    },
+    {
+      "epoch": 0.8050371036653924,
+      "grad_norm": 3.0654786668548546,
+      "learning_rate": 9.160980697395114e-06,
+      "loss": 0.4207,
+      "step": 716
+    },
+    {
+      "epoch": 0.8061614571621317,
+      "grad_norm": 3.0965043625675515,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 0.5999,
+      "step": 717
+    },
+    {
+      "epoch": 0.8072858106588712,
+      "grad_norm": 3.121987430409932,
+      "learning_rate": 9.153708302127759e-06,
+      "loss": 0.4865,
+      "step": 718
+    },
+    {
+      "epoch": 0.8084101641556105,
+      "grad_norm": 3.1093288555103986,
+      "learning_rate": 9.150061425476839e-06,
+      "loss": 0.4808,
+      "step": 719
+    },
+    {
+      "epoch": 0.8095345176523498,
+      "grad_norm": 3.119425828932225,
+      "learning_rate": 9.14640743780879e-06,
+      "loss": 0.494,
+      "step": 720
+    },
+    {
+      "epoch": 0.8106588711490893,
+      "grad_norm": 3.079612590468189,
+      "learning_rate": 9.142746345384619e-06,
+      "loss": 0.5177,
+      "step": 721
+    },
+    {
+      "epoch": 0.8117832246458286,
+      "grad_norm": 3.4720698234166982,
+      "learning_rate": 9.139078154477512e-06,
+      "loss": 0.5409,
+      "step": 722
+    },
+    {
+      "epoch": 0.812907578142568,
+      "grad_norm": 3.2828392488948164,
+      "learning_rate": 9.13540287137281e-06,
+      "loss": 0.6447,
+      "step": 723
+    },
+    {
+      "epoch": 0.8140319316393074,
+      "grad_norm": 3.2510824795008584,
+      "learning_rate": 9.131720502368013e-06,
+      "loss": 0.5668,
+      "step": 724
+    },
+    {
+      "epoch": 0.8151562851360468,
+      "grad_norm": 2.7430845114318325,
+      "learning_rate": 9.12803105377276e-06,
+      "loss": 0.4691,
+      "step": 725
+    },
+    {
+      "epoch": 0.8162806386327861,
+      "grad_norm": 3.1318889883829586,
+      "learning_rate": 9.124334531908818e-06,
+      "loss": 0.6329,
+      "step": 726
+    },
+    {
+      "epoch": 0.8174049921295256,
+      "grad_norm": 2.9894288059820306,
+      "learning_rate": 9.120630943110078e-06,
+      "loss": 0.4049,
+      "step": 727
+    },
+    {
+      "epoch": 0.8185293456262649,
+      "grad_norm": 3.288441123474717,
+      "learning_rate": 9.11692029372254e-06,
+      "loss": 0.5551,
+      "step": 728
+    },
+    {
+      "epoch": 0.8196536991230042,
+      "grad_norm": 3.2343311773378742,
+      "learning_rate": 9.1132025901043e-06,
+      "loss": 0.5524,
+      "step": 729
+    },
+    {
+      "epoch": 0.8207780526197437,
+      "grad_norm": 3.1954240429536123,
+      "learning_rate": 9.10947783862554e-06,
+      "loss": 0.5488,
+      "step": 730
+    },
+    {
+      "epoch": 0.821902406116483,
+      "grad_norm": 3.1181995356742025,
+      "learning_rate": 9.10574604566852e-06,
+      "loss": 0.561,
+      "step": 731
+    },
+    {
+      "epoch": 0.8230267596132224,
+      "grad_norm": 3.3007706204174188,
+      "learning_rate": 9.102007217627568e-06,
+      "loss": 0.4816,
+      "step": 732
+    },
+    {
+      "epoch": 0.8241511131099618,
+      "grad_norm": 3.098959802248871,
+      "learning_rate": 9.098261360909064e-06,
+      "loss": 0.6114,
+      "step": 733
+    },
+    {
+      "epoch": 0.8252754666067011,
+      "grad_norm": 2.7408655941911,
+      "learning_rate": 9.094508481931428e-06,
+      "loss": 0.5079,
+      "step": 734
+    },
+    {
+      "epoch": 0.8263998201034405,
+      "grad_norm": 3.379854126381703,
+      "learning_rate": 9.090748587125118e-06,
+      "loss": 0.6316,
+      "step": 735
+    },
+    {
+      "epoch": 0.8275241736001799,
+      "grad_norm": 3.7169287785606104,
+      "learning_rate": 9.086981682932612e-06,
+      "loss": 0.4767,
+      "step": 736
+    },
+    {
+      "epoch": 0.8286485270969193,
+      "grad_norm": 11.855947411405696,
+      "learning_rate": 9.083207775808395e-06,
+      "loss": 0.5761,
+      "step": 737
+    },
+    {
+      "epoch": 0.8297728805936586,
+      "grad_norm": 3.470058982781829,
+      "learning_rate": 9.079426872218958e-06,
+      "loss": 0.5819,
+      "step": 738
+    },
+    {
+      "epoch": 0.8308972340903981,
+      "grad_norm": 3.0833740106776424,
+      "learning_rate": 9.07563897864277e-06,
+      "loss": 0.5912,
+      "step": 739
+    },
+    {
+      "epoch": 0.8320215875871374,
+      "grad_norm": 2.873714977425552,
+      "learning_rate": 9.071844101570291e-06,
+      "loss": 0.49,
+      "step": 740
+    },
+    {
+      "epoch": 0.8331459410838767,
+      "grad_norm": 2.8568996972948786,
+      "learning_rate": 9.068042247503937e-06,
+      "loss": 0.4727,
+      "step": 741
+    },
+    {
+      "epoch": 0.8342702945806162,
+      "grad_norm": 3.3331831900390014,
+      "learning_rate": 9.064233422958078e-06,
+      "loss": 0.5266,
+      "step": 742
+    },
+    {
+      "epoch": 0.8353946480773555,
+      "grad_norm": 2.8131564740790775,
+      "learning_rate": 9.060417634459032e-06,
+      "loss": 0.5103,
+      "step": 743
+    },
+    {
+      "epoch": 0.8365190015740949,
+      "grad_norm": 3.1111706346131616,
+      "learning_rate": 9.05659488854505e-06,
+      "loss": 0.4656,
+      "step": 744
+    },
+    {
+      "epoch": 0.8376433550708343,
+      "grad_norm": 3.138515399309868,
+      "learning_rate": 9.052765191766304e-06,
+      "loss": 0.4523,
+      "step": 745
+    },
+    {
+      "epoch": 0.8387677085675737,
+      "grad_norm": 3.086756720576352,
+      "learning_rate": 9.048928550684873e-06,
+      "loss": 0.4612,
+      "step": 746
+    },
+    {
+      "epoch": 0.839892062064313,
+      "grad_norm": 2.942844018158255,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.5185,
+      "step": 747
+    },
+    {
+      "epoch": 0.8410164155610524,
+      "grad_norm": 3.095834719714255,
+      "learning_rate": 9.041234461921766e-06,
+      "loss": 0.4256,
+      "step": 748
+    },
+    {
+      "epoch": 0.8421407690577918,
+      "grad_norm": 3.009744890920842,
+      "learning_rate": 9.0373770274237e-06,
+      "loss": 0.575,
+      "step": 749
+    },
+    {
+      "epoch": 0.8432651225545311,
+      "grad_norm": 2.9002670381872315,
+      "learning_rate": 9.033512674990151e-06,
+      "loss": 0.4775,
+      "step": 750
+    },
+    {
+      "epoch": 0.8443894760512706,
+      "grad_norm": 2.7824101321756545,
+      "learning_rate": 9.02964141124258e-06,
+      "loss": 0.4943,
+      "step": 751
+    },
+    {
+      "epoch": 0.8455138295480099,
+      "grad_norm": 2.9002876492102834,
+      "learning_rate": 9.025763242814291e-06,
+      "loss": 0.4448,
+      "step": 752
+    },
+    {
+      "epoch": 0.8466381830447492,
+      "grad_norm": 2.7934623724354233,
+      "learning_rate": 9.021878176350422e-06,
+      "loss": 0.4636,
+      "step": 753
+    },
+    {
+      "epoch": 0.8477625365414887,
+      "grad_norm": 3.017113218564395,
+      "learning_rate": 9.01798621850793e-06,
+      "loss": 0.4988,
+      "step": 754
+    },
+    {
+      "epoch": 0.848886890038228,
+      "grad_norm": 3.199798676799069,
+      "learning_rate": 9.014087375955574e-06,
+      "loss": 0.514,
+      "step": 755
+    },
+    {
+      "epoch": 0.8500112435349674,
+      "grad_norm": 2.9560909234773685,
+      "learning_rate": 9.010181655373918e-06,
+      "loss": 0.4864,
+      "step": 756
+    },
+    {
+      "epoch": 0.8511355970317067,
+      "grad_norm": 2.642759584717876,
+      "learning_rate": 9.006269063455305e-06,
+      "loss": 0.3787,
+      "step": 757
+    },
+    {
+      "epoch": 0.8522599505284462,
+      "grad_norm": 3.0813433337704845,
+      "learning_rate": 9.002349606903854e-06,
+      "loss": 0.5596,
+      "step": 758
+    },
+    {
+      "epoch": 0.8533843040251855,
+      "grad_norm": 3.907425211423246,
+      "learning_rate": 8.998423292435455e-06,
+      "loss": 0.57,
+      "step": 759
+    },
+    {
+      "epoch": 0.8545086575219248,
+      "grad_norm": 3.0825384820579513,
+      "learning_rate": 8.994490126777731e-06,
+      "loss": 0.4872,
+      "step": 760
+    },
+    {
+      "epoch": 0.8556330110186643,
+      "grad_norm": 2.7314078105935113,
+      "learning_rate": 8.990550116670059e-06,
+      "loss": 0.4826,
+      "step": 761
+    },
+    {
+      "epoch": 0.8567573645154036,
+      "grad_norm": 3.067778008788354,
+      "learning_rate": 8.986603268863536e-06,
+      "loss": 0.4698,
+      "step": 762
+    },
+    {
+      "epoch": 0.857881718012143,
+      "grad_norm": 3.1136410393636136,
+      "learning_rate": 8.982649590120982e-06,
+      "loss": 0.5015,
+      "step": 763
+    },
+    {
+      "epoch": 0.8590060715088824,
+      "grad_norm": 3.096745450459727,
+      "learning_rate": 8.978689087216917e-06,
+      "loss": 0.5395,
+      "step": 764
+    },
+    {
+      "epoch": 0.8601304250056218,
+      "grad_norm": 3.2229539871812976,
+      "learning_rate": 8.97472176693755e-06,
+      "loss": 0.516,
+      "step": 765
+    },
+    {
+      "epoch": 0.8612547785023611,
+      "grad_norm": 3.109680464220826,
+      "learning_rate": 8.970747636080783e-06,
+      "loss": 0.4514,
+      "step": 766
+    },
+    {
+      "epoch": 0.8623791319991005,
+      "grad_norm": 2.9901198551243406,
+      "learning_rate": 8.966766701456177e-06,
+      "loss": 0.4647,
+      "step": 767
+    },
+    {
+      "epoch": 0.8635034854958399,
+      "grad_norm": 3.3135298558861477,
+      "learning_rate": 8.962778969884956e-06,
+      "loss": 0.577,
+      "step": 768
+    },
+    {
+      "epoch": 0.8646278389925792,
+      "grad_norm": 2.669343823030547,
+      "learning_rate": 8.958784448199987e-06,
+      "loss": 0.3989,
+      "step": 769
+    },
+    {
+      "epoch": 0.8657521924893187,
+      "grad_norm": 3.2110616275015103,
+      "learning_rate": 8.954783143245778e-06,
+      "loss": 0.5665,
+      "step": 770
+    },
+    {
+      "epoch": 0.866876545986058,
+      "grad_norm": 3.0517839999115237,
+      "learning_rate": 8.950775061878453e-06,
+      "loss": 0.5279,
+      "step": 771
+    },
+    {
+      "epoch": 0.8680008994827974,
+      "grad_norm": 3.0843322054343947,
+      "learning_rate": 8.94676021096575e-06,
+      "loss": 0.5105,
+      "step": 772
+    },
+    {
+      "epoch": 0.8691252529795368,
+      "grad_norm": 2.9151526140133392,
+      "learning_rate": 8.94273859738701e-06,
+      "loss": 0.6153,
+      "step": 773
+    },
+    {
+      "epoch": 0.8702496064762761,
+      "grad_norm": 2.945712443919617,
+      "learning_rate": 8.938710228033155e-06,
+      "loss": 0.4815,
+      "step": 774
+    },
+    {
+      "epoch": 0.8713739599730155,
+      "grad_norm": 3.041656123013539,
+      "learning_rate": 8.934675109806688e-06,
+      "loss": 0.5637,
+      "step": 775
+    },
+    {
+      "epoch": 0.8724983134697549,
+      "grad_norm": 3.0900116293513666,
+      "learning_rate": 8.930633249621673e-06,
+      "loss": 0.4553,
+      "step": 776
+    },
+    {
+      "epoch": 0.8736226669664943,
+      "grad_norm": 2.9188745123124256,
+      "learning_rate": 8.926584654403725e-06,
+      "loss": 0.3863,
+      "step": 777
+    },
+    {
+      "epoch": 0.8747470204632336,
+      "grad_norm": 3.2342032214112617,
+      "learning_rate": 8.922529331090006e-06,
+      "loss": 0.6126,
+      "step": 778
+    },
+    {
+      "epoch": 0.875871373959973,
+      "grad_norm": 3.7953682023367166,
+      "learning_rate": 8.9184672866292e-06,
+      "loss": 0.5582,
+      "step": 779
+    },
+    {
+      "epoch": 0.8769957274567124,
+      "grad_norm": 3.355090844389576,
+      "learning_rate": 8.91439852798151e-06,
+      "loss": 0.5652,
+      "step": 780
+    },
+    {
+      "epoch": 0.8781200809534517,
+      "grad_norm": 3.7288668364755257,
+      "learning_rate": 8.91032306211864e-06,
+      "loss": 0.5102,
+      "step": 781
+    },
+    {
+      "epoch": 0.8792444344501912,
+      "grad_norm": 2.7977221885235006,
+      "learning_rate": 8.906240896023794e-06,
+      "loss": 0.4723,
+      "step": 782
+    },
+    {
+      "epoch": 0.8803687879469305,
+      "grad_norm": 3.2160255021183097,
+      "learning_rate": 8.902152036691649e-06,
+      "loss": 0.6058,
+      "step": 783
+    },
+    {
+      "epoch": 0.8814931414436699,
+      "grad_norm": 3.0292726174752502,
+      "learning_rate": 8.898056491128358e-06,
+      "loss": 0.5134,
+      "step": 784
+    },
+    {
+      "epoch": 0.8826174949404093,
+      "grad_norm": 2.7286711077506496,
+      "learning_rate": 8.893954266351522e-06,
+      "loss": 0.4538,
+      "step": 785
+    },
+    {
+      "epoch": 0.8837418484371486,
+      "grad_norm": 2.6728107624164212,
+      "learning_rate": 8.889845369390193e-06,
+      "loss": 0.4143,
+      "step": 786
+    },
+    {
+      "epoch": 0.884866201933888,
+      "grad_norm": 2.705847973673183,
+      "learning_rate": 8.885729807284855e-06,
+      "loss": 0.4026,
+      "step": 787
+    },
+    {
+      "epoch": 0.8859905554306274,
+      "grad_norm": 3.156033442908449,
+      "learning_rate": 8.881607587087412e-06,
+      "loss": 0.5609,
+      "step": 788
+    },
+    {
+      "epoch": 0.8871149089273668,
+      "grad_norm": 2.814062836921905,
+      "learning_rate": 8.877478715861173e-06,
+      "loss": 0.4869,
+      "step": 789
+    },
+    {
+      "epoch": 0.8882392624241061,
+      "grad_norm": 3.345750449849639,
+      "learning_rate": 8.873343200680848e-06,
+      "loss": 0.4553,
+      "step": 790
+    },
+    {
+      "epoch": 0.8893636159208456,
+      "grad_norm": 2.788662798354574,
+      "learning_rate": 8.869201048632531e-06,
+      "loss": 0.489,
+      "step": 791
+    },
+    {
+      "epoch": 0.8904879694175849,
+      "grad_norm": 2.797191596507121,
+      "learning_rate": 8.865052266813686e-06,
+      "loss": 0.4841,
+      "step": 792
+    },
+    {
+      "epoch": 0.8916123229143242,
+      "grad_norm": 3.001472535210457,
+      "learning_rate": 8.860896862333135e-06,
+      "loss": 0.4069,
+      "step": 793
+    },
+    {
+      "epoch": 0.8927366764110637,
+      "grad_norm": 3.275736148758101,
+      "learning_rate": 8.856734842311053e-06,
+      "loss": 0.5344,
+      "step": 794
+    },
+    {
+      "epoch": 0.893861029907803,
+      "grad_norm": 2.9851694260849655,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 0.4479,
+      "step": 795
+    },
+    {
+      "epoch": 0.8949853834045424,
+      "grad_norm": 3.244781039430138,
+      "learning_rate": 8.848390984179647e-06,
+      "loss": 0.5512,
+      "step": 796
+    },
+    {
+      "epoch": 0.8961097369012818,
+      "grad_norm": 3.1077188692851547,
+      "learning_rate": 8.844209160367298e-06,
+      "loss": 0.5329,
+      "step": 797
+    },
+    {
+      "epoch": 0.8972340903980212,
+      "grad_norm": 3.0929950405793356,
+      "learning_rate": 8.84002074960734e-06,
+      "loss": 0.5125,
+      "step": 798
+    },
+    {
+      "epoch": 0.8983584438947605,
+      "grad_norm": 3.4687138253487197,
+      "learning_rate": 8.8358257590765e-06,
+      "loss": 0.5359,
+      "step": 799
+    },
+    {
+      "epoch": 0.8994827973914998,
+      "grad_norm": 2.7645099066105727,
+      "learning_rate": 8.831624195962778e-06,
+      "loss": 0.5558,
+      "step": 800
+    },
+    {
+      "epoch": 0.9006071508882393,
+      "grad_norm": 2.609479825191883,
+      "learning_rate": 8.827416067465442e-06,
+      "loss": 0.555,
+      "step": 801
+    },
+    {
+      "epoch": 0.9017315043849786,
+      "grad_norm": 3.1529345778907403,
+      "learning_rate": 8.823201380795003e-06,
+      "loss": 0.4678,
+      "step": 802
+    },
+    {
+      "epoch": 0.902855857881718,
+      "grad_norm": 2.955295525802485,
+      "learning_rate": 8.818980143173212e-06,
+      "loss": 0.4069,
+      "step": 803
+    },
+    {
+      "epoch": 0.9039802113784574,
+      "grad_norm": 3.0322812948475115,
+      "learning_rate": 8.814752361833045e-06,
+      "loss": 0.4405,
+      "step": 804
+    },
+    {
+      "epoch": 0.9051045648751967,
+      "grad_norm": 3.273933069022393,
+      "learning_rate": 8.81051804401869e-06,
+      "loss": 0.5436,
+      "step": 805
+    },
+    {
+      "epoch": 0.9062289183719361,
+      "grad_norm": 3.0490851518787303,
+      "learning_rate": 8.806277196985536e-06,
+      "loss": 0.5254,
+      "step": 806
+    },
+    {
+      "epoch": 0.9073532718686755,
+      "grad_norm": 2.9516463802227073,
+      "learning_rate": 8.802029828000157e-06,
+      "loss": 0.4935,
+      "step": 807
+    },
+    {
+      "epoch": 0.9084776253654149,
+      "grad_norm": 3.048535531931336,
+      "learning_rate": 8.797775944340303e-06,
+      "loss": 0.5715,
+      "step": 808
+    },
+    {
+      "epoch": 0.9096019788621542,
+      "grad_norm": 3.632497158640449,
+      "learning_rate": 8.793515553294892e-06,
+      "loss": 0.6392,
+      "step": 809
+    },
+    {
+      "epoch": 0.9107263323588937,
+      "grad_norm": 2.9896310106566486,
+      "learning_rate": 8.789248662163985e-06,
+      "loss": 0.5984,
+      "step": 810
+    },
+    {
+      "epoch": 0.911850685855633,
+      "grad_norm": 2.953378610680634,
+      "learning_rate": 8.784975278258783e-06,
+      "loss": 0.5422,
+      "step": 811
+    },
+    {
+      "epoch": 0.9129750393523723,
+      "grad_norm": 3.1619650699429664,
+      "learning_rate": 8.780695408901613e-06,
+      "loss": 0.4736,
+      "step": 812
+    },
+    {
+      "epoch": 0.9140993928491118,
+      "grad_norm": 3.0232675386182235,
+      "learning_rate": 8.776409061425919e-06,
+      "loss": 0.5297,
+      "step": 813
+    },
+    {
+      "epoch": 0.9152237463458511,
+      "grad_norm": 3.5766434910622635,
+      "learning_rate": 8.772116243176235e-06,
+      "loss": 0.5294,
+      "step": 814
+    },
+    {
+      "epoch": 0.9163480998425905,
+      "grad_norm": 2.933881553382151,
+      "learning_rate": 8.767816961508191e-06,
+      "loss": 0.4507,
+      "step": 815
+    },
+    {
+      "epoch": 0.9174724533393299,
+      "grad_norm": 2.846121788281873,
+      "learning_rate": 8.763511223788485e-06,
+      "loss": 0.533,
+      "step": 816
+    },
+    {
+      "epoch": 0.9185968068360693,
+      "grad_norm": 3.1392470532883623,
+      "learning_rate": 8.759199037394888e-06,
+      "loss": 0.6066,
+      "step": 817
+    },
+    {
+      "epoch": 0.9197211603328086,
+      "grad_norm": 2.856841990607503,
+      "learning_rate": 8.754880409716208e-06,
+      "loss": 0.6032,
+      "step": 818
+    },
+    {
+      "epoch": 0.920845513829548,
+      "grad_norm": 2.902420522278277,
+      "learning_rate": 8.750555348152299e-06,
+      "loss": 0.5239,
+      "step": 819
+    },
+    {
+      "epoch": 0.9219698673262874,
+      "grad_norm": 3.056145314330929,
+      "learning_rate": 8.746223860114033e-06,
+      "loss": 0.5858,
+      "step": 820
+    },
+    {
+      "epoch": 0.9230942208230267,
+      "grad_norm": 3.1008678556514364,
+      "learning_rate": 8.741885953023302e-06,
+      "loss": 0.562,
+      "step": 821
+    },
+    {
+      "epoch": 0.9242185743197662,
+      "grad_norm": 3.243942409040953,
+      "learning_rate": 8.737541634312985e-06,
+      "loss": 0.5085,
+      "step": 822
+    },
+    {
+      "epoch": 0.9253429278165055,
+      "grad_norm": 2.730760687361521,
+      "learning_rate": 8.733190911426957e-06,
+      "loss": 0.4323,
+      "step": 823
+    },
+    {
+      "epoch": 0.9264672813132449,
+      "grad_norm": 3.1795363820493794,
+      "learning_rate": 8.728833791820064e-06,
+      "loss": 0.5649,
+      "step": 824
+    },
+    {
+      "epoch": 0.9275916348099843,
+      "grad_norm": 2.890208222309272,
+      "learning_rate": 8.72447028295811e-06,
+      "loss": 0.4839,
+      "step": 825
+    },
+    {
+      "epoch": 0.9287159883067236,
+      "grad_norm": 2.787570325148229,
+      "learning_rate": 8.720100392317851e-06,
+      "loss": 0.4636,
+      "step": 826
+    },
+    {
+      "epoch": 0.929840341803463,
+      "grad_norm": 3.2824360330329325,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.5288,
+      "step": 827
+    },
+    {
+      "epoch": 0.9309646953002024,
+      "grad_norm": 3.0187955942720883,
+      "learning_rate": 8.711341495664085e-06,
+      "loss": 0.5872,
+      "step": 828
+    },
+    {
+      "epoch": 0.9320890487969418,
+      "grad_norm": 3.014681725541038,
+      "learning_rate": 8.706952504658712e-06,
+      "loss": 0.4903,
+      "step": 829
+    },
+    {
+      "epoch": 0.9332134022936811,
+      "grad_norm": 2.720315178948105,
+      "learning_rate": 8.702557161891265e-06,
+      "loss": 0.5116,
+      "step": 830
+    },
+    {
+      "epoch": 0.9343377557904206,
+      "grad_norm": 3.089381322317717,
+      "learning_rate": 8.69815547489305e-06,
+      "loss": 0.6088,
+      "step": 831
+    },
+    {
+      "epoch": 0.9354621092871599,
+      "grad_norm": 3.379751551734169,
+      "learning_rate": 8.693747451206231e-06,
+      "loss": 0.5457,
+      "step": 832
+    },
+    {
+      "epoch": 0.9365864627838992,
+      "grad_norm": 2.996818010827567,
+      "learning_rate": 8.689333098383843e-06,
+      "loss": 0.5019,
+      "step": 833
+    },
+    {
+      "epoch": 0.9377108162806387,
+      "grad_norm": 3.0915568764083727,
+      "learning_rate": 8.684912423989755e-06,
+      "loss": 0.4927,
+      "step": 834
+    },
+    {
+      "epoch": 0.938835169777378,
+      "grad_norm": 2.590565999834234,
+      "learning_rate": 8.680485435598674e-06,
+      "loss": 0.4292,
+      "step": 835
+    },
+    {
+      "epoch": 0.9399595232741174,
+      "grad_norm": 2.802161714814858,
+      "learning_rate": 8.67605214079612e-06,
+      "loss": 0.4588,
+      "step": 836
+    },
+    {
+      "epoch": 0.9410838767708568,
+      "grad_norm": 2.736103716314624,
+      "learning_rate": 8.671612547178428e-06,
+      "loss": 0.4284,
+      "step": 837
+    },
+    {
+      "epoch": 0.9422082302675961,
+      "grad_norm": 3.5381219002251925,
+      "learning_rate": 8.66716666235272e-06,
+      "loss": 0.4447,
+      "step": 838
+    },
+    {
+      "epoch": 0.9433325837643355,
+      "grad_norm": 2.771222503349902,
+      "learning_rate": 8.662714493936895e-06,
+      "loss": 0.475,
+      "step": 839
+    },
+    {
+      "epoch": 0.9444569372610749,
+      "grad_norm": 2.8283582876627342,
+      "learning_rate": 8.658256049559624e-06,
+      "loss": 0.4948,
+      "step": 840
+    },
+    {
+      "epoch": 0.9455812907578143,
+      "grad_norm": 2.86026619238831,
+      "learning_rate": 8.653791336860331e-06,
+      "loss": 0.4756,
+      "step": 841
+    },
+    {
+      "epoch": 0.9467056442545536,
+      "grad_norm": 2.9124974736821607,
+      "learning_rate": 8.649320363489178e-06,
+      "loss": 0.5096,
+      "step": 842
+    },
+    {
+      "epoch": 0.947829997751293,
+      "grad_norm": 3.061547973972135,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 0.479,
+      "step": 843
+    },
+    {
+      "epoch": 0.9489543512480324,
+      "grad_norm": 3.0447996623655995,
+      "learning_rate": 8.640359665385575e-06,
+      "loss": 0.5929,
+      "step": 844
+    },
+    {
+      "epoch": 0.9500787047447717,
+      "grad_norm": 2.978527128934878,
+      "learning_rate": 8.635869956007034e-06,
+      "loss": 0.5586,
+      "step": 845
+    },
+    {
+      "epoch": 0.9512030582415111,
+      "grad_norm": 3.566486006363544,
+      "learning_rate": 8.631374016664434e-06,
+      "loss": 0.5509,
+      "step": 846
+    },
+    {
+      "epoch": 0.9523274117382505,
+      "grad_norm": 2.85871676868812,
+      "learning_rate": 8.626871855061438e-06,
+      "loss": 0.5143,
+      "step": 847
+    },
+    {
+      "epoch": 0.9534517652349899,
+      "grad_norm": 2.5498593289900016,
+      "learning_rate": 8.622363478912384e-06,
+      "loss": 0.4587,
+      "step": 848
+    },
+    {
+      "epoch": 0.9545761187317292,
+      "grad_norm": 2.9403688659831735,
+      "learning_rate": 8.617848895942246e-06,
+      "loss": 0.5529,
+      "step": 849
+    },
+    {
+      "epoch": 0.9557004722284687,
+      "grad_norm": 2.9465773832169866,
+      "learning_rate": 8.613328113886644e-06,
+      "loss": 0.4959,
+      "step": 850
+    },
+    {
+      "epoch": 0.956824825725208,
+      "grad_norm": 2.9482612668984713,
+      "learning_rate": 8.608801140491811e-06,
+      "loss": 0.4992,
+      "step": 851
+    },
+    {
+      "epoch": 0.9579491792219473,
+      "grad_norm": 2.98387683641091,
+      "learning_rate": 8.604267983514595e-06,
+      "loss": 0.4986,
+      "step": 852
+    },
+    {
+      "epoch": 0.9590735327186868,
+      "grad_norm": 2.7836707825255544,
+      "learning_rate": 8.599728650722435e-06,
+      "loss": 0.4102,
+      "step": 853
+    },
+    {
+      "epoch": 0.9601978862154261,
+      "grad_norm": 2.60806714362599,
+      "learning_rate": 8.595183149893353e-06,
+      "loss": 0.4394,
+      "step": 854
+    },
+    {
+      "epoch": 0.9613222397121655,
+      "grad_norm": 2.868515476959801,
+      "learning_rate": 8.590631488815945e-06,
+      "loss": 0.5104,
+      "step": 855
+    },
+    {
+      "epoch": 0.9624465932089049,
+      "grad_norm": 2.887627862416009,
+      "learning_rate": 8.586073675289355e-06,
+      "loss": 0.5252,
+      "step": 856
+    },
+    {
+      "epoch": 0.9635709467056442,
+      "grad_norm": 2.890126330889477,
+      "learning_rate": 8.581509717123272e-06,
+      "loss": 0.5088,
+      "step": 857
+    },
+    {
+      "epoch": 0.9646953002023836,
+      "grad_norm": 3.077054036431364,
+      "learning_rate": 8.576939622137915e-06,
+      "loss": 0.4816,
+      "step": 858
+    },
+    {
+      "epoch": 0.965819653699123,
+      "grad_norm": 2.8284888068174165,
+      "learning_rate": 8.572363398164017e-06,
+      "loss": 0.5208,
+      "step": 859
+    },
+    {
+      "epoch": 0.9669440071958624,
+      "grad_norm": 2.646072290731001,
+      "learning_rate": 8.567781053042813e-06,
+      "loss": 0.4088,
+      "step": 860
+    },
+    {
+      "epoch": 0.9680683606926017,
+      "grad_norm": 3.1079449358216227,
+      "learning_rate": 8.563192594626027e-06,
+      "loss": 0.616,
+      "step": 861
+    },
+    {
+      "epoch": 0.9691927141893412,
+      "grad_norm": 2.944728511036786,
+      "learning_rate": 8.558598030775857e-06,
+      "loss": 0.482,
+      "step": 862
+    },
+    {
+      "epoch": 0.9703170676860805,
+      "grad_norm": 2.737039620571577,
+      "learning_rate": 8.553997369364964e-06,
+      "loss": 0.4522,
+      "step": 863
+    },
+    {
+      "epoch": 0.9714414211828198,
+      "grad_norm": 2.6278667535966203,
+      "learning_rate": 8.549390618276452e-06,
+      "loss": 0.4213,
+      "step": 864
+    },
+    {
+      "epoch": 0.9725657746795593,
+      "grad_norm": 3.5843513943262226,
+      "learning_rate": 8.544777785403869e-06,
+      "loss": 0.527,
+      "step": 865
+    },
+    {
+      "epoch": 0.9736901281762986,
+      "grad_norm": 3.2068216994066114,
+      "learning_rate": 8.540158878651173e-06,
+      "loss": 0.5175,
+      "step": 866
+    },
+    {
+      "epoch": 0.974814481673038,
+      "grad_norm": 2.404746055652824,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4418,
+      "step": 867
+    },
+    {
+      "epoch": 0.9759388351697774,
+      "grad_norm": 3.3887264655772076,
+      "learning_rate": 8.530902875173329e-06,
+      "loss": 0.5521,
+      "step": 868
+    },
+    {
+      "epoch": 0.9770631886665168,
+      "grad_norm": 3.2251229962463865,
+      "learning_rate": 8.526265794308088e-06,
+      "loss": 0.4521,
+      "step": 869
+    },
+    {
+      "epoch": 0.9781875421632561,
+      "grad_norm": 3.3025923834907975,
+      "learning_rate": 8.521622671282532e-06,
+      "loss": 0.4626,
+      "step": 870
+    },
+    {
+      "epoch": 0.9793118956599955,
+      "grad_norm": 2.7506802588053487,
+      "learning_rate": 8.51697351405252e-06,
+      "loss": 0.4438,
+      "step": 871
+    },
+    {
+      "epoch": 0.9804362491567349,
+      "grad_norm": 3.336477118287713,
+      "learning_rate": 8.51231833058426e-06,
+      "loss": 0.5539,
+      "step": 872
+    },
+    {
+      "epoch": 0.9815606026534742,
+      "grad_norm": 2.818153272115007,
+      "learning_rate": 8.50765712885428e-06,
+      "loss": 0.4555,
+      "step": 873
+    },
+    {
+      "epoch": 0.9826849561502137,
+      "grad_norm": 3.3124729755407936,
+      "learning_rate": 8.502989916849422e-06,
+      "loss": 0.572,
+      "step": 874
+    },
+    {
+      "epoch": 0.983809309646953,
+      "grad_norm": 2.8002041917884606,
+      "learning_rate": 8.498316702566828e-06,
+      "loss": 0.4916,
+      "step": 875
+    },
+    {
+      "epoch": 0.9849336631436924,
+      "grad_norm": 3.107616322212631,
+      "learning_rate": 8.493637494013922e-06,
+      "loss": 0.5143,
+      "step": 876
+    },
+    {
+      "epoch": 0.9860580166404318,
+      "grad_norm": 3.088533850820032,
+      "learning_rate": 8.488952299208402e-06,
+      "loss": 0.5498,
+      "step": 877
+    },
+    {
+      "epoch": 0.9871823701371711,
+      "grad_norm": 3.2479430993698877,
+      "learning_rate": 8.48426112617822e-06,
+      "loss": 0.5091,
+      "step": 878
+    },
+    {
+      "epoch": 0.9883067236339105,
+      "grad_norm": 3.183438847450482,
+      "learning_rate": 8.479563982961572e-06,
+      "loss": 0.5405,
+      "step": 879
+    },
+    {
+      "epoch": 0.9894310771306499,
+      "grad_norm": 2.644912085886114,
+      "learning_rate": 8.474860877606887e-06,
+      "loss": 0.452,
+      "step": 880
+    },
+    {
+      "epoch": 0.9905554306273893,
+      "grad_norm": 3.255510181893681,
+      "learning_rate": 8.47015181817281e-06,
+      "loss": 0.4597,
+      "step": 881
+    },
+    {
+      "epoch": 0.9916797841241286,
+      "grad_norm": 2.7322631892414204,
+      "learning_rate": 8.465436812728181e-06,
+      "loss": 0.4729,
+      "step": 882
+    },
+    {
+      "epoch": 0.992804137620868,
+      "grad_norm": 3.2017300524246046,
+      "learning_rate": 8.460715869352035e-06,
+      "loss": 0.5045,
+      "step": 883
+    },
+    {
+      "epoch": 0.9939284911176074,
+      "grad_norm": 2.935536468718181,
+      "learning_rate": 8.455988996133583e-06,
+      "loss": 0.5671,
+      "step": 884
+    },
+    {
+      "epoch": 0.9950528446143467,
+      "grad_norm": 2.9618968291507484,
+      "learning_rate": 8.451256201172186e-06,
+      "loss": 0.5082,
+      "step": 885
+    },
+    {
+      "epoch": 0.9961771981110861,
+      "grad_norm": 2.9729167655528905,
+      "learning_rate": 8.446517492577368e-06,
+      "loss": 0.5274,
+      "step": 886
+    },
+    {
+      "epoch": 0.9973015516078255,
+      "grad_norm": 2.696551403581966,
+      "learning_rate": 8.44177287846877e-06,
+      "loss": 0.4691,
+      "step": 887
+    },
+    {
+      "epoch": 0.9984259051045649,
+      "grad_norm": 2.9156442750630163,
+      "learning_rate": 8.437022366976165e-06,
+      "loss": 0.6015,
+      "step": 888
+    },
+    {
+      "epoch": 0.9995502586013042,
+      "grad_norm": 2.944217932610134,
+      "learning_rate": 8.432265966239419e-06,
+      "loss": 0.4508,
+      "step": 889
+    },
+    {
+      "epoch": 1.0006746120980436,
+      "grad_norm": 2.8373692207098844,
+      "learning_rate": 8.427503684408498e-06,
+      "loss": 0.4514,
+      "step": 890
+    },
+    {
+      "epoch": 1.001798965594783,
+      "grad_norm": 2.421494877150214,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 0.3348,
+      "step": 891
+    },
+    {
+      "epoch": 1.0029233190915223,
+      "grad_norm": 3.1271036601912137,
+      "learning_rate": 8.417961510114357e-06,
+      "loss": 0.393,
+      "step": 892
+    },
+    {
+      "epoch": 1.0040476725882617,
+      "grad_norm": 2.1384023838974215,
+      "learning_rate": 8.41318163400139e-06,
+      "loss": 0.2838,
+      "step": 893
+    },
+    {
+      "epoch": 1.0051720260850012,
+      "grad_norm": 2.5582384822827704,
+      "learning_rate": 8.408395909494733e-06,
+      "loss": 0.371,
+      "step": 894
+    },
+    {
+      "epoch": 1.0062963795817406,
+      "grad_norm": 2.265354192014569,
+      "learning_rate": 8.40360434479459e-06,
+      "loss": 0.307,
+      "step": 895
+    },
+    {
+      "epoch": 1.00742073307848,
+      "grad_norm": 2.4271071173294305,
+      "learning_rate": 8.398806948111181e-06,
+      "loss": 0.3428,
+      "step": 896
+    },
+    {
+      "epoch": 1.0085450865752192,
+      "grad_norm": 2.2663171754544615,
+      "learning_rate": 8.39400372766471e-06,
+      "loss": 0.3603,
+      "step": 897
+    },
+    {
+      "epoch": 1.0096694400719586,
+      "grad_norm": 2.3978806235829455,
+      "learning_rate": 8.389194691685368e-06,
+      "loss": 0.302,
+      "step": 898
+    },
+    {
+      "epoch": 1.010793793568698,
+      "grad_norm": 2.724683991463916,
+      "learning_rate": 8.384379848413304e-06,
+      "loss": 0.2669,
+      "step": 899
+    },
+    {
+      "epoch": 1.0119181470654375,
+      "grad_norm": 2.52241418794142,
+      "learning_rate": 8.379559206098625e-06,
+      "loss": 0.2754,
+      "step": 900
+    },
+    {
+      "epoch": 1.0130425005621768,
+      "grad_norm": 3.014709383498324,
+      "learning_rate": 8.374732773001367e-06,
+      "loss": 0.3692,
+      "step": 901
+    },
+    {
+      "epoch": 1.0141668540589162,
+      "grad_norm": 2.755121369269042,
+      "learning_rate": 8.36990055739149e-06,
+      "loss": 0.3333,
+      "step": 902
+    },
+    {
+      "epoch": 1.0152912075556555,
+      "grad_norm": 2.4474500654207345,
+      "learning_rate": 8.365062567548868e-06,
+      "loss": 0.2562,
+      "step": 903
+    },
+    {
+      "epoch": 1.0164155610523948,
+      "grad_norm": 2.6001967850626673,
+      "learning_rate": 8.36021881176326e-06,
+      "loss": 0.272,
+      "step": 904
+    },
+    {
+      "epoch": 1.0175399145491342,
+      "grad_norm": 2.7089841494253126,
+      "learning_rate": 8.355369298334316e-06,
+      "loss": 0.3236,
+      "step": 905
+    },
+    {
+      "epoch": 1.0186642680458737,
+      "grad_norm": 3.1413520025837465,
+      "learning_rate": 8.350514035571539e-06,
+      "loss": 0.336,
+      "step": 906
+    },
+    {
+      "epoch": 1.019788621542613,
+      "grad_norm": 8.667446558884613,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.3125,
+      "step": 907
+    },
+    {
+      "epoch": 1.0209129750393524,
+      "grad_norm": 2.7220654482281152,
+      "learning_rate": 8.340786295331772e-06,
+      "loss": 0.3242,
+      "step": 908
+    },
+    {
+      "epoch": 1.0220373285360917,
+      "grad_norm": 2.9392681078678446,
+      "learning_rate": 8.335913834522999e-06,
+      "loss": 0.3381,
+      "step": 909
+    },
+    {
+      "epoch": 1.023161682032831,
+      "grad_norm": 3.1025555853424103,
+      "learning_rate": 8.331035657716802e-06,
+      "loss": 0.3388,
+      "step": 910
+    },
+    {
+      "epoch": 1.0242860355295704,
+      "grad_norm": 2.4134981239919413,
+      "learning_rate": 8.326151773271805e-06,
+      "loss": 0.227,
+      "step": 911
+    },
+    {
+      "epoch": 1.0254103890263098,
+      "grad_norm": 3.027628308180082,
+      "learning_rate": 8.32126218955641e-06,
+      "loss": 0.2623,
+      "step": 912
+    },
+    {
+      "epoch": 1.0265347425230493,
+      "grad_norm": 2.7700677662348743,
+      "learning_rate": 8.316366914948783e-06,
+      "loss": 0.3002,
+      "step": 913
+    },
+    {
+      "epoch": 1.0276590960197887,
+      "grad_norm": 2.8927954497283728,
+      "learning_rate": 8.311465957836848e-06,
+      "loss": 0.3095,
+      "step": 914
+    },
+    {
+      "epoch": 1.028783449516528,
+      "grad_norm": 2.9498357628117637,
+      "learning_rate": 8.30655932661826e-06,
+      "loss": 0.3919,
+      "step": 915
+    },
+    {
+      "epoch": 1.0299078030132673,
+      "grad_norm": 2.4501640573943915,
+      "learning_rate": 8.301647029700396e-06,
+      "loss": 0.2285,
+      "step": 916
+    },
+    {
+      "epoch": 1.0310321565100067,
+      "grad_norm": 2.96370051162635,
+      "learning_rate": 8.296729075500345e-06,
+      "loss": 0.295,
+      "step": 917
+    },
+    {
+      "epoch": 1.032156510006746,
+      "grad_norm": 3.9607065031330233,
+      "learning_rate": 8.291805472444887e-06,
+      "loss": 0.317,
+      "step": 918
+    },
+    {
+      "epoch": 1.0332808635034856,
+      "grad_norm": 3.0456517300841353,
+      "learning_rate": 8.28687622897048e-06,
+      "loss": 0.3286,
+      "step": 919
+    },
+    {
+      "epoch": 1.034405217000225,
+      "grad_norm": 2.795273036156408,
+      "learning_rate": 8.28194135352325e-06,
+      "loss": 0.2652,
+      "step": 920
+    },
+    {
+      "epoch": 1.0355295704969643,
+      "grad_norm": 3.696104442684044,
+      "learning_rate": 8.27700085455897e-06,
+      "loss": 0.2788,
+      "step": 921
+    },
+    {
+      "epoch": 1.0366539239937036,
+      "grad_norm": 2.490867499502248,
+      "learning_rate": 8.272054740543053e-06,
+      "loss": 0.2839,
+      "step": 922
+    },
+    {
+      "epoch": 1.037778277490443,
+      "grad_norm": 2.6932480020966083,
+      "learning_rate": 8.267103019950529e-06,
+      "loss": 0.3306,
+      "step": 923
+    },
+    {
+      "epoch": 1.0389026309871823,
+      "grad_norm": 2.510338411712444,
+      "learning_rate": 8.262145701266034e-06,
+      "loss": 0.2729,
+      "step": 924
+    },
+    {
+      "epoch": 1.0400269844839218,
+      "grad_norm": 2.674830915570119,
+      "learning_rate": 8.257182792983802e-06,
+      "loss": 0.277,
+      "step": 925
+    },
+    {
+      "epoch": 1.0411513379806612,
+      "grad_norm": 3.035526911165084,
+      "learning_rate": 8.252214303607639e-06,
+      "loss": 0.366,
+      "step": 926
+    },
+    {
+      "epoch": 1.0422756914774005,
+      "grad_norm": 2.7692473475734514,
+      "learning_rate": 8.247240241650918e-06,
+      "loss": 0.2897,
+      "step": 927
+    },
+    {
+      "epoch": 1.0434000449741399,
+      "grad_norm": 2.536303962286788,
+      "learning_rate": 8.24226061563656e-06,
+      "loss": 0.2376,
+      "step": 928
+    },
+    {
+      "epoch": 1.0445243984708792,
+      "grad_norm": 2.904352676095586,
+      "learning_rate": 8.237275434097011e-06,
+      "loss": 0.2552,
+      "step": 929
+    },
+    {
+      "epoch": 1.0456487519676185,
+      "grad_norm": 3.441783039978973,
+      "learning_rate": 8.232284705574251e-06,
+      "loss": 0.3121,
+      "step": 930
+    },
+    {
+      "epoch": 1.046773105464358,
+      "grad_norm": 2.752464386921845,
+      "learning_rate": 8.227288438619754e-06,
+      "loss": 0.3095,
+      "step": 931
+    },
+    {
+      "epoch": 1.0478974589610974,
+      "grad_norm": 2.5255327563288943,
+      "learning_rate": 8.222286641794488e-06,
+      "loss": 0.2717,
+      "step": 932
+    },
+    {
+      "epoch": 1.0490218124578368,
+      "grad_norm": 2.5887681257914825,
+      "learning_rate": 8.217279323668895e-06,
+      "loss": 0.3057,
+      "step": 933
+    },
+    {
+      "epoch": 1.050146165954576,
+      "grad_norm": 6.139432906918701,
+      "learning_rate": 8.21226649282288e-06,
+      "loss": 0.3112,
+      "step": 934
+    },
+    {
+      "epoch": 1.0512705194513154,
+      "grad_norm": 3.216955306669303,
+      "learning_rate": 8.20724815784579e-06,
+      "loss": 0.2733,
+      "step": 935
+    },
+    {
+      "epoch": 1.0523948729480548,
+      "grad_norm": 2.8831623698208695,
+      "learning_rate": 8.202224327336406e-06,
+      "loss": 0.3012,
+      "step": 936
+    },
+    {
+      "epoch": 1.0535192264447943,
+      "grad_norm": 3.2853145208040337,
+      "learning_rate": 8.197195009902924e-06,
+      "loss": 0.327,
+      "step": 937
+    },
+    {
+      "epoch": 1.0546435799415337,
+      "grad_norm": 3.0742014255938437,
+      "learning_rate": 8.192160214162943e-06,
+      "loss": 0.3684,
+      "step": 938
+    },
+    {
+      "epoch": 1.055767933438273,
+      "grad_norm": 3.078255499590101,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 0.3425,
+      "step": 939
+    },
+    {
+      "epoch": 1.0568922869350124,
+      "grad_norm": 2.611442029307992,
+      "learning_rate": 8.182074222280798e-06,
+      "loss": 0.2793,
+      "step": 940
+    },
+    {
+      "epoch": 1.0580166404317517,
+      "grad_norm": 2.472649621651664,
+      "learning_rate": 8.177023043420706e-06,
+      "loss": 0.2821,
+      "step": 941
+    },
+    {
+      "epoch": 1.059140993928491,
+      "grad_norm": 2.6632801805199793,
+      "learning_rate": 8.171966420818227e-06,
+      "loss": 0.3047,
+      "step": 942
+    },
+    {
+      "epoch": 1.0602653474252306,
+      "grad_norm": 2.847986080746418,
+      "learning_rate": 8.166904363137752e-06,
+      "loss": 0.2831,
+      "step": 943
+    },
+    {
+      "epoch": 1.06138970092197,
+      "grad_norm": 2.929109853483684,
+      "learning_rate": 8.161836879052973e-06,
+      "loss": 0.341,
+      "step": 944
+    },
+    {
+      "epoch": 1.0625140544187093,
+      "grad_norm": 2.6412790614632238,
+      "learning_rate": 8.15676397724689e-06,
+      "loss": 0.267,
+      "step": 945
+    },
+    {
+      "epoch": 1.0636384079154486,
+      "grad_norm": 2.638730286195992,
+      "learning_rate": 8.15168566641178e-06,
+      "loss": 0.2622,
+      "step": 946
+    },
+    {
+      "epoch": 1.064762761412188,
+      "grad_norm": 2.669678293337342,
+      "learning_rate": 8.146601955249187e-06,
+      "loss": 0.2549,
+      "step": 947
+    },
+    {
+      "epoch": 1.0658871149089273,
+      "grad_norm": 2.885252685129536,
+      "learning_rate": 8.141512852469919e-06,
+      "loss": 0.2776,
+      "step": 948
+    },
+    {
+      "epoch": 1.0670114684056666,
+      "grad_norm": 2.4695891955909315,
+      "learning_rate": 8.13641836679401e-06,
+      "loss": 0.2963,
+      "step": 949
+    },
+    {
+      "epoch": 1.0681358219024062,
+      "grad_norm": 2.693010396780015,
+      "learning_rate": 8.13131850695072e-06,
+      "loss": 0.2613,
+      "step": 950
+    },
+    {
+      "epoch": 1.0692601753991455,
+      "grad_norm": 2.5025212771556,
+      "learning_rate": 8.126213281678527e-06,
+      "loss": 0.2914,
+      "step": 951
+    },
+    {
+      "epoch": 1.0703845288958849,
+      "grad_norm": 2.772492169426272,
+      "learning_rate": 8.12110269972509e-06,
+      "loss": 0.2689,
+      "step": 952
+    },
+    {
+      "epoch": 1.0715088823926242,
+      "grad_norm": 2.7695154169243414,
+      "learning_rate": 8.115986769847253e-06,
+      "loss": 0.2514,
+      "step": 953
+    },
+    {
+      "epoch": 1.0726332358893635,
+      "grad_norm": 2.8879402501985436,
+      "learning_rate": 8.110865500811022e-06,
+      "loss": 0.301,
+      "step": 954
+    },
+    {
+      "epoch": 1.073757589386103,
+      "grad_norm": 3.6494859162379214,
+      "learning_rate": 8.105738901391553e-06,
+      "loss": 0.3618,
+      "step": 955
+    },
+    {
+      "epoch": 1.0748819428828424,
+      "grad_norm": 2.5463621397134846,
+      "learning_rate": 8.100606980373133e-06,
+      "loss": 0.2973,
+      "step": 956
+    },
+    {
+      "epoch": 1.0760062963795818,
+      "grad_norm": 2.5990847114916966,
+      "learning_rate": 8.095469746549172e-06,
+      "loss": 0.2773,
+      "step": 957
+    },
+    {
+      "epoch": 1.0771306498763211,
+      "grad_norm": 2.3111605451157557,
+      "learning_rate": 8.090327208722176e-06,
+      "loss": 0.1786,
+      "step": 958
+    },
+    {
+      "epoch": 1.0782550033730605,
+      "grad_norm": 2.875532675879894,
+      "learning_rate": 8.085179375703745e-06,
+      "loss": 0.2905,
+      "step": 959
+    },
+    {
+      "epoch": 1.0793793568697998,
+      "grad_norm": 3.382762664563662,
+      "learning_rate": 8.08002625631455e-06,
+      "loss": 0.3215,
+      "step": 960
+    },
+    {
+      "epoch": 1.0805037103665391,
+      "grad_norm": 2.46347420590634,
+      "learning_rate": 8.074867859384322e-06,
+      "loss": 0.2936,
+      "step": 961
+    },
+    {
+      "epoch": 1.0816280638632787,
+      "grad_norm": 2.404432542630058,
+      "learning_rate": 8.069704193751834e-06,
+      "loss": 0.2593,
+      "step": 962
+    },
+    {
+      "epoch": 1.082752417360018,
+      "grad_norm": 3.214551180769464,
+      "learning_rate": 8.064535268264883e-06,
+      "loss": 0.3675,
+      "step": 963
+    },
+    {
+      "epoch": 1.0838767708567574,
+      "grad_norm": 2.7974627751021957,
+      "learning_rate": 8.059361091780285e-06,
+      "loss": 0.3174,
+      "step": 964
+    },
+    {
+      "epoch": 1.0850011243534967,
+      "grad_norm": 2.5952837614586763,
+      "learning_rate": 8.05418167316385e-06,
+      "loss": 0.2869,
+      "step": 965
+    },
+    {
+      "epoch": 1.086125477850236,
+      "grad_norm": 2.629250204753398,
+      "learning_rate": 8.04899702129037e-06,
+      "loss": 0.2983,
+      "step": 966
+    },
+    {
+      "epoch": 1.0872498313469754,
+      "grad_norm": 2.6192848094451793,
+      "learning_rate": 8.043807145043604e-06,
+      "loss": 0.3204,
+      "step": 967
+    },
+    {
+      "epoch": 1.088374184843715,
+      "grad_norm": 2.488706673471521,
+      "learning_rate": 8.038612053316264e-06,
+      "loss": 0.2795,
+      "step": 968
+    },
+    {
+      "epoch": 1.0894985383404543,
+      "grad_norm": 2.47370192954873,
+      "learning_rate": 8.033411755009999e-06,
+      "loss": 0.2659,
+      "step": 969
+    },
+    {
+      "epoch": 1.0906228918371936,
+      "grad_norm": 2.6079212050944642,
+      "learning_rate": 8.028206259035378e-06,
+      "loss": 0.2821,
+      "step": 970
+    },
+    {
+      "epoch": 1.091747245333933,
+      "grad_norm": 3.6298050306866094,
+      "learning_rate": 8.022995574311876e-06,
+      "loss": 0.3266,
+      "step": 971
+    },
+    {
+      "epoch": 1.0928715988306723,
+      "grad_norm": 2.775266732883481,
+      "learning_rate": 8.017779709767857e-06,
+      "loss": 0.2199,
+      "step": 972
+    },
+    {
+      "epoch": 1.0939959523274116,
+      "grad_norm": 2.760739856512836,
+      "learning_rate": 8.012558674340567e-06,
+      "loss": 0.2855,
+      "step": 973
+    },
+    {
+      "epoch": 1.0951203058241512,
+      "grad_norm": 2.855329661594668,
+      "learning_rate": 8.007332476976106e-06,
+      "loss": 0.3089,
+      "step": 974
+    },
+    {
+      "epoch": 1.0962446593208905,
+      "grad_norm": 2.9762603273432187,
+      "learning_rate": 8.002101126629422e-06,
+      "loss": 0.3417,
+      "step": 975
+    },
+    {
+      "epoch": 1.0973690128176299,
+      "grad_norm": 2.843740637749373,
+      "learning_rate": 7.996864632264287e-06,
+      "loss": 0.2905,
+      "step": 976
+    },
+    {
+      "epoch": 1.0984933663143692,
+      "grad_norm": 3.335780584873073,
+      "learning_rate": 7.991623002853296e-06,
+      "loss": 0.277,
+      "step": 977
+    },
+    {
+      "epoch": 1.0996177198111086,
+      "grad_norm": 2.938345079454318,
+      "learning_rate": 7.986376247377835e-06,
+      "loss": 0.3245,
+      "step": 978
+    },
+    {
+      "epoch": 1.100742073307848,
+      "grad_norm": 2.833478528627048,
+      "learning_rate": 7.981124374828079e-06,
+      "loss": 0.3074,
+      "step": 979
+    },
+    {
+      "epoch": 1.1018664268045875,
+      "grad_norm": 3.2326496735459305,
+      "learning_rate": 7.975867394202968e-06,
+      "loss": 0.2949,
+      "step": 980
+    },
+    {
+      "epoch": 1.1029907803013268,
+      "grad_norm": 3.1946375170172874,
+      "learning_rate": 7.970605314510194e-06,
+      "loss": 0.3464,
+      "step": 981
+    },
+    {
+      "epoch": 1.1041151337980661,
+      "grad_norm": 2.7783161055056507,
+      "learning_rate": 7.965338144766186e-06,
+      "loss": 0.282,
+      "step": 982
+    },
+    {
+      "epoch": 1.1052394872948055,
+      "grad_norm": 2.4328346493202835,
+      "learning_rate": 7.960065893996099e-06,
+      "loss": 0.2469,
+      "step": 983
+    },
+    {
+      "epoch": 1.1063638407915448,
+      "grad_norm": 2.8962553571764738,
+      "learning_rate": 7.954788571233788e-06,
+      "loss": 0.3045,
+      "step": 984
+    },
+    {
+      "epoch": 1.1074881942882842,
+      "grad_norm": 3.042569281967086,
+      "learning_rate": 7.949506185521802e-06,
+      "loss": 0.375,
+      "step": 985
+    },
+    {
+      "epoch": 1.1086125477850235,
+      "grad_norm": 2.6742877653053863,
+      "learning_rate": 7.944218745911367e-06,
+      "loss": 0.2824,
+      "step": 986
+    },
+    {
+      "epoch": 1.109736901281763,
+      "grad_norm": 2.9870370374869593,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.336,
+      "step": 987
+    },
+    {
+      "epoch": 1.1108612547785024,
+      "grad_norm": 3.2042419486351226,
+      "learning_rate": 7.933628741243326e-06,
+      "loss": 0.293,
+      "step": 988
+    },
+    {
+      "epoch": 1.1119856082752417,
+      "grad_norm": 2.9697170409591886,
+      "learning_rate": 7.928326194331404e-06,
+      "loss": 0.2875,
+      "step": 989
+    },
+    {
+      "epoch": 1.113109961771981,
+      "grad_norm": 2.7587778845607978,
+      "learning_rate": 7.923018629812369e-06,
+      "loss": 0.3428,
+      "step": 990
+    },
+    {
+      "epoch": 1.1142343152687204,
+      "grad_norm": 2.6357221736921184,
+      "learning_rate": 7.917706056780588e-06,
+      "loss": 0.2729,
+      "step": 991
+    },
+    {
+      "epoch": 1.11535866876546,
+      "grad_norm": 3.765832474513344,
+      "learning_rate": 7.912388484339012e-06,
+      "loss": 0.2746,
+      "step": 992
+    },
+    {
+      "epoch": 1.1164830222621993,
+      "grad_norm": 2.8776992587156203,
+      "learning_rate": 7.907065921599153e-06,
+      "loss": 0.2829,
+      "step": 993
+    },
+    {
+      "epoch": 1.1176073757589386,
+      "grad_norm": 2.9935506487327177,
+      "learning_rate": 7.901738377681081e-06,
+      "loss": 0.2946,
+      "step": 994
+    },
+    {
+      "epoch": 1.118731729255678,
+      "grad_norm": 2.526213083952562,
+      "learning_rate": 7.896405861713393e-06,
+      "loss": 0.3102,
+      "step": 995
+    },
+    {
+      "epoch": 1.1198560827524173,
+      "grad_norm": 2.731217026919832,
+      "learning_rate": 7.891068382833216e-06,
+      "loss": 0.2933,
+      "step": 996
+    },
+    {
+      "epoch": 1.1209804362491567,
+      "grad_norm": 3.346272035914983,
+      "learning_rate": 7.88572595018617e-06,
+      "loss": 0.364,
+      "step": 997
+    },
+    {
+      "epoch": 1.122104789745896,
+      "grad_norm": 2.6514780255475716,
+      "learning_rate": 7.880378572926367e-06,
+      "loss": 0.2404,
+      "step": 998
+    },
+    {
+      "epoch": 1.1232291432426356,
+      "grad_norm": 2.8102158731473184,
+      "learning_rate": 7.875026260216395e-06,
+      "loss": 0.3255,
+      "step": 999
+    },
+    {
+      "epoch": 1.124353496739375,
+      "grad_norm": 2.831225374864997,
+      "learning_rate": 7.869669021227293e-06,
+      "loss": 0.2729,
+      "step": 1000
+    },
+    {
+      "epoch": 1.124353496739375,
+      "eval_loss": 0.2562812864780426,
+      "eval_runtime": 358.6637,
+      "eval_samples_per_second": 148.772,
+      "eval_steps_per_second": 4.651,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1254778502361142,
+      "grad_norm": 3.1751928413794115,
+      "learning_rate": 7.864306865138546e-06,
+      "loss": 0.3635,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1266022037328536,
+      "grad_norm": 2.445218236064315,
+      "learning_rate": 7.858939801138061e-06,
+      "loss": 0.287,
+      "step": 1002
+    },
+    {
+      "epoch": 1.127726557229593,
+      "grad_norm": 2.4130535575245102,
+      "learning_rate": 7.85356783842216e-06,
+      "loss": 0.2268,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1288509107263325,
+      "grad_norm": 2.6477234133836887,
+      "learning_rate": 7.84819098619555e-06,
+      "loss": 0.2528,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1299752642230718,
+      "grad_norm": 4.219596318462011,
+      "learning_rate": 7.842809253671321e-06,
+      "loss": 0.3201,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1310996177198112,
+      "grad_norm": 2.478385718529922,
+      "learning_rate": 7.837422650070927e-06,
+      "loss": 0.2872,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1322239712165505,
+      "grad_norm": 2.8445082108470534,
+      "learning_rate": 7.832031184624165e-06,
+      "loss": 0.3065,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1333483247132898,
+      "grad_norm": 2.6058182348591576,
+      "learning_rate": 7.826634866569164e-06,
+      "loss": 0.3621,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1344726782100292,
+      "grad_norm": 2.885943817814243,
+      "learning_rate": 7.821233705152371e-06,
+      "loss": 0.3418,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1355970317067685,
+      "grad_norm": 2.776112519805635,
+      "learning_rate": 7.815827709628526e-06,
+      "loss": 0.2093,
+      "step": 1010
+    },
+    {
+      "epoch": 1.136721385203508,
+      "grad_norm": 3.2383389491687886,
+      "learning_rate": 7.810416889260653e-06,
+      "loss": 0.2891,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1378457387002474,
+      "grad_norm": 2.6054894900218093,
+      "learning_rate": 7.80500125332005e-06,
+      "loss": 0.2408,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1389700921969867,
+      "grad_norm": 3.095618740033099,
+      "learning_rate": 7.799580811086257e-06,
+      "loss": 0.3466,
+      "step": 1013
+    },
+    {
+      "epoch": 1.140094445693726,
+      "grad_norm": 2.9287778175334487,
+      "learning_rate": 7.794155571847058e-06,
+      "loss": 0.2904,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1412187991904654,
+      "grad_norm": 2.6289552567758205,
+      "learning_rate": 7.788725544898452e-06,
+      "loss": 0.2868,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1423431526872048,
+      "grad_norm": 3.043173257710223,
+      "learning_rate": 7.783290739544639e-06,
+      "loss": 0.3477,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1434675061839443,
+      "grad_norm": 2.4280278817374303,
+      "learning_rate": 7.777851165098012e-06,
+      "loss": 0.2732,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1445918596806837,
+      "grad_norm": 2.6563477431016245,
+      "learning_rate": 7.772406830879133e-06,
+      "loss": 0.3164,
+      "step": 1018
+    },
+    {
+      "epoch": 1.145716213177423,
+      "grad_norm": 4.250128599281835,
+      "learning_rate": 7.76695774621672e-06,
+      "loss": 0.2315,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1468405666741623,
+      "grad_norm": 2.704696977984871,
+      "learning_rate": 7.761503920447636e-06,
+      "loss": 0.2401,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1479649201709017,
+      "grad_norm": 3.1938225831440583,
+      "learning_rate": 7.756045362916853e-06,
+      "loss": 0.2936,
+      "step": 1021
+    },
+    {
+      "epoch": 1.149089273667641,
+      "grad_norm": 2.684105197492418,
+      "learning_rate": 7.750582082977468e-06,
+      "loss": 0.2509,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1502136271643804,
+      "grad_norm": 2.9837601977881096,
+      "learning_rate": 7.74511408999066e-06,
+      "loss": 0.3195,
+      "step": 1023
+    },
+    {
+      "epoch": 1.15133798066112,
+      "grad_norm": 2.432175267610356,
+      "learning_rate": 7.739641393325685e-06,
+      "loss": 0.295,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1524623341578593,
+      "grad_norm": 2.769093064287888,
+      "learning_rate": 7.734164002359864e-06,
+      "loss": 0.2776,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1535866876545986,
+      "grad_norm": 2.845714122934292,
+      "learning_rate": 7.72868192647855e-06,
+      "loss": 0.3235,
+      "step": 1026
+    },
+    {
+      "epoch": 1.154711041151338,
+      "grad_norm": 2.7823239983100736,
+      "learning_rate": 7.723195175075136e-06,
+      "loss": 0.2984,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1558353946480773,
+      "grad_norm": 2.8108853456923777,
+      "learning_rate": 7.71770375755102e-06,
+      "loss": 0.3068,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1569597481448168,
+      "grad_norm": 3.0135041944307135,
+      "learning_rate": 7.712207683315595e-06,
+      "loss": 0.3323,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1580841016415562,
+      "grad_norm": 3.422332185417169,
+      "learning_rate": 7.706706961786235e-06,
+      "loss": 0.2831,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1592084551382955,
+      "grad_norm": 2.5693851756198964,
+      "learning_rate": 7.701201602388276e-06,
+      "loss": 0.2737,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1603328086350349,
+      "grad_norm": 2.9976595023721773,
+      "learning_rate": 7.695691614555002e-06,
+      "loss": 0.3361,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1614571621317742,
+      "grad_norm": 3.5009927044362734,
+      "learning_rate": 7.690177007727626e-06,
+      "loss": 0.3077,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1625815156285135,
+      "grad_norm": 2.929049504193217,
+      "learning_rate": 7.68465779135528e-06,
+      "loss": 0.296,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1637058691252529,
+      "grad_norm": 2.5956607094216193,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 0.2741,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1648302226219924,
+      "grad_norm": 2.9001088925865894,
+      "learning_rate": 7.673605567811652e-06,
+      "loss": 0.3551,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1659545761187318,
+      "grad_norm": 2.729775132599058,
+      "learning_rate": 7.66807257957806e-06,
+      "loss": 0.2836,
+      "step": 1037
+    },
+    {
+      "epoch": 1.167078929615471,
+      "grad_norm": 2.8082393539897788,
+      "learning_rate": 7.662535019674828e-06,
+      "loss": 0.3329,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1682032831122104,
+      "grad_norm": 2.641522749610266,
+      "learning_rate": 7.656992897590416e-06,
+      "loss": 0.2894,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1693276366089498,
+      "grad_norm": 2.812579687617896,
+      "learning_rate": 7.651446222821099e-06,
+      "loss": 0.3009,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1704519901056893,
+      "grad_norm": 2.9029980812057445,
+      "learning_rate": 7.645895004870953e-06,
+      "loss": 0.302,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1715763436024287,
+      "grad_norm": 4.660174402018597,
+      "learning_rate": 7.64033925325184e-06,
+      "loss": 0.3289,
+      "step": 1042
+    },
+    {
+      "epoch": 1.172700697099168,
+      "grad_norm": 3.090540724121755,
+      "learning_rate": 7.634778977483389e-06,
+      "loss": 0.288,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1738250505959074,
+      "grad_norm": 3.159891019603476,
+      "learning_rate": 7.6292141870929784e-06,
+      "loss": 0.3357,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1749494040926467,
+      "grad_norm": 2.761035238361628,
+      "learning_rate": 7.6236448916157265e-06,
+      "loss": 0.293,
+      "step": 1045
+    },
+    {
+      "epoch": 1.176073757589386,
+      "grad_norm": 3.1591546821572716,
+      "learning_rate": 7.618071100594469e-06,
+      "loss": 0.2343,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1771981110861254,
+      "grad_norm": 2.781164401464889,
+      "learning_rate": 7.612492823579744e-06,
+      "loss": 0.3791,
+      "step": 1047
+    },
+    {
+      "epoch": 1.178322464582865,
+      "grad_norm": 2.3823948108311175,
+      "learning_rate": 7.60691007012978e-06,
+      "loss": 0.2656,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1794468180796043,
+      "grad_norm": 2.627214007380113,
+      "learning_rate": 7.6013228498104705e-06,
+      "loss": 0.2787,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1805711715763436,
+      "grad_norm": 2.315399013928376,
+      "learning_rate": 7.5957311721953656e-06,
+      "loss": 0.2718,
+      "step": 1050
+    },
+    {
+      "epoch": 1.181695525073083,
+      "grad_norm": 2.668688729845943,
+      "learning_rate": 7.590135046865652e-06,
+      "loss": 0.2856,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1828198785698223,
+      "grad_norm": 2.5817022383154518,
+      "learning_rate": 7.584534483410137e-06,
+      "loss": 0.3109,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1839442320665616,
+      "grad_norm": 2.3851881453383474,
+      "learning_rate": 7.5789294914252376e-06,
+      "loss": 0.2893,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1850685855633012,
+      "grad_norm": 2.943868999213503,
+      "learning_rate": 7.573320080514951e-06,
+      "loss": 0.3341,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1861929390600405,
+      "grad_norm": 2.683253539657934,
+      "learning_rate": 7.5677062602908515e-06,
+      "loss": 0.3072,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1873172925567799,
+      "grad_norm": 2.4173042132960334,
+      "learning_rate": 7.562088040372067e-06,
+      "loss": 0.2578,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1884416460535192,
+      "grad_norm": 2.6312722380030666,
+      "learning_rate": 7.55646543038526e-06,
+      "loss": 0.269,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1895659995502585,
+      "grad_norm": 2.511276050584004,
+      "learning_rate": 7.550838439964627e-06,
+      "loss": 0.2257,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1906903530469979,
+      "grad_norm": 3.612602574225538,
+      "learning_rate": 7.545207078751858e-06,
+      "loss": 0.2535,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1918147065437374,
+      "grad_norm": 3.7777879054509507,
+      "learning_rate": 7.539571356396136e-06,
+      "loss": 0.3751,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1929390600404768,
+      "grad_norm": 2.993781150409025,
+      "learning_rate": 7.533931282554121e-06,
+      "loss": 0.2844,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1940634135372161,
+      "grad_norm": 3.52989850145406,
+      "learning_rate": 7.528286866889924e-06,
+      "loss": 0.3154,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1951877670339555,
+      "grad_norm": 2.9460195878866533,
+      "learning_rate": 7.522638119075097e-06,
+      "loss": 0.2633,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1963121205306948,
+      "grad_norm": 2.960064344694475,
+      "learning_rate": 7.516985048788617e-06,
+      "loss": 0.3141,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1974364740274341,
+      "grad_norm": 2.9638040230321776,
+      "learning_rate": 7.511327665716863e-06,
+      "loss": 0.3087,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1985608275241737,
+      "grad_norm": 3.5387791973020466,
+      "learning_rate": 7.505665979553609e-06,
+      "loss": 0.346,
+      "step": 1066
+    },
+    {
+      "epoch": 1.199685181020913,
+      "grad_norm": 2.8334304008517024,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3561,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2008095345176524,
+      "grad_norm": 2.7970707111161026,
+      "learning_rate": 7.494329736764538e-06,
+      "loss": 0.2808,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2019338880143917,
+      "grad_norm": 3.054187444653746,
+      "learning_rate": 7.4886551995630615e-06,
+      "loss": 0.3518,
+      "step": 1069
+    },
+    {
+      "epoch": 1.203058241511131,
+      "grad_norm": 2.8894355809432684,
+      "learning_rate": 7.482976398118737e-06,
+      "loss": 0.2856,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2041825950078704,
+      "grad_norm": 2.905959128134695,
+      "learning_rate": 7.477293342162038e-06,
+      "loss": 0.3161,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2053069485046097,
+      "grad_norm": 3.2816850364734385,
+      "learning_rate": 7.471606041430724e-06,
+      "loss": 0.3119,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2064313020013493,
+      "grad_norm": 2.84026420974853,
+      "learning_rate": 7.465914505669829e-06,
+      "loss": 0.2835,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2075556554980886,
+      "grad_norm": 2.902917383149602,
+      "learning_rate": 7.4602187446316456e-06,
+      "loss": 0.3238,
+      "step": 1074
+    },
+    {
+      "epoch": 1.208680008994828,
+      "grad_norm": 2.7670132838147206,
+      "learning_rate": 7.454518768075705e-06,
+      "loss": 0.2669,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2098043624915673,
+      "grad_norm": 2.445128234610265,
+      "learning_rate": 7.448814585768761e-06,
+      "loss": 0.2425,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2109287159883066,
+      "grad_norm": 2.7478427853721676,
+      "learning_rate": 7.443106207484776e-06,
+      "loss": 0.2947,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2120530694850462,
+      "grad_norm": 2.8176446022128605,
+      "learning_rate": 7.4373936430048985e-06,
+      "loss": 0.2534,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2131774229817855,
+      "grad_norm": 2.6217208071545994,
+      "learning_rate": 7.431676902117453e-06,
+      "loss": 0.2828,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2143017764785249,
+      "grad_norm": 2.8620784354863646,
+      "learning_rate": 7.425955994617919e-06,
+      "loss": 0.2838,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2154261299752642,
+      "grad_norm": 2.622240490842645,
+      "learning_rate": 7.420230930308917e-06,
+      "loss": 0.2634,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2165504834720036,
+      "grad_norm": 3.100033658500169,
+      "learning_rate": 7.414501719000187e-06,
+      "loss": 0.3586,
+      "step": 1082
+    },
+    {
+      "epoch": 1.217674836968743,
+      "grad_norm": 3.1150069672719747,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 0.3154,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2187991904654822,
+      "grad_norm": 2.752085697295658,
+      "learning_rate": 7.403030894658023e-06,
+      "loss": 0.2863,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2199235439622218,
+      "grad_norm": 2.6095279298959797,
+      "learning_rate": 7.397289301279533e-06,
+      "loss": 0.2927,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2210478974589611,
+      "grad_norm": 2.423849305063145,
+      "learning_rate": 7.391543600211173e-06,
+      "loss": 0.2744,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2221722509557005,
+      "grad_norm": 2.928314102813541,
+      "learning_rate": 7.3857938012980425e-06,
+      "loss": 0.2877,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2232966044524398,
+      "grad_norm": 2.715141165727771,
+      "learning_rate": 7.380039914392267e-06,
+      "loss": 0.281,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2244209579491792,
+      "grad_norm": 2.9428484395491346,
+      "learning_rate": 7.3742819493529725e-06,
+      "loss": 0.3227,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2255453114459187,
+      "grad_norm": 2.8506282044977387,
+      "learning_rate": 7.36851991604628e-06,
+      "loss": 0.2931,
+      "step": 1090
+    },
+    {
+      "epoch": 1.226669664942658,
+      "grad_norm": 3.617892986016753,
+      "learning_rate": 7.362753824345271e-06,
+      "loss": 0.306,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2277940184393974,
+      "grad_norm": 3.016881654583372,
+      "learning_rate": 7.3569836841299905e-06,
+      "loss": 0.3403,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2289183719361367,
+      "grad_norm": 2.6774754466605843,
+      "learning_rate": 7.351209505287412e-06,
+      "loss": 0.2857,
+      "step": 1093
+    },
+    {
+      "epoch": 1.230042725432876,
+      "grad_norm": 3.065389025199878,
+      "learning_rate": 7.345431297711437e-06,
+      "loss": 0.3539,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2311670789296154,
+      "grad_norm": 2.8743090403470295,
+      "learning_rate": 7.3396490713028674e-06,
+      "loss": 0.2364,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2322914324263547,
+      "grad_norm": 3.033763225564396,
+      "learning_rate": 7.333862835969389e-06,
+      "loss": 0.2828,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2334157859230943,
+      "grad_norm": 2.596336040088367,
+      "learning_rate": 7.328072601625558e-06,
+      "loss": 0.2436,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2345401394198336,
+      "grad_norm": 3.0388324455820075,
+      "learning_rate": 7.322278378192783e-06,
+      "loss": 0.3557,
+      "step": 1098
+    },
+    {
+      "epoch": 1.235664492916573,
+      "grad_norm": 2.9751783658508804,
+      "learning_rate": 7.31648017559931e-06,
+      "loss": 0.3394,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2367888464133123,
+      "grad_norm": 2.9526605706778604,
+      "learning_rate": 7.310678003780197e-06,
+      "loss": 0.281,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2379131999100517,
+      "grad_norm": 2.9181471986877052,
+      "learning_rate": 7.304871872677313e-06,
+      "loss": 0.3245,
+      "step": 1101
+    },
+    {
+      "epoch": 1.239037553406791,
+      "grad_norm": 2.571544148935548,
+      "learning_rate": 7.2990617922393e-06,
+      "loss": 0.3027,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2401619069035306,
+      "grad_norm": 2.3946620569618435,
+      "learning_rate": 7.293247772421577e-06,
+      "loss": 0.275,
+      "step": 1103
+    },
+    {
+      "epoch": 1.24128626040027,
+      "grad_norm": 2.439411972694527,
+      "learning_rate": 7.2874298231863025e-06,
+      "loss": 0.2486,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2424106138970092,
+      "grad_norm": 2.853920777858357,
+      "learning_rate": 7.28160795450238e-06,
+      "loss": 0.3067,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2435349673937486,
+      "grad_norm": 2.521956172369948,
+      "learning_rate": 7.27578217634542e-06,
+      "loss": 0.2391,
+      "step": 1106
+    },
+    {
+      "epoch": 1.244659320890488,
+      "grad_norm": 2.9052958470318035,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.3192,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2457836743872273,
+      "grad_norm": 2.561047725817857,
+      "learning_rate": 7.264118931548318e-06,
+      "loss": 0.2769,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2469080278839666,
+      "grad_norm": 2.8819870607054523,
+      "learning_rate": 7.258281484892828e-06,
+      "loss": 0.3194,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2480323813807062,
+      "grad_norm": 2.5289304188881148,
+      "learning_rate": 7.252440168733572e-06,
+      "loss": 0.3052,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2491567348774455,
+      "grad_norm": 2.63806102009759,
+      "learning_rate": 7.246594993079483e-06,
+      "loss": 0.2418,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2502810883741848,
+      "grad_norm": 2.8308730849675596,
+      "learning_rate": 7.240745967946113e-06,
+      "loss": 0.3236,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2514054418709242,
+      "grad_norm": 3.4641227572964985,
+      "learning_rate": 7.2348931033556065e-06,
+      "loss": 0.3179,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2525297953676635,
+      "grad_norm": 2.6215191957181876,
+      "learning_rate": 7.229036409336687e-06,
+      "loss": 0.288,
+      "step": 1114
+    },
+    {
+      "epoch": 1.253654148864403,
+      "grad_norm": 2.5964112847510696,
+      "learning_rate": 7.223175895924638e-06,
+      "loss": 0.274,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2547785023611424,
+      "grad_norm": 3.168966997723176,
+      "learning_rate": 7.217311573161293e-06,
+      "loss": 0.2803,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2559028558578818,
+      "grad_norm": 2.697131661469735,
+      "learning_rate": 7.211443451095007e-06,
+      "loss": 0.2727,
+      "step": 1117
+    },
+    {
+      "epoch": 1.257027209354621,
+      "grad_norm": 2.83056211327441,
+      "learning_rate": 7.205571539780648e-06,
+      "loss": 0.3393,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2581515628513604,
+      "grad_norm": 2.32188979897206,
+      "learning_rate": 7.199695849279576e-06,
+      "loss": 0.254,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2592759163480998,
+      "grad_norm": 2.2483160042286663,
+      "learning_rate": 7.193816389659627e-06,
+      "loss": 0.249,
+      "step": 1120
+    },
+    {
+      "epoch": 1.260400269844839,
+      "grad_norm": 3.097120152368278,
+      "learning_rate": 7.1879331709950935e-06,
+      "loss": 0.2768,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2615246233415787,
+      "grad_norm": 2.7986510776611104,
+      "learning_rate": 7.18204620336671e-06,
+      "loss": 0.2585,
+      "step": 1122
+    },
+    {
+      "epoch": 1.262648976838318,
+      "grad_norm": 2.5754735178618913,
+      "learning_rate": 7.176155496861639e-06,
+      "loss": 0.3118,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2637733303350573,
+      "grad_norm": 2.663626198816178,
+      "learning_rate": 7.17026106157344e-06,
+      "loss": 0.2992,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2648976838317967,
+      "grad_norm": 2.693282412472543,
+      "learning_rate": 7.164362907602072e-06,
+      "loss": 0.2865,
+      "step": 1125
+    },
+    {
+      "epoch": 1.266022037328536,
+      "grad_norm": 2.5391648506728464,
+      "learning_rate": 7.158461045053856e-06,
+      "loss": 0.2681,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2671463908252756,
+      "grad_norm": 2.240825361238931,
+      "learning_rate": 7.1525554840414765e-06,
+      "loss": 0.231,
+      "step": 1127
+    },
+    {
+      "epoch": 1.268270744322015,
+      "grad_norm": 2.558301729544823,
+      "learning_rate": 7.14664623468395e-06,
+      "loss": 0.2306,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2693950978187543,
+      "grad_norm": 2.7884287674771664,
+      "learning_rate": 7.140733307106615e-06,
+      "loss": 0.2911,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2705194513154936,
+      "grad_norm": 2.821793233501533,
+      "learning_rate": 7.134816711441112e-06,
+      "loss": 0.2815,
+      "step": 1130
+    },
+    {
+      "epoch": 1.271643804812233,
+      "grad_norm": 2.5593453607466303,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 0.2796,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2727681583089723,
+      "grad_norm": 2.7242793274037584,
+      "learning_rate": 7.1229725564035665e-06,
+      "loss": 0.2856,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2738925118057116,
+      "grad_norm": 2.8745856920414985,
+      "learning_rate": 7.117045017326162e-06,
+      "loss": 0.3198,
+      "step": 1133
+    },
+    {
+      "epoch": 1.275016865302451,
+      "grad_norm": 3.2341038937907003,
+      "learning_rate": 7.111113850749828e-06,
+      "loss": 0.3298,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2761412187991905,
+      "grad_norm": 2.7664007855252293,
+      "learning_rate": 7.105179066837456e-06,
+      "loss": 0.2664,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2772655722959299,
+      "grad_norm": 2.5389843103191096,
+      "learning_rate": 7.099240675758137e-06,
+      "loss": 0.2801,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2783899257926692,
+      "grad_norm": 2.6755965105207493,
+      "learning_rate": 7.093298687687141e-06,
+      "loss": 0.2915,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2795142792894085,
+      "grad_norm": 2.5359890910198875,
+      "learning_rate": 7.087353112805905e-06,
+      "loss": 0.288,
+      "step": 1138
+    },
+    {
+      "epoch": 1.280638632786148,
+      "grad_norm": 3.2954850527313875,
+      "learning_rate": 7.081403961302007e-06,
+      "loss": 0.3235,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2817629862828874,
+      "grad_norm": 2.813983154778963,
+      "learning_rate": 7.075451243369157e-06,
+      "loss": 0.3073,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2828873397796268,
+      "grad_norm": 3.469631923708284,
+      "learning_rate": 7.069494969207175e-06,
+      "loss": 0.2918,
+      "step": 1141
+    },
+    {
+      "epoch": 1.284011693276366,
+      "grad_norm": 3.133005690947805,
+      "learning_rate": 7.063535149021974e-06,
+      "loss": 0.3504,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2851360467731054,
+      "grad_norm": 2.5450963509817544,
+      "learning_rate": 7.057571793025545e-06,
+      "loss": 0.2285,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2862604002698448,
+      "grad_norm": 2.649068784059692,
+      "learning_rate": 7.051604911435933e-06,
+      "loss": 0.2639,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2873847537665841,
+      "grad_norm": 2.5033283973546525,
+      "learning_rate": 7.04563451447723e-06,
+      "loss": 0.2764,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2885091072633235,
+      "grad_norm": 2.9977198706009647,
+      "learning_rate": 7.0396606123795465e-06,
+      "loss": 0.3662,
+      "step": 1146
+    },
+    {
+      "epoch": 1.289633460760063,
+      "grad_norm": 3.036966449353263,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.3343,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2907578142568024,
+      "grad_norm": 2.8043853839500534,
+      "learning_rate": 7.0277023337177016e-06,
+      "loss": 0.3248,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2918821677535417,
+      "grad_norm": 2.8356440752738097,
+      "learning_rate": 7.021717977643726e-06,
+      "loss": 0.2928,
+      "step": 1149
+    },
+    {
+      "epoch": 1.293006521250281,
+      "grad_norm": 2.706734481115516,
+      "learning_rate": 7.015730157411103e-06,
+      "loss": 0.307,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2941308747470206,
+      "grad_norm": 2.8098408648781477,
+      "learning_rate": 7.009738883279802e-06,
+      "loss": 0.3472,
+      "step": 1151
+    },
+    {
+      "epoch": 1.29525522824376,
+      "grad_norm": 2.881231157926553,
+      "learning_rate": 7.0037441655157045e-06,
+      "loss": 0.3232,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2963795817404993,
+      "grad_norm": 2.621661342793719,
+      "learning_rate": 6.997746014390601e-06,
+      "loss": 0.3231,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2975039352372386,
+      "grad_norm": 2.94211074490897,
+      "learning_rate": 6.991744440182156e-06,
+      "loss": 0.3489,
+      "step": 1154
+    },
+    {
+      "epoch": 1.298628288733978,
+      "grad_norm": 2.8425703878041033,
+      "learning_rate": 6.985739453173903e-06,
+      "loss": 0.3055,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2997526422307173,
+      "grad_norm": 2.5310000663520404,
+      "learning_rate": 6.979731063655227e-06,
+      "loss": 0.2405,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3008769957274566,
+      "grad_norm": 2.69290729133707,
+      "learning_rate": 6.973719281921336e-06,
+      "loss": 0.281,
+      "step": 1157
+    },
+    {
+      "epoch": 1.302001349224196,
+      "grad_norm": 2.643656317191291,
+      "learning_rate": 6.967704118273257e-06,
+      "loss": 0.2786,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3031257027209355,
+      "grad_norm": 2.5146322782713555,
+      "learning_rate": 6.961685583017808e-06,
+      "loss": 0.2735,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3042500562176749,
+      "grad_norm": 2.7677400688480365,
+      "learning_rate": 6.955663686467585e-06,
+      "loss": 0.2969,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3053744097144142,
+      "grad_norm": 3.27052637290744,
+      "learning_rate": 6.949638438940942e-06,
+      "loss": 0.3432,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3064987632111535,
+      "grad_norm": 2.814723829619485,
+      "learning_rate": 6.943609850761979e-06,
+      "loss": 0.2977,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3076231167078929,
+      "grad_norm": 2.943483669810458,
+      "learning_rate": 6.9375779322605154e-06,
+      "loss": 0.3468,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3087474702046324,
+      "grad_norm": 2.364578666635158,
+      "learning_rate": 6.931542693772081e-06,
+      "loss": 0.2379,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3098718237013718,
+      "grad_norm": 2.4999103359129,
+      "learning_rate": 6.925504145637891e-06,
+      "loss": 0.3014,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3109961771981111,
+      "grad_norm": 2.1695566380010116,
+      "learning_rate": 6.919462298204834e-06,
+      "loss": 0.2174,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3121205306948505,
+      "grad_norm": 2.7987420630694,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.2996,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3132448841915898,
+      "grad_norm": 2.806386321879188,
+      "learning_rate": 6.907368746857914e-06,
+      "loss": 0.2804,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3143692376883291,
+      "grad_norm": 3.0657484411056184,
+      "learning_rate": 6.9013170636660255e-06,
+      "loss": 0.2948,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3154935911850685,
+      "grad_norm": 2.4250259691156586,
+      "learning_rate": 6.895262122619174e-06,
+      "loss": 0.2948,
+      "step": 1170
+    },
+    {
+      "epoch": 1.316617944681808,
+      "grad_norm": 2.992507610915035,
+      "learning_rate": 6.889203934092337e-06,
+      "loss": 0.3056,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3177422981785474,
+      "grad_norm": 2.6504725249430705,
+      "learning_rate": 6.883142508466054e-06,
+      "loss": 0.2662,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3188666516752867,
+      "grad_norm": 2.927921846439051,
+      "learning_rate": 6.877077856126416e-06,
+      "loss": 0.3025,
+      "step": 1173
+    },
+    {
+      "epoch": 1.319991005172026,
+      "grad_norm": 2.915194377779703,
+      "learning_rate": 6.871009987465037e-06,
+      "loss": 0.2936,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3211153586687654,
+      "grad_norm": 2.9824206618349622,
+      "learning_rate": 6.8649389128790455e-06,
+      "loss": 0.3139,
+      "step": 1175
+    },
+    {
+      "epoch": 1.322239712165505,
+      "grad_norm": 2.6332599852495995,
+      "learning_rate": 6.858864642771062e-06,
+      "loss": 0.3253,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3233640656622443,
+      "grad_norm": 3.064410476764093,
+      "learning_rate": 6.852787187549182e-06,
+      "loss": 0.3302,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3244884191589836,
+      "grad_norm": 2.5464093551575226,
+      "learning_rate": 6.846706557626961e-06,
+      "loss": 0.2705,
+      "step": 1178
+    },
+    {
+      "epoch": 1.325612772655723,
+      "grad_norm": 3.273650579051752,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 0.3582,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3267371261524623,
+      "grad_norm": 2.971170267605882,
+      "learning_rate": 6.834535815362888e-06,
+      "loss": 0.3285,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3278614796492016,
+      "grad_norm": 2.532793966080974,
+      "learning_rate": 6.828445723875272e-06,
+      "loss": 0.2404,
+      "step": 1181
+    },
+    {
+      "epoch": 1.328985833145941,
+      "grad_norm": 2.7375693010099376,
+      "learning_rate": 6.822352499395751e-06,
+      "loss": 0.2677,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3301101866426803,
+      "grad_norm": 2.689219905873509,
+      "learning_rate": 6.816256152364893e-06,
+      "loss": 0.2929,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3312345401394199,
+      "grad_norm": 3.221945659605454,
+      "learning_rate": 6.810156693228627e-06,
+      "loss": 0.3258,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3323588936361592,
+      "grad_norm": 2.6933136800732256,
+      "learning_rate": 6.804054132438209e-06,
+      "loss": 0.2379,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3334832471328986,
+      "grad_norm": 2.3225774704015265,
+      "learning_rate": 6.797948480450212e-06,
+      "loss": 0.2676,
+      "step": 1186
+    },
+    {
+      "epoch": 1.334607600629638,
+      "grad_norm": 2.4588978211096406,
+      "learning_rate": 6.7918397477265e-06,
+      "loss": 0.275,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3357319541263775,
+      "grad_norm": 3.1795072323120563,
+      "learning_rate": 6.785727944734228e-06,
+      "loss": 0.3584,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3368563076231168,
+      "grad_norm": 2.650973721648376,
+      "learning_rate": 6.779613081945795e-06,
+      "loss": 0.3556,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3379806611198561,
+      "grad_norm": 2.5745724180602276,
+      "learning_rate": 6.773495169838856e-06,
+      "loss": 0.2857,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3391050146165955,
+      "grad_norm": 2.917261350497406,
+      "learning_rate": 6.767374218896286e-06,
+      "loss": 0.3185,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3402293681133348,
+      "grad_norm": 2.8324035752473526,
+      "learning_rate": 6.7612502396061685e-06,
+      "loss": 0.3048,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3413537216100742,
+      "grad_norm": 2.824301835496274,
+      "learning_rate": 6.755123242461774e-06,
+      "loss": 0.376,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3424780751068135,
+      "grad_norm": 3.631483591401779,
+      "learning_rate": 6.748993237961544e-06,
+      "loss": 0.3421,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3436024286035528,
+      "grad_norm": 2.4843289301574014,
+      "learning_rate": 6.7428602366090764e-06,
+      "loss": 0.3024,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3447267821002924,
+      "grad_norm": 3.017086238421862,
+      "learning_rate": 6.736724248913102e-06,
+      "loss": 0.327,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3458511355970317,
+      "grad_norm": 2.664912564443586,
+      "learning_rate": 6.730585285387465e-06,
+      "loss": 0.3093,
+      "step": 1197
+    },
+    {
+      "epoch": 1.346975489093771,
+      "grad_norm": 2.896952830645061,
+      "learning_rate": 6.724443356551116e-06,
+      "loss": 0.3259,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3480998425905104,
+      "grad_norm": 2.8832897159866255,
+      "learning_rate": 6.718298472928082e-06,
+      "loss": 0.288,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3492241960872497,
+      "grad_norm": 2.5122036223092947,
+      "learning_rate": 6.71215064504745e-06,
+      "loss": 0.3036,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3503485495839893,
+      "grad_norm": 2.8939076676266016,
+      "learning_rate": 6.70599988344336e-06,
+      "loss": 0.3222,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3514729030807286,
+      "grad_norm": 2.748732728733348,
+      "learning_rate": 6.6998461986549715e-06,
+      "loss": 0.3041,
+      "step": 1202
+    },
+    {
+      "epoch": 1.352597256577468,
+      "grad_norm": 2.544159935433011,
+      "learning_rate": 6.693689601226458e-06,
+      "loss": 0.2693,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3537216100742073,
+      "grad_norm": 2.6880726042580894,
+      "learning_rate": 6.687530101706978e-06,
+      "loss": 0.3216,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3548459635709467,
+      "grad_norm": 2.8042989174896507,
+      "learning_rate": 6.6813677106506705e-06,
+      "loss": 0.2484,
+      "step": 1205
+    },
+    {
+      "epoch": 1.355970317067686,
+      "grad_norm": 2.451849699035043,
+      "learning_rate": 6.67520243861662e-06,
+      "loss": 0.2809,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3570946705644253,
+      "grad_norm": 2.2931669740924847,
+      "learning_rate": 6.669034296168855e-06,
+      "loss": 0.268,
+      "step": 1207
+    },
+    {
+      "epoch": 1.358219024061165,
+      "grad_norm": 2.854653791382171,
+      "learning_rate": 6.662863293876319e-06,
+      "loss": 0.3097,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3593433775579042,
+      "grad_norm": 2.6146309250480972,
+      "learning_rate": 6.656689442312855e-06,
+      "loss": 0.297,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3604677310546436,
+      "grad_norm": 2.8058480136228536,
+      "learning_rate": 6.6505127520571914e-06,
+      "loss": 0.2441,
+      "step": 1210
+    },
+    {
+      "epoch": 1.361592084551383,
+      "grad_norm": 2.6869543793657655,
+      "learning_rate": 6.644333233692917e-06,
+      "loss": 0.269,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3627164380481223,
+      "grad_norm": 2.4020636061068372,
+      "learning_rate": 6.638150897808469e-06,
+      "loss": 0.2536,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3638407915448618,
+      "grad_norm": 3.0280847399912494,
+      "learning_rate": 6.63196575499711e-06,
+      "loss": 0.2996,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3649651450416012,
+      "grad_norm": 2.687005988626177,
+      "learning_rate": 6.625777815856915e-06,
+      "loss": 0.3185,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3660894985383405,
+      "grad_norm": 2.723821993554223,
+      "learning_rate": 6.619587090990748e-06,
+      "loss": 0.2766,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3672138520350798,
+      "grad_norm": 2.788548240450778,
+      "learning_rate": 6.613393591006247e-06,
+      "loss": 0.2822,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3683382055318192,
+      "grad_norm": 2.9628781605091086,
+      "learning_rate": 6.607197326515808e-06,
+      "loss": 0.254,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3694625590285585,
+      "grad_norm": 3.3546273377543265,
+      "learning_rate": 6.600998308136559e-06,
+      "loss": 0.3113,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3705869125252979,
+      "grad_norm": 3.1004002606382604,
+      "learning_rate": 6.594796546490351e-06,
+      "loss": 0.3084,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3717112660220372,
+      "grad_norm": 3.051239377044502,
+      "learning_rate": 6.58859205220373e-06,
+      "loss": 0.2795,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3728356195187768,
+      "grad_norm": 2.5674864705213265,
+      "learning_rate": 6.582384835907931e-06,
+      "loss": 0.3075,
+      "step": 1221
+    },
+    {
+      "epoch": 1.373959973015516,
+      "grad_norm": 2.6648881882592836,
+      "learning_rate": 6.57617490823885e-06,
+      "loss": 0.3027,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3750843265122554,
+      "grad_norm": 2.729606456583687,
+      "learning_rate": 6.569962279837025e-06,
+      "loss": 0.3051,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3762086800089948,
+      "grad_norm": 2.9081353157253265,
+      "learning_rate": 6.56374696134763e-06,
+      "loss": 0.2872,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3773330335057343,
+      "grad_norm": 2.816961863834089,
+      "learning_rate": 6.557528963420442e-06,
+      "loss": 0.3275,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3784573870024737,
+      "grad_norm": 2.938357492633034,
+      "learning_rate": 6.55130829670983e-06,
+      "loss": 0.341,
+      "step": 1226
+    },
+    {
+      "epoch": 1.379581740499213,
+      "grad_norm": 2.954779695996704,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2902,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3807060939959523,
+      "grad_norm": 2.5683036465443645,
+      "learning_rate": 6.5388589995786635e-06,
+      "loss": 0.2779,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3818304474926917,
+      "grad_norm": 2.7574145005815405,
+      "learning_rate": 6.5326303904896395e-06,
+      "loss": 0.265,
+      "step": 1229
+    },
+    {
+      "epoch": 1.382954800989431,
+      "grad_norm": 3.0572590107654927,
+      "learning_rate": 6.526399155280218e-06,
+      "loss": 0.2786,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3840791544861704,
+      "grad_norm": 2.3426765947540473,
+      "learning_rate": 6.520165304627452e-06,
+      "loss": 0.2394,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3852035079829097,
+      "grad_norm": 2.3070639111197124,
+      "learning_rate": 6.513928849212874e-06,
+      "loss": 0.2644,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3863278614796493,
+      "grad_norm": 2.641701218592941,
+      "learning_rate": 6.507689799722479e-06,
+      "loss": 0.2279,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3874522149763886,
+      "grad_norm": 2.702301737029662,
+      "learning_rate": 6.501448166846711e-06,
+      "loss": 0.2535,
+      "step": 1234
+    },
+    {
+      "epoch": 1.388576568473128,
+      "grad_norm": 3.012167779871616,
+      "learning_rate": 6.495203961280434e-06,
+      "loss": 0.3448,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3897009219698673,
+      "grad_norm": 3.0716208153673645,
+      "learning_rate": 6.4889571937229275e-06,
+      "loss": 0.2894,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3908252754666068,
+      "grad_norm": 2.13140782498267,
+      "learning_rate": 6.482707874877855e-06,
+      "loss": 0.211,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3919496289633462,
+      "grad_norm": 2.7643432292751955,
+      "learning_rate": 6.476456015453257e-06,
+      "loss": 0.3193,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3930739824600855,
+      "grad_norm": 2.482841913844934,
+      "learning_rate": 6.47020162616152e-06,
+      "loss": 0.2631,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3941983359568249,
+      "grad_norm": 3.2518187873514055,
+      "learning_rate": 6.4639447177193745e-06,
+      "loss": 0.3584,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3953226894535642,
+      "grad_norm": 2.761306650753673,
+      "learning_rate": 6.457685300847858e-06,
+      "loss": 0.3225,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3964470429503035,
+      "grad_norm": 2.5713080005995743,
+      "learning_rate": 6.451423386272312e-06,
+      "loss": 0.2563,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3975713964470429,
+      "grad_norm": 2.6636202432951346,
+      "learning_rate": 6.445158984722358e-06,
+      "loss": 0.252,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3986957499437822,
+      "grad_norm": 3.1010056377285795,
+      "learning_rate": 6.438892106931876e-06,
+      "loss": 0.3156,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3998201034405218,
+      "grad_norm": 2.720467041807045,
+      "learning_rate": 6.432622763638993e-06,
+      "loss": 0.2584,
+      "step": 1245
+    },
+    {
+      "epoch": 1.400944456937261,
+      "grad_norm": 2.359371604344945,
+      "learning_rate": 6.426350965586053e-06,
+      "loss": 0.2182,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4020688104340004,
+      "grad_norm": 2.529814459133507,
+      "learning_rate": 6.420076723519615e-06,
+      "loss": 0.2441,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4031931639307398,
+      "grad_norm": 2.331333023795072,
+      "learning_rate": 6.413800048190417e-06,
+      "loss": 0.2324,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4043175174274791,
+      "grad_norm": 2.85706278111809,
+      "learning_rate": 6.407520950353377e-06,
+      "loss": 0.3124,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4054418709242187,
+      "grad_norm": 2.6996908849976022,
+      "learning_rate": 6.401239440767555e-06,
+      "loss": 0.3082,
+      "step": 1250
+    },
+    {
+      "epoch": 1.406566224420958,
+      "grad_norm": 3.0119468091336157,
+      "learning_rate": 6.3949555301961474e-06,
+      "loss": 0.2821,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4076905779176974,
+      "grad_norm": 3.4311307556221964,
+      "learning_rate": 6.388669229406462e-06,
+      "loss": 0.3882,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4088149314144367,
+      "grad_norm": 2.645481306356194,
+      "learning_rate": 6.382380549169905e-06,
+      "loss": 0.3381,
+      "step": 1253
+    },
+    {
+      "epoch": 1.409939284911176,
+      "grad_norm": 2.859479025437835,
+      "learning_rate": 6.376089500261958e-06,
+      "loss": 0.3689,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4110636384079154,
+      "grad_norm": 2.968103897456334,
+      "learning_rate": 6.369796093462164e-06,
+      "loss": 0.3165,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4121879919046547,
+      "grad_norm": 2.8503052546448315,
+      "learning_rate": 6.363500339554099e-06,
+      "loss": 0.3098,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4133123454013943,
+      "grad_norm": 2.6879311644483357,
+      "learning_rate": 6.3572022493253715e-06,
+      "loss": 0.28,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4144366988981336,
+      "grad_norm": 4.247080664784111,
+      "learning_rate": 6.350901833567586e-06,
+      "loss": 0.2921,
+      "step": 1258
+    },
+    {
+      "epoch": 1.415561052394873,
+      "grad_norm": 2.5560777045436645,
+      "learning_rate": 6.344599103076329e-06,
+      "loss": 0.2821,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4166854058916123,
+      "grad_norm": 2.707521923537777,
+      "learning_rate": 6.3382940686511625e-06,
+      "loss": 0.3197,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4178097593883516,
+      "grad_norm": 2.617457425082368,
+      "learning_rate": 6.331986741095588e-06,
+      "loss": 0.2648,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4189341128850912,
+      "grad_norm": 2.7332090635371777,
+      "learning_rate": 6.325677131217041e-06,
+      "loss": 0.2433,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4200584663818305,
+      "grad_norm": 3.1774116558893466,
+      "learning_rate": 6.3193652498268656e-06,
+      "loss": 0.3473,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4211828198785699,
+      "grad_norm": 2.7747560803243405,
+      "learning_rate": 6.313051107740298e-06,
+      "loss": 0.263,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4223071733753092,
+      "grad_norm": 2.5074607419684254,
+      "learning_rate": 6.306734715776448e-06,
+      "loss": 0.304,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4234315268720485,
+      "grad_norm": 2.724534482108307,
+      "learning_rate": 6.300416084758284e-06,
+      "loss": 0.3094,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4245558803687879,
+      "grad_norm": 2.748385612499274,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3062,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4256802338655272,
+      "grad_norm": 2.8673032362892124,
+      "learning_rate": 6.287772148870032e-06,
+      "loss": 0.2907,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4268045873622666,
+      "grad_norm": 2.760436267097503,
+      "learning_rate": 6.281446865664984e-06,
+      "loss": 0.3078,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4279289408590061,
+      "grad_norm": 2.7930469409530834,
+      "learning_rate": 6.275119386735662e-06,
+      "loss": 0.2954,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4290532943557455,
+      "grad_norm": 3.139858867318902,
+      "learning_rate": 6.268789722924029e-06,
+      "loss": 0.3409,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4301776478524848,
+      "grad_norm": 2.669954477779559,
+      "learning_rate": 6.26245788507579e-06,
+      "loss": 0.3124,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4313020013492241,
+      "grad_norm": 2.7328073466271565,
+      "learning_rate": 6.256123884040378e-06,
+      "loss": 0.3824,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4324263548459637,
+      "grad_norm": 4.358596896116515,
+      "learning_rate": 6.249787730670932e-06,
+      "loss": 0.4066,
+      "step": 1274
+    },
+    {
+      "epoch": 1.433550708342703,
+      "grad_norm": 2.8530944540844314,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 0.2845,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4346750618394424,
+      "grad_norm": 3.543131346082617,
+      "learning_rate": 6.237109010360904e-06,
+      "loss": 0.2823,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4357994153361817,
+      "grad_norm": 2.553904624787297,
+      "learning_rate": 6.230766465144966e-06,
+      "loss": 0.2537,
+      "step": 1277
+    },
+    {
+      "epoch": 1.436923768832921,
+      "grad_norm": 2.6017823559876008,
+      "learning_rate": 6.224421811044238e-06,
+      "loss": 0.2941,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4380481223296604,
+      "grad_norm": 2.8824799984529967,
+      "learning_rate": 6.218075058930113e-06,
+      "loss": 0.2949,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4391724758263997,
+      "grad_norm": 2.479541590536001,
+      "learning_rate": 6.211726219677579e-06,
+      "loss": 0.2652,
+      "step": 1280
+    },
+    {
+      "epoch": 1.440296829323139,
+      "grad_norm": 2.8143048267025814,
+      "learning_rate": 6.205375304165194e-06,
+      "loss": 0.3572,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4414211828198786,
+      "grad_norm": 3.06294101517481,
+      "learning_rate": 6.199022323275083e-06,
+      "loss": 0.3322,
+      "step": 1282
+    },
+    {
+      "epoch": 1.442545536316618,
+      "grad_norm": 2.325740462751447,
+      "learning_rate": 6.192667287892905e-06,
+      "loss": 0.3013,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4436698898133573,
+      "grad_norm": 2.4312622709006555,
+      "learning_rate": 6.18631020890784e-06,
+      "loss": 0.2067,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4447942433100966,
+      "grad_norm": 2.457422615631194,
+      "learning_rate": 6.179951097212566e-06,
+      "loss": 0.2604,
+      "step": 1285
+    },
+    {
+      "epoch": 1.445918596806836,
+      "grad_norm": 2.4757939607121746,
+      "learning_rate": 6.173589963703255e-06,
+      "loss": 0.2766,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4470429503035755,
+      "grad_norm": 2.860212641410032,
+      "learning_rate": 6.1672268192795285e-06,
+      "loss": 0.2876,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4481673038003149,
+      "grad_norm": 2.592867004418506,
+      "learning_rate": 6.160861674844464e-06,
+      "loss": 0.313,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4492916572970542,
+      "grad_norm": 2.664951670209397,
+      "learning_rate": 6.154494541304562e-06,
+      "loss": 0.2663,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4504160107937936,
+      "grad_norm": 2.922693776465058,
+      "learning_rate": 6.148125429569735e-06,
+      "loss": 0.2829,
+      "step": 1290
+    },
+    {
+      "epoch": 1.451540364290533,
+      "grad_norm": 2.88550277227972,
+      "learning_rate": 6.141754350553279e-06,
+      "loss": 0.3209,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4526647177872722,
+      "grad_norm": 2.6982660670426446,
+      "learning_rate": 6.135381315171867e-06,
+      "loss": 0.2618,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4537890712840116,
+      "grad_norm": 2.570657480457353,
+      "learning_rate": 6.1290063343455196e-06,
+      "loss": 0.3187,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4549134247807511,
+      "grad_norm": 2.767106742375816,
+      "learning_rate": 6.1226294189975925e-06,
+      "loss": 0.2625,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4560377782774905,
+      "grad_norm": 2.601002757931117,
+      "learning_rate": 6.116250580054758e-06,
+      "loss": 0.2682,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4571621317742298,
+      "grad_norm": 2.8009960090914605,
+      "learning_rate": 6.10986982844698e-06,
+      "loss": 0.254,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4582864852709692,
+      "grad_norm": 3.1797149648538183,
+      "learning_rate": 6.103487175107508e-06,
+      "loss": 0.2897,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4594108387677085,
+      "grad_norm": 2.6484924022594574,
+      "learning_rate": 6.097102630972837e-06,
+      "loss": 0.2467,
+      "step": 1298
+    },
+    {
+      "epoch": 1.460535192264448,
+      "grad_norm": 2.714295991335401,
+      "learning_rate": 6.090716206982714e-06,
+      "loss": 0.3044,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4616595457611874,
+      "grad_norm": 2.9301263628917193,
+      "learning_rate": 6.084327914080101e-06,
+      "loss": 0.3289,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4627838992579267,
+      "grad_norm": 2.4351636450448533,
+      "learning_rate": 6.077937763211166e-06,
+      "loss": 0.2797,
+      "step": 1301
+    },
+    {
+      "epoch": 1.463908252754666,
+      "grad_norm": 2.687277723198637,
+      "learning_rate": 6.071545765325254e-06,
+      "loss": 0.2773,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4650326062514054,
+      "grad_norm": 3.1472646339738364,
+      "learning_rate": 6.065151931374884e-06,
+      "loss": 0.2846,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4661569597481448,
+      "grad_norm": 2.539936540151918,
+      "learning_rate": 6.058756272315713e-06,
+      "loss": 0.2891,
+      "step": 1304
+    },
+    {
+      "epoch": 1.467281313244884,
+      "grad_norm": 2.7320600557468238,
+      "learning_rate": 6.052358799106528e-06,
+      "loss": 0.2619,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4684056667416236,
+      "grad_norm": 2.5751029364102207,
+      "learning_rate": 6.045959522709228e-06,
+      "loss": 0.2557,
+      "step": 1306
+    },
+    {
+      "epoch": 1.469530020238363,
+      "grad_norm": 2.632159177917145,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.2795,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4706543737351023,
+      "grad_norm": 2.656197156020335,
+      "learning_rate": 6.033155604213291e-06,
+      "loss": 0.2931,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4717787272318417,
+      "grad_norm": 2.8311824169997974,
+      "learning_rate": 6.026750984053821e-06,
+      "loss": 0.3246,
+      "step": 1309
+    },
+    {
+      "epoch": 1.472903080728581,
+      "grad_norm": 2.463841175256915,
+      "learning_rate": 6.020344604584526e-06,
+      "loss": 0.2522,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4740274342253206,
+      "grad_norm": 2.8690736584864704,
+      "learning_rate": 6.013936476782563e-06,
+      "loss": 0.3071,
+      "step": 1311
+    },
+    {
+      "epoch": 1.47515178772206,
+      "grad_norm": 2.5621424092200935,
+      "learning_rate": 6.0075266116280865e-06,
+      "loss": 0.251,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4762761412187992,
+      "grad_norm": 2.4648039653457943,
+      "learning_rate": 6.001115020104223e-06,
+      "loss": 0.2492,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4774004947155386,
+      "grad_norm": 2.513208858299552,
+      "learning_rate": 5.994701713197063e-06,
+      "loss": 0.2839,
+      "step": 1314
+    },
+    {
+      "epoch": 1.478524848212278,
+      "grad_norm": 2.3532304787350538,
+      "learning_rate": 5.988286701895631e-06,
+      "loss": 0.2235,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4796492017090173,
+      "grad_norm": 2.4710901206569207,
+      "learning_rate": 5.981869997191876e-06,
+      "loss": 0.2698,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4807735552057566,
+      "grad_norm": 2.8689567396066993,
+      "learning_rate": 5.975451610080643e-06,
+      "loss": 0.3589,
+      "step": 1317
+    },
+    {
+      "epoch": 1.481897908702496,
+      "grad_norm": 3.1682870012730033,
+      "learning_rate": 5.969031551559664e-06,
+      "loss": 0.2961,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4830222621992355,
+      "grad_norm": 2.7614967291466943,
+      "learning_rate": 5.962609832629538e-06,
+      "loss": 0.2812,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4841466156959748,
+      "grad_norm": 2.685710123366342,
+      "learning_rate": 5.956186464293703e-06,
+      "loss": 0.2774,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4852709691927142,
+      "grad_norm": 3.032936018630151,
+      "learning_rate": 5.949761457558424e-06,
+      "loss": 0.3174,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4863953226894535,
+      "grad_norm": 2.7822815281368842,
+      "learning_rate": 5.943334823432777e-06,
+      "loss": 0.3273,
+      "step": 1322
+    },
+    {
+      "epoch": 1.487519676186193,
+      "grad_norm": 2.885389933182994,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 0.3383,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4886440296829324,
+      "grad_norm": 2.2461867874623245,
+      "learning_rate": 5.930476717060596e-06,
+      "loss": 0.2176,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4897683831796718,
+      "grad_norm": 2.7106193659680873,
+      "learning_rate": 5.9240452668460775e-06,
+      "loss": 0.3026,
+      "step": 1325
+    },
+    {
+      "epoch": 1.490892736676411,
+      "grad_norm": 2.9177764670265782,
+      "learning_rate": 5.917612233305183e-06,
+      "loss": 0.2991,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4920170901731504,
+      "grad_norm": 2.8405325535912307,
+      "learning_rate": 5.911177627460739e-06,
+      "loss": 0.2907,
+      "step": 1327
+    },
+    {
+      "epoch": 1.4931414436698898,
+      "grad_norm": 2.622108909708301,
+      "learning_rate": 5.904741460338266e-06,
+      "loss": 0.2503,
+      "step": 1328
+    },
+    {
+      "epoch": 1.494265797166629,
+      "grad_norm": 3.0487960594067904,
+      "learning_rate": 5.898303742965964e-06,
+      "loss": 0.3503,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4953901506633684,
+      "grad_norm": 2.539172534273024,
+      "learning_rate": 5.891864486374683e-06,
+      "loss": 0.2806,
+      "step": 1330
+    },
+    {
+      "epoch": 1.496514504160108,
+      "grad_norm": 2.5812952166880887,
+      "learning_rate": 5.885423701597918e-06,
+      "loss": 0.2599,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4976388576568473,
+      "grad_norm": 3.012995809163071,
+      "learning_rate": 5.878981399671774e-06,
+      "loss": 0.3664,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4987632111535867,
+      "grad_norm": 2.808724105215205,
+      "learning_rate": 5.87253759163496e-06,
+      "loss": 0.2889,
+      "step": 1333
+    },
+    {
+      "epoch": 1.499887564650326,
+      "grad_norm": 2.6482668079548697,
+      "learning_rate": 5.866092288528768e-06,
+      "loss": 0.2888,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5010119181470656,
+      "grad_norm": 2.512640391543218,
+      "learning_rate": 5.859645501397048e-06,
+      "loss": 0.2211,
+      "step": 1335
+    },
+    {
+      "epoch": 1.502136271643805,
+      "grad_norm": 2.562764877198543,
+      "learning_rate": 5.853197241286194e-06,
+      "loss": 0.2698,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5032606251405443,
+      "grad_norm": 2.7331542964038107,
+      "learning_rate": 5.846747519245123e-06,
+      "loss": 0.2617,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5043849786372836,
+      "grad_norm": 3.414621690091624,
+      "learning_rate": 5.8402963463252605e-06,
+      "loss": 0.2716,
+      "step": 1338
+    },
+    {
+      "epoch": 1.505509332134023,
+      "grad_norm": 2.8490678305265513,
+      "learning_rate": 5.8338437335805124e-06,
+      "loss": 0.2705,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5066336856307623,
+      "grad_norm": 2.7825525769685804,
+      "learning_rate": 5.827389692067256e-06,
+      "loss": 0.2912,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5077580391275016,
+      "grad_norm": 2.7145395683616567,
+      "learning_rate": 5.820934232844315e-06,
+      "loss": 0.2865,
+      "step": 1341
+    },
+    {
+      "epoch": 1.508882392624241,
+      "grad_norm": 2.7275739549362323,
+      "learning_rate": 5.814477366972945e-06,
+      "loss": 0.2606,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5100067461209803,
+      "grad_norm": 2.6313556264290185,
+      "learning_rate": 5.8080191055168064e-06,
+      "loss": 0.2453,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5111310996177199,
+      "grad_norm": 2.636517556125716,
+      "learning_rate": 5.801559459541956e-06,
+      "loss": 0.2748,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5122554531144592,
+      "grad_norm": 2.409408073592718,
+      "learning_rate": 5.795098440116822e-06,
+      "loss": 0.2661,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5133798066111985,
+      "grad_norm": 2.9881692201258745,
+      "learning_rate": 5.7886360583121825e-06,
+      "loss": 0.2627,
+      "step": 1346
+    },
+    {
+      "epoch": 1.514504160107938,
+      "grad_norm": 2.7949985744909505,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.2926,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5156285136046774,
+      "grad_norm": 2.5442486830707884,
+      "learning_rate": 5.775707251859168e-06,
+      "loss": 0.2554,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5167528671014168,
+      "grad_norm": 2.735196517319595,
+      "learning_rate": 5.7692408493639514e-06,
+      "loss": 0.3225,
+      "step": 1349
+    },
+    {
+      "epoch": 1.517877220598156,
+      "grad_norm": 2.610514298652375,
+      "learning_rate": 5.7627731287955054e-06,
+      "loss": 0.3057,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5190015740948954,
+      "grad_norm": 2.797648541159533,
+      "learning_rate": 5.7563041012360975e-06,
+      "loss": 0.3168,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5201259275916348,
+      "grad_norm": 2.679767002997822,
+      "learning_rate": 5.749833777770225e-06,
+      "loss": 0.2869,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5212502810883741,
+      "grad_norm": 2.7626728479008635,
+      "learning_rate": 5.743362169484617e-06,
+      "loss": 0.2828,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5223746345851135,
+      "grad_norm": 2.521672008608302,
+      "learning_rate": 5.7368892874681905e-06,
+      "loss": 0.3334,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5234989880818528,
+      "grad_norm": 2.6123196228953445,
+      "learning_rate": 5.730415142812059e-06,
+      "loss": 0.2484,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5246233415785924,
+      "grad_norm": 2.731266535233032,
+      "learning_rate": 5.72393974660949e-06,
+      "loss": 0.2902,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5257476950753317,
+      "grad_norm": 2.4321214797327255,
+      "learning_rate": 5.717463109955896e-06,
+      "loss": 0.2648,
+      "step": 1357
+    },
+    {
+      "epoch": 1.526872048572071,
+      "grad_norm": 2.286039354739911,
+      "learning_rate": 5.710985243948823e-06,
+      "loss": 0.2786,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5279964020688106,
+      "grad_norm": 2.860932290989918,
+      "learning_rate": 5.704506159687914e-06,
+      "loss": 0.2943,
+      "step": 1359
+    },
+    {
+      "epoch": 1.52912075556555,
+      "grad_norm": 2.7985107251544776,
+      "learning_rate": 5.698025868274903e-06,
+      "loss": 0.2828,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5302451090622893,
+      "grad_norm": 2.596928274285701,
+      "learning_rate": 5.6915443808135964e-06,
+      "loss": 0.269,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5313694625590286,
+      "grad_norm": 2.352052258442473,
+      "learning_rate": 5.6850617084098416e-06,
+      "loss": 0.231,
+      "step": 1362
+    },
+    {
+      "epoch": 1.532493816055768,
+      "grad_norm": 2.7676769587484134,
+      "learning_rate": 5.678577862171523e-06,
+      "loss": 0.3404,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5336181695525073,
+      "grad_norm": 2.6684057394548364,
+      "learning_rate": 5.672092853208535e-06,
+      "loss": 0.2582,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5347425230492466,
+      "grad_norm": 2.289132883138945,
+      "learning_rate": 5.665606692632762e-06,
+      "loss": 0.2496,
+      "step": 1365
+    },
+    {
+      "epoch": 1.535866876545986,
+      "grad_norm": 2.628564854458908,
+      "learning_rate": 5.659119391558065e-06,
+      "loss": 0.3007,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5369912300427253,
+      "grad_norm": 2.4887721532199034,
+      "learning_rate": 5.65263096110026e-06,
+      "loss": 0.226,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5381155835394646,
+      "grad_norm": 2.8181161261528778,
+      "learning_rate": 5.646141412377089e-06,
+      "loss": 0.3389,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5392399370362042,
+      "grad_norm": 2.5420781424504586,
+      "learning_rate": 5.639650756508222e-06,
+      "loss": 0.247,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5403642905329435,
+      "grad_norm": 2.412676115561819,
+      "learning_rate": 5.63315900461522e-06,
+      "loss": 0.2318,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5414886440296829,
+      "grad_norm": 2.7877756305176074,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 0.3059,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5426129975264224,
+      "grad_norm": 3.4413559213703913,
+      "learning_rate": 5.620172257252427e-06,
+      "loss": 0.2985,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5437373510231618,
+      "grad_norm": 2.667113632914508,
+      "learning_rate": 5.613677284035075e-06,
+      "loss": 0.3091,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5448617045199011,
+      "grad_norm": 2.5832888084063947,
+      "learning_rate": 5.607181259298424e-06,
+      "loss": 0.3208,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5459860580166405,
+      "grad_norm": 2.679620885313586,
+      "learning_rate": 5.600684194173236e-06,
+      "loss": 0.2878,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5471104115133798,
+      "grad_norm": 2.935844724741026,
+      "learning_rate": 5.594186099792055e-06,
+      "loss": 0.2937,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5482347650101191,
+      "grad_norm": 2.9902596219822155,
+      "learning_rate": 5.587686987289189e-06,
+      "loss": 0.3624,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5493591185068585,
+      "grad_norm": 2.589225936817787,
+      "learning_rate": 5.58118686780069e-06,
+      "loss": 0.268,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5504834720035978,
+      "grad_norm": 2.6492435481523198,
+      "learning_rate": 5.5746857524643335e-06,
+      "loss": 0.2861,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5516078255003372,
+      "grad_norm": 2.7391156290418586,
+      "learning_rate": 5.5681836524196065e-06,
+      "loss": 0.2694,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5527321789970767,
+      "grad_norm": 2.897429229501401,
+      "learning_rate": 5.561680578807677e-06,
+      "loss": 0.342,
+      "step": 1381
+    },
+    {
+      "epoch": 1.553856532493816,
+      "grad_norm": 2.2976640555408196,
+      "learning_rate": 5.555176542771389e-06,
+      "loss": 0.2813,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5549808859905554,
+      "grad_norm": 2.9784389470866692,
+      "learning_rate": 5.548671555455226e-06,
+      "loss": 0.3411,
+      "step": 1383
+    },
+    {
+      "epoch": 1.556105239487295,
+      "grad_norm": 2.636475001848819,
+      "learning_rate": 5.542165628005311e-06,
+      "loss": 0.2725,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5572295929840343,
+      "grad_norm": 2.8234961947345063,
+      "learning_rate": 5.535658771569369e-06,
+      "loss": 0.3088,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5583539464807736,
+      "grad_norm": 2.546906116674042,
+      "learning_rate": 5.529150997296724e-06,
+      "loss": 0.3197,
+      "step": 1386
+    },
+    {
+      "epoch": 1.559478299977513,
+      "grad_norm": 2.7739728179828407,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.2977,
+      "step": 1387
+    },
+    {
+      "epoch": 1.5606026534742523,
+      "grad_norm": 2.638817664728077,
+      "learning_rate": 5.51613273984645e-06,
+      "loss": 0.2487,
+      "step": 1388
+    },
+    {
+      "epoch": 1.5617270069709916,
+      "grad_norm": 2.6907666870251314,
+      "learning_rate": 5.50962227897525e-06,
+      "loss": 0.2746,
+      "step": 1389
+    },
+    {
+      "epoch": 1.562851360467731,
+      "grad_norm": 2.5055288157969313,
+      "learning_rate": 5.503110944880169e-06,
+      "loss": 0.2641,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5639757139644703,
+      "grad_norm": 2.938333590950068,
+      "learning_rate": 5.496598748718196e-06,
+      "loss": 0.3272,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5651000674612097,
+      "grad_norm": 2.846596639566283,
+      "learning_rate": 5.490085701647805e-06,
+      "loss": 0.2867,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5662244209579492,
+      "grad_norm": 3.0859537685586425,
+      "learning_rate": 5.483571814828921e-06,
+      "loss": 0.296,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5673487744546886,
+      "grad_norm": 2.1825068584523604,
+      "learning_rate": 5.477057099422913e-06,
+      "loss": 0.2313,
+      "step": 1394
+    },
+    {
+      "epoch": 1.568473127951428,
+      "grad_norm": 3.676847513677605,
+      "learning_rate": 5.470541566592573e-06,
+      "loss": 0.3308,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5695974814481675,
+      "grad_norm": 2.7604916698273563,
+      "learning_rate": 5.464025227502082e-06,
+      "loss": 0.281,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5707218349449068,
+      "grad_norm": 2.6393275663137934,
+      "learning_rate": 5.457508093317013e-06,
+      "loss": 0.2395,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5718461884416461,
+      "grad_norm": 2.396943497172001,
+      "learning_rate": 5.450990175204296e-06,
+      "loss": 0.2613,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5729705419383855,
+      "grad_norm": 2.612344916108006,
+      "learning_rate": 5.4444714843322085e-06,
+      "loss": 0.3046,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5740948954351248,
+      "grad_norm": 2.8228762360758264,
+      "learning_rate": 5.437952031870347e-06,
+      "loss": 0.3062,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5752192489318642,
+      "grad_norm": 3.059576933871892,
+      "learning_rate": 5.431431828989618e-06,
+      "loss": 0.3328,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5763436024286035,
+      "grad_norm": 2.4833817731026833,
+      "learning_rate": 5.4249108868622095e-06,
+      "loss": 0.2405,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5774679559253428,
+      "grad_norm": 2.8957894028582682,
+      "learning_rate": 5.41838921666158e-06,
+      "loss": 0.3279,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5785923094220822,
+      "grad_norm": 2.8368846135763786,
+      "learning_rate": 5.411866829562429e-06,
+      "loss": 0.2461,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5797166629188215,
+      "grad_norm": 2.8162134354569064,
+      "learning_rate": 5.405343736740695e-06,
+      "loss": 0.3042,
+      "step": 1405
+    },
+    {
+      "epoch": 1.580841016415561,
+      "grad_norm": 2.4373094382129685,
+      "learning_rate": 5.398819949373516e-06,
+      "loss": 0.2238,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5819653699123004,
+      "grad_norm": 2.7390927415204858,
+      "learning_rate": 5.392295478639226e-06,
+      "loss": 0.2793,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5830897234090398,
+      "grad_norm": 2.5903207096597898,
+      "learning_rate": 5.385770335717327e-06,
+      "loss": 0.2983,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5842140769057793,
+      "grad_norm": 2.398978752525999,
+      "learning_rate": 5.37924453178847e-06,
+      "loss": 0.2658,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5853384304025187,
+      "grad_norm": 2.5882774804407167,
+      "learning_rate": 5.372718078034449e-06,
+      "loss": 0.2336,
+      "step": 1410
+    },
+    {
+      "epoch": 1.586462783899258,
+      "grad_norm": 2.8610576712533944,
+      "learning_rate": 5.366190985638159e-06,
+      "loss": 0.3173,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5875871373959973,
+      "grad_norm": 2.674642398405688,
+      "learning_rate": 5.3596632657835975e-06,
+      "loss": 0.3202,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5887114908927367,
+      "grad_norm": 2.4529857967155317,
+      "learning_rate": 5.353134929655834e-06,
+      "loss": 0.2414,
+      "step": 1413
+    },
+    {
+      "epoch": 1.589835844389476,
+      "grad_norm": 3.187717543988481,
+      "learning_rate": 5.346605988440995e-06,
+      "loss": 0.3312,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5909601978862153,
+      "grad_norm": 3.1284766068227827,
+      "learning_rate": 5.3400764533262415e-06,
+      "loss": 0.3489,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5920845513829547,
+      "grad_norm": 2.4913311762035417,
+      "learning_rate": 5.333546335499756e-06,
+      "loss": 0.2428,
+      "step": 1416
+    },
+    {
+      "epoch": 1.593208904879694,
+      "grad_norm": 2.8323844837417176,
+      "learning_rate": 5.327015646150716e-06,
+      "loss": 0.3047,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5943332583764336,
+      "grad_norm": 2.7240348402774632,
+      "learning_rate": 5.320484396469281e-06,
+      "loss": 0.315,
+      "step": 1418
+    },
+    {
+      "epoch": 1.595457611873173,
+      "grad_norm": 2.8986598881561836,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 0.3102,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5965819653699123,
+      "grad_norm": 2.563026827418521,
+      "learning_rate": 5.3074202608746365e-06,
+      "loss": 0.2511,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5977063188666518,
+      "grad_norm": 2.517384879030703,
+      "learning_rate": 5.300887397346468e-06,
+      "loss": 0.2743,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5988306723633912,
+      "grad_norm": 2.7814223711735817,
+      "learning_rate": 5.294354018255945e-06,
+      "loss": 0.3474,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5999550258601305,
+      "grad_norm": 2.536243633191927,
+      "learning_rate": 5.287820134797837e-06,
+      "loss": 0.2992,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6010793793568698,
+      "grad_norm": 2.2374418958787348,
+      "learning_rate": 5.2812857581677735e-06,
+      "loss": 0.2068,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6022037328536092,
+      "grad_norm": 2.5802045300235252,
+      "learning_rate": 5.27475089956223e-06,
+      "loss": 0.2281,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6033280863503485,
+      "grad_norm": 2.64152305028715,
+      "learning_rate": 5.268215570178509e-06,
+      "loss": 0.2782,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6044524398470879,
+      "grad_norm": 2.615610427996027,
+      "learning_rate": 5.2616797812147205e-06,
+      "loss": 0.2685,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6055767933438272,
+      "grad_norm": 2.2483500889106387,
+      "learning_rate": 5.255143543869759e-06,
+      "loss": 0.2354,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6067011468405665,
+      "grad_norm": 2.71284567829115,
+      "learning_rate": 5.24860686934329e-06,
+      "loss": 0.2719,
+      "step": 1429
+    },
+    {
+      "epoch": 1.607825500337306,
+      "grad_norm": 2.4641881988993504,
+      "learning_rate": 5.242069768835728e-06,
+      "loss": 0.2401,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6089498538340454,
+      "grad_norm": 2.96588117466104,
+      "learning_rate": 5.235532253548213e-06,
+      "loss": 0.2808,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6100742073307848,
+      "grad_norm": 2.844200590884009,
+      "learning_rate": 5.228994334682605e-06,
+      "loss": 0.3159,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6111985608275243,
+      "grad_norm": 3.2162981045262584,
+      "learning_rate": 5.2224560234414446e-06,
+      "loss": 0.2848,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6123229143242637,
+      "grad_norm": 3.3804173051657322,
+      "learning_rate": 5.215917331027952e-06,
+      "loss": 0.2996,
+      "step": 1434
+    },
+    {
+      "epoch": 1.613447267821003,
+      "grad_norm": 2.525681218114005,
+      "learning_rate": 5.209378268645998e-06,
+      "loss": 0.2841,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6145716213177423,
+      "grad_norm": 2.6487550159467443,
+      "learning_rate": 5.20283884750009e-06,
+      "loss": 0.2864,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6156959748144817,
+      "grad_norm": 2.317052286727151,
+      "learning_rate": 5.1962990787953436e-06,
+      "loss": 0.2729,
+      "step": 1437
+    },
+    {
+      "epoch": 1.616820328311221,
+      "grad_norm": 2.5675092651306843,
+      "learning_rate": 5.189758973737478e-06,
+      "loss": 0.2791,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6179446818079604,
+      "grad_norm": 2.9084938995318512,
+      "learning_rate": 5.183218543532782e-06,
+      "loss": 0.2969,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6190690353046997,
+      "grad_norm": 2.5892462878632534,
+      "learning_rate": 5.176677799388107e-06,
+      "loss": 0.243,
+      "step": 1440
+    },
+    {
+      "epoch": 1.620193388801439,
+      "grad_norm": 2.81029449209127,
+      "learning_rate": 5.170136752510837e-06,
+      "loss": 0.2623,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6213177422981786,
+      "grad_norm": 2.6147524470715298,
+      "learning_rate": 5.1635954141088815e-06,
+      "loss": 0.2812,
+      "step": 1442
+    },
+    {
+      "epoch": 1.622442095794918,
+      "grad_norm": 2.240347988975483,
+      "learning_rate": 5.157053795390642e-06,
+      "loss": 0.2283,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6235664492916573,
+      "grad_norm": 2.3963874681809374,
+      "learning_rate": 5.150511907565006e-06,
+      "loss": 0.2938,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6246908027883968,
+      "grad_norm": 3.5427507210825566,
+      "learning_rate": 5.143969761841317e-06,
+      "loss": 0.3007,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6258151562851362,
+      "grad_norm": 2.5489453109444495,
+      "learning_rate": 5.1374273694293676e-06,
+      "loss": 0.2823,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6269395097818755,
+      "grad_norm": 3.0859460549968882,
+      "learning_rate": 5.130884741539367e-06,
+      "loss": 0.3658,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6280638632786149,
+      "grad_norm": 2.35918701064532,
+      "learning_rate": 5.124341889381929e-06,
+      "loss": 0.2418,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6291882167753542,
+      "grad_norm": 2.5921090178447814,
+      "learning_rate": 5.117798824168052e-06,
+      "loss": 0.3061,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6303125702720935,
+      "grad_norm": 2.470693539138189,
+      "learning_rate": 5.1112555571091005e-06,
+      "loss": 0.2256,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6314369237688329,
+      "grad_norm": 2.9049506725244734,
+      "learning_rate": 5.1047120994167855e-06,
+      "loss": 0.3095,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6325612772655722,
+      "grad_norm": 2.562246961243346,
+      "learning_rate": 5.098168462303141e-06,
+      "loss": 0.2974,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6336856307623115,
+      "grad_norm": 2.6746923861261496,
+      "learning_rate": 5.091624656980515e-06,
+      "loss": 0.3231,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6348099842590509,
+      "grad_norm": 3.1053490212850545,
+      "learning_rate": 5.085080694661535e-06,
+      "loss": 0.3036,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6359343377557904,
+      "grad_norm": 2.7312787542926302,
+      "learning_rate": 5.078536586559104e-06,
+      "loss": 0.3016,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6370586912525298,
+      "grad_norm": 2.679929159190367,
+      "learning_rate": 5.071992343886373e-06,
+      "loss": 0.2574,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6381830447492691,
+      "grad_norm": 2.4995730078463363,
+      "learning_rate": 5.065447977856723e-06,
+      "loss": 0.2885,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6393073982460087,
+      "grad_norm": 2.777410177535025,
+      "learning_rate": 5.058903499683746e-06,
+      "loss": 0.2926,
+      "step": 1458
+    },
+    {
+      "epoch": 1.640431751742748,
+      "grad_norm": 2.748998621771593,
+      "learning_rate": 5.05235892058123e-06,
+      "loss": 0.3031,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6415561052394874,
+      "grad_norm": 2.7243275887270055,
+      "learning_rate": 5.04581425176313e-06,
+      "loss": 0.33,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6426804587362267,
+      "grad_norm": 2.4476143426961996,
+      "learning_rate": 5.039269504443557e-06,
+      "loss": 0.265,
+      "step": 1461
+    },
+    {
+      "epoch": 1.643804812232966,
+      "grad_norm": 2.4671076980323217,
+      "learning_rate": 5.0327246898367595e-06,
+      "loss": 0.2849,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6449291657297054,
+      "grad_norm": 2.50732101763821,
+      "learning_rate": 5.0261798191570975e-06,
+      "loss": 0.2777,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6460535192264447,
+      "grad_norm": 2.6850348201326724,
+      "learning_rate": 5.019634903619031e-06,
+      "loss": 0.3005,
+      "step": 1464
+    },
+    {
+      "epoch": 1.647177872723184,
+      "grad_norm": 2.7798565844284995,
+      "learning_rate": 5.013089954437091e-06,
+      "loss": 0.3218,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6483022262199234,
+      "grad_norm": 3.6185778007931013,
+      "learning_rate": 5.006544982825872e-06,
+      "loss": 0.2991,
+      "step": 1466
+    },
+    {
+      "epoch": 1.649426579716663,
+      "grad_norm": 2.6541163177996614,
+      "learning_rate": 5e-06,
+      "loss": 0.2353,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6505509332134023,
+      "grad_norm": 2.8087791246935327,
+      "learning_rate": 4.993455017174132e-06,
+      "loss": 0.2932,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6516752867101416,
+      "grad_norm": 2.644409006883856,
+      "learning_rate": 4.9869100455629105e-06,
+      "loss": 0.2821,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6527996402068812,
+      "grad_norm": 2.425505042349941,
+      "learning_rate": 4.980365096380971e-06,
+      "loss": 0.2788,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6539239937036205,
+      "grad_norm": 2.4500276830545245,
+      "learning_rate": 4.9738201808429025e-06,
+      "loss": 0.2651,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6550483472003599,
+      "grad_norm": 2.7161888373926915,
+      "learning_rate": 4.967275310163241e-06,
+      "loss": 0.3397,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6561727006970992,
+      "grad_norm": 2.3399247094057043,
+      "learning_rate": 4.9607304955564456e-06,
+      "loss": 0.2385,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6572970541938385,
+      "grad_norm": 2.6006009251014044,
+      "learning_rate": 4.954185748236871e-06,
+      "loss": 0.2781,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6584214076905779,
+      "grad_norm": 2.577898097198441,
+      "learning_rate": 4.9476410794187726e-06,
+      "loss": 0.2399,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6595457611873172,
+      "grad_norm": 2.8408541664549833,
+      "learning_rate": 4.941096500316254e-06,
+      "loss": 0.3243,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6606701146840566,
+      "grad_norm": 2.211291907580132,
+      "learning_rate": 4.934552022143279e-06,
+      "loss": 0.2193,
+      "step": 1477
+    },
+    {
+      "epoch": 1.661794468180796,
+      "grad_norm": 2.419711707973134,
+      "learning_rate": 4.928007656113628e-06,
+      "loss": 0.2547,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6629188216775355,
+      "grad_norm": 2.639071724986486,
+      "learning_rate": 4.921463413440898e-06,
+      "loss": 0.3428,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6640431751742748,
+      "grad_norm": 2.7233127079517847,
+      "learning_rate": 4.914919305338468e-06,
+      "loss": 0.244,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6651675286710141,
+      "grad_norm": 2.6555770781271613,
+      "learning_rate": 4.908375343019487e-06,
+      "loss": 0.2929,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6662918821677537,
+      "grad_norm": 2.692090318345248,
+      "learning_rate": 4.90183153769686e-06,
+      "loss": 0.2108,
+      "step": 1482
+    },
+    {
+      "epoch": 1.667416235664493,
+      "grad_norm": 2.6768417244863167,
+      "learning_rate": 4.895287900583216e-06,
+      "loss": 0.3233,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6685405891612324,
+      "grad_norm": 3.1578387073594896,
+      "learning_rate": 4.888744442890901e-06,
+      "loss": 0.2831,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6696649426579717,
+      "grad_norm": 2.731315638546983,
+      "learning_rate": 4.88220117583195e-06,
+      "loss": 0.2875,
+      "step": 1485
+    },
+    {
+      "epoch": 1.670789296154711,
+      "grad_norm": 2.5589005996937346,
+      "learning_rate": 4.875658110618074e-06,
+      "loss": 0.2872,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6719136496514504,
+      "grad_norm": 2.7183396097716233,
+      "learning_rate": 4.869115258460636e-06,
+      "loss": 0.2569,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6730380031481897,
+      "grad_norm": 2.5877376163972756,
+      "learning_rate": 4.862572630570633e-06,
+      "loss": 0.2644,
+      "step": 1488
+    },
+    {
+      "epoch": 1.674162356644929,
+      "grad_norm": 2.2123438403486313,
+      "learning_rate": 4.8560302381586835e-06,
+      "loss": 0.204,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6752867101416684,
+      "grad_norm": 2.8406375088805795,
+      "learning_rate": 4.849488092434995e-06,
+      "loss": 0.3101,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6764110636384078,
+      "grad_norm": 2.7839852680507255,
+      "learning_rate": 4.842946204609359e-06,
+      "loss": 0.2764,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6775354171351473,
+      "grad_norm": 3.0386007401425745,
+      "learning_rate": 4.83640458589112e-06,
+      "loss": 0.2733,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6786597706318866,
+      "grad_norm": 2.4529149429418804,
+      "learning_rate": 4.829863247489163e-06,
+      "loss": 0.2783,
+      "step": 1493
+    },
+    {
+      "epoch": 1.679784124128626,
+      "grad_norm": 2.6770221106109275,
+      "learning_rate": 4.823322200611895e-06,
+      "loss": 0.2794,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6809084776253655,
+      "grad_norm": 2.835039042285269,
+      "learning_rate": 4.816781456467218e-06,
+      "loss": 0.3439,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6820328311221049,
+      "grad_norm": 3.0450828041214053,
+      "learning_rate": 4.810241026262524e-06,
+      "loss": 0.2975,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6831571846188442,
+      "grad_norm": 3.038261078749211,
+      "learning_rate": 4.803700921204659e-06,
+      "loss": 0.3236,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6842815381155836,
+      "grad_norm": 2.7273820565156743,
+      "learning_rate": 4.797161152499911e-06,
+      "loss": 0.2965,
+      "step": 1498
+    },
+    {
+      "epoch": 1.685405891612323,
+      "grad_norm": 2.6246773671759356,
+      "learning_rate": 4.7906217313540035e-06,
+      "loss": 0.3494,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6865302451090622,
+      "grad_norm": 2.4438399041632968,
+      "learning_rate": 4.784082668972048e-06,
+      "loss": 0.2743,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6865302451090622,
+      "eval_loss": 0.24364010989665985,
+      "eval_runtime": 358.87,
+      "eval_samples_per_second": 148.686,
+      "eval_steps_per_second": 4.648,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6876545986058016,
+      "grad_norm": 2.5895939074395646,
+      "learning_rate": 4.777543976558557e-06,
+      "loss": 0.2363,
+      "step": 1501
+    },
+    {
+      "epoch": 1.688778952102541,
+      "grad_norm": 3.502672107470009,
+      "learning_rate": 4.771005665317398e-06,
+      "loss": 0.2597,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6899033055992803,
+      "grad_norm": 2.599272828341689,
+      "learning_rate": 4.7644677464517874e-06,
+      "loss": 0.298,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6910276590960198,
+      "grad_norm": 2.4968744232763598,
+      "learning_rate": 4.757930231164275e-06,
+      "loss": 0.2679,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6921520125927592,
+      "grad_norm": 2.530700458806166,
+      "learning_rate": 4.751393130656712e-06,
+      "loss": 0.2727,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6932763660894985,
+      "grad_norm": 2.2519639987253517,
+      "learning_rate": 4.744856456130243e-06,
+      "loss": 0.2395,
+      "step": 1506
+    },
+    {
+      "epoch": 1.694400719586238,
+      "grad_norm": 2.502075493277466,
+      "learning_rate": 4.738320218785281e-06,
+      "loss": 0.2344,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6955250730829774,
+      "grad_norm": 2.5575146419605357,
+      "learning_rate": 4.731784429821493e-06,
+      "loss": 0.2964,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6966494265797167,
+      "grad_norm": 2.655646684760487,
+      "learning_rate": 4.725249100437773e-06,
+      "loss": 0.29,
+      "step": 1509
+    },
+    {
+      "epoch": 1.697773780076456,
+      "grad_norm": 2.6531033394585877,
+      "learning_rate": 4.718714241832228e-06,
+      "loss": 0.2352,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6988981335731954,
+      "grad_norm": 2.9021312084344566,
+      "learning_rate": 4.712179865202164e-06,
+      "loss": 0.3147,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7000224870699348,
+      "grad_norm": 2.509384433276035,
+      "learning_rate": 4.705645981744055e-06,
+      "loss": 0.282,
+      "step": 1512
+    },
+    {
+      "epoch": 1.701146840566674,
+      "grad_norm": 2.552395381709117,
+      "learning_rate": 4.6991126026535336e-06,
+      "loss": 0.2912,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7022711940634134,
+      "grad_norm": 2.6041263231543423,
+      "learning_rate": 4.692579739125365e-06,
+      "loss": 0.2924,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7033955475601528,
+      "grad_norm": 2.4993266572226225,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 0.2185,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7045199010568923,
+      "grad_norm": 2.898533809098515,
+      "learning_rate": 4.679515603530721e-06,
+      "loss": 0.2672,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7056442545536317,
+      "grad_norm": 3.10925501485894,
+      "learning_rate": 4.672984353849285e-06,
+      "loss": 0.3179,
+      "step": 1517
+    },
+    {
+      "epoch": 1.706768608050371,
+      "grad_norm": 2.8407308290767976,
+      "learning_rate": 4.6664536645002456e-06,
+      "loss": 0.3212,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7078929615471106,
+      "grad_norm": 2.8256402585051,
+      "learning_rate": 4.659923546673761e-06,
+      "loss": 0.2999,
+      "step": 1519
+    },
+    {
+      "epoch": 1.70901731504385,
+      "grad_norm": 2.4586431100321016,
+      "learning_rate": 4.653394011559007e-06,
+      "loss": 0.2577,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7101416685405892,
+      "grad_norm": 2.6811786937655255,
+      "learning_rate": 4.646865070344168e-06,
+      "loss": 0.2442,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7112660220373286,
+      "grad_norm": 2.610683096529766,
+      "learning_rate": 4.640336734216403e-06,
+      "loss": 0.3432,
+      "step": 1522
+    },
+    {
+      "epoch": 1.712390375534068,
+      "grad_norm": 2.4022792318848794,
+      "learning_rate": 4.6338090143618435e-06,
+      "loss": 0.3264,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7135147290308073,
+      "grad_norm": 2.5784957158207447,
+      "learning_rate": 4.627281921965552e-06,
+      "loss": 0.2659,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7146390825275466,
+      "grad_norm": 2.8318187201826315,
+      "learning_rate": 4.620755468211531e-06,
+      "loss": 0.2404,
+      "step": 1525
+    },
+    {
+      "epoch": 1.715763436024286,
+      "grad_norm": 2.6721861235389857,
+      "learning_rate": 4.614229664282677e-06,
+      "loss": 0.2814,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7168877895210253,
+      "grad_norm": 2.3684641441002734,
+      "learning_rate": 4.6077045213607765e-06,
+      "loss": 0.2509,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7180121430177648,
+      "grad_norm": 2.548625314998276,
+      "learning_rate": 4.601180050626486e-06,
+      "loss": 0.2489,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7191364965145042,
+      "grad_norm": 2.5268470851484794,
+      "learning_rate": 4.594656263259307e-06,
+      "loss": 0.295,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7202608500112435,
+      "grad_norm": 3.30755277458132,
+      "learning_rate": 4.588133170437572e-06,
+      "loss": 0.347,
+      "step": 1530
+    },
+    {
+      "epoch": 1.721385203507983,
+      "grad_norm": 2.5766764298869185,
+      "learning_rate": 4.581610783338424e-06,
+      "loss": 0.3254,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7225095570047224,
+      "grad_norm": 2.4978218492385214,
+      "learning_rate": 4.575089113137792e-06,
+      "loss": 0.2596,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7236339105014618,
+      "grad_norm": 2.9209807801549137,
+      "learning_rate": 4.568568171010384e-06,
+      "loss": 0.3373,
+      "step": 1533
+    },
+    {
+      "epoch": 1.724758263998201,
+      "grad_norm": 2.7363941601611304,
+      "learning_rate": 4.562047968129654e-06,
+      "loss": 0.264,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7258826174949404,
+      "grad_norm": 2.5249393446878754,
+      "learning_rate": 4.555528515667793e-06,
+      "loss": 0.2409,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7270069709916798,
+      "grad_norm": 2.6903751944329803,
+      "learning_rate": 4.5490098247957045e-06,
+      "loss": 0.2984,
+      "step": 1536
+    },
+    {
+      "epoch": 1.728131324488419,
+      "grad_norm": 3.0263010466944285,
+      "learning_rate": 4.542491906682988e-06,
+      "loss": 0.256,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7292556779851584,
+      "grad_norm": 2.6749768078807215,
+      "learning_rate": 4.5359747724979205e-06,
+      "loss": 0.2548,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7303800314818978,
+      "grad_norm": 2.9866627980072797,
+      "learning_rate": 4.529458433407429e-06,
+      "loss": 0.3352,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7315043849786371,
+      "grad_norm": 2.654866698117159,
+      "learning_rate": 4.5229429005770875e-06,
+      "loss": 0.2712,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7326287384753767,
+      "grad_norm": 2.9836808764424245,
+      "learning_rate": 4.516428185171079e-06,
+      "loss": 0.303,
+      "step": 1541
+    },
+    {
+      "epoch": 1.733753091972116,
+      "grad_norm": 2.5861985769593265,
+      "learning_rate": 4.509914298352197e-06,
+      "loss": 0.2517,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7348774454688554,
+      "grad_norm": 3.0995271947123486,
+      "learning_rate": 4.5034012512818065e-06,
+      "loss": 0.3216,
+      "step": 1543
+    },
+    {
+      "epoch": 1.736001798965595,
+      "grad_norm": 2.5773947788346208,
+      "learning_rate": 4.496889055119832e-06,
+      "loss": 0.2928,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7371261524623343,
+      "grad_norm": 2.644851996447548,
+      "learning_rate": 4.490377721024751e-06,
+      "loss": 0.2836,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7382505059590736,
+      "grad_norm": 2.462064989481647,
+      "learning_rate": 4.483867260153551e-06,
+      "loss": 0.2783,
+      "step": 1546
+    },
+    {
+      "epoch": 1.739374859455813,
+      "grad_norm": 2.737521466503658,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.2796,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7404992129525523,
+      "grad_norm": 2.985886108411822,
+      "learning_rate": 4.470849002703279e-06,
+      "loss": 0.318,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7416235664492916,
+      "grad_norm": 2.818305661178237,
+      "learning_rate": 4.464341228430632e-06,
+      "loss": 0.2624,
+      "step": 1549
+    },
+    {
+      "epoch": 1.742747919946031,
+      "grad_norm": 2.3936503940050837,
+      "learning_rate": 4.457834371994691e-06,
+      "loss": 0.2253,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7438722734427703,
+      "grad_norm": 3.1190631833914138,
+      "learning_rate": 4.451328444544774e-06,
+      "loss": 0.3106,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7449966269395096,
+      "grad_norm": 2.5634945361397357,
+      "learning_rate": 4.4448234572286126e-06,
+      "loss": 0.2508,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7461209804362492,
+      "grad_norm": 2.7801056840481877,
+      "learning_rate": 4.438319421192323e-06,
+      "loss": 0.3245,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7472453339329885,
+      "grad_norm": 2.845805447833734,
+      "learning_rate": 4.431816347580395e-06,
+      "loss": 0.2344,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7483696874297279,
+      "grad_norm": 3.1565127652877942,
+      "learning_rate": 4.425314247535668e-06,
+      "loss": 0.2265,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7494940409264674,
+      "grad_norm": 3.028627145295381,
+      "learning_rate": 4.418813132199312e-06,
+      "loss": 0.268,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7506183944232068,
+      "grad_norm": 2.599832095926638,
+      "learning_rate": 4.4123130127108125e-06,
+      "loss": 0.2988,
+      "step": 1557
+    },
+    {
+      "epoch": 1.751742747919946,
+      "grad_norm": 2.824561350995646,
+      "learning_rate": 4.405813900207945e-06,
+      "loss": 0.2948,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7528671014166854,
+      "grad_norm": 2.7938050444800036,
+      "learning_rate": 4.399315805826765e-06,
+      "loss": 0.2629,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7539914549134248,
+      "grad_norm": 2.74349947403414,
+      "learning_rate": 4.392818740701579e-06,
+      "loss": 0.3017,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7551158084101641,
+      "grad_norm": 2.1906051365653183,
+      "learning_rate": 4.386322715964926e-06,
+      "loss": 0.2524,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7562401619069035,
+      "grad_norm": 2.470296485225604,
+      "learning_rate": 4.379827742747575e-06,
+      "loss": 0.2387,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7573645154036428,
+      "grad_norm": 2.591154068590832,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 0.2341,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7584888689003821,
+      "grad_norm": 2.8121839368112704,
+      "learning_rate": 4.3668409953847825e-06,
+      "loss": 0.3268,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7596132223971217,
+      "grad_norm": 2.5820821317475575,
+      "learning_rate": 4.3603492434917784e-06,
+      "loss": 0.2599,
+      "step": 1565
+    },
+    {
+      "epoch": 1.760737575893861,
+      "grad_norm": 2.970159189621139,
+      "learning_rate": 4.353858587622913e-06,
+      "loss": 0.2958,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7618619293906004,
+      "grad_norm": 2.6655809541092497,
+      "learning_rate": 4.347369038899744e-06,
+      "loss": 0.2496,
+      "step": 1567
+    },
+    {
+      "epoch": 1.76298628288734,
+      "grad_norm": 2.407254477225243,
+      "learning_rate": 4.340880608441935e-06,
+      "loss": 0.2735,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7641106363840793,
+      "grad_norm": 2.483256445435613,
+      "learning_rate": 4.3343933073672395e-06,
+      "loss": 0.2405,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7652349898808186,
+      "grad_norm": 2.5805481301895044,
+      "learning_rate": 4.327907146791467e-06,
+      "loss": 0.2639,
+      "step": 1570
+    },
+    {
+      "epoch": 1.766359343377558,
+      "grad_norm": 2.6149869770511547,
+      "learning_rate": 4.321422137828479e-06,
+      "loss": 0.3097,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7674836968742973,
+      "grad_norm": 2.640263099646082,
+      "learning_rate": 4.314938291590161e-06,
+      "loss": 0.2594,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7686080503710366,
+      "grad_norm": 2.68907642740071,
+      "learning_rate": 4.308455619186406e-06,
+      "loss": 0.2915,
+      "step": 1573
+    },
+    {
+      "epoch": 1.769732403867776,
+      "grad_norm": 2.7847510989592856,
+      "learning_rate": 4.301974131725098e-06,
+      "loss": 0.2758,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7708567573645153,
+      "grad_norm": 2.7653312327853126,
+      "learning_rate": 4.295493840312087e-06,
+      "loss": 0.2742,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7719811108612546,
+      "grad_norm": 2.739238538236603,
+      "learning_rate": 4.289014756051178e-06,
+      "loss": 0.3014,
+      "step": 1576
+    },
+    {
+      "epoch": 1.773105464357994,
+      "grad_norm": 3.2416231687105013,
+      "learning_rate": 4.282536890044105e-06,
+      "loss": 0.2968,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7742298178547335,
+      "grad_norm": 2.311425624128964,
+      "learning_rate": 4.276060253390511e-06,
+      "loss": 0.2092,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7753541713514729,
+      "grad_norm": 2.653159376778407,
+      "learning_rate": 4.269584857187942e-06,
+      "loss": 0.3189,
+      "step": 1579
+    },
+    {
+      "epoch": 1.7764785248482122,
+      "grad_norm": 2.7117514147717747,
+      "learning_rate": 4.263110712531809e-06,
+      "loss": 0.2722,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7776028783449518,
+      "grad_norm": 2.5733608394731013,
+      "learning_rate": 4.256637830515385e-06,
+      "loss": 0.2895,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7787272318416911,
+      "grad_norm": 3.07996547297896,
+      "learning_rate": 4.250166222229775e-06,
+      "loss": 0.305,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7798515853384305,
+      "grad_norm": 2.8326212006059532,
+      "learning_rate": 4.243695898763904e-06,
+      "loss": 0.3067,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7809759388351698,
+      "grad_norm": 2.5293983473908894,
+      "learning_rate": 4.237226871204496e-06,
+      "loss": 0.2723,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7821002923319091,
+      "grad_norm": 2.872394120632912,
+      "learning_rate": 4.230759150636049e-06,
+      "loss": 0.3459,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7832246458286485,
+      "grad_norm": 2.5736489365800717,
+      "learning_rate": 4.2242927481408335e-06,
+      "loss": 0.2843,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7843489993253878,
+      "grad_norm": 2.7024836852899075,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.2862,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7854733528221272,
+      "grad_norm": 2.5030270714510343,
+      "learning_rate": 4.211363941687819e-06,
+      "loss": 0.2232,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7865977063188665,
+      "grad_norm": 3.24373195157353,
+      "learning_rate": 4.2049015598831805e-06,
+      "loss": 0.2977,
+      "step": 1589
+    },
+    {
+      "epoch": 1.787722059815606,
+      "grad_norm": 2.6167463779114226,
+      "learning_rate": 4.198440540458045e-06,
+      "loss": 0.2649,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7888464133123454,
+      "grad_norm": 2.6978306937411145,
+      "learning_rate": 4.191980894483195e-06,
+      "loss": 0.2784,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7899707668090847,
+      "grad_norm": 2.8210479600166476,
+      "learning_rate": 4.185522633027057e-06,
+      "loss": 0.3017,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7910951203058243,
+      "grad_norm": 2.7021302484070255,
+      "learning_rate": 4.179065767155686e-06,
+      "loss": 0.3178,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7922194738025636,
+      "grad_norm": 2.814207524007699,
+      "learning_rate": 4.172610307932747e-06,
+      "loss": 0.2399,
+      "step": 1594
+    },
+    {
+      "epoch": 1.793343827299303,
+      "grad_norm": 2.677475004313743,
+      "learning_rate": 4.166156266419489e-06,
+      "loss": 0.2766,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7944681807960423,
+      "grad_norm": 2.650282462841417,
+      "learning_rate": 4.159703653674741e-06,
+      "loss": 0.2686,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7955925342927817,
+      "grad_norm": 2.392243660707686,
+      "learning_rate": 4.1532524807548776e-06,
+      "loss": 0.2678,
+      "step": 1597
+    },
+    {
+      "epoch": 1.796716887789521,
+      "grad_norm": 2.8080939003272047,
+      "learning_rate": 4.146802758713807e-06,
+      "loss": 0.2651,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7978412412862603,
+      "grad_norm": 2.4056901987429744,
+      "learning_rate": 4.140354498602952e-06,
+      "loss": 0.2365,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7989655947829997,
+      "grad_norm": 3.145595420009361,
+      "learning_rate": 4.1339077114712324e-06,
+      "loss": 0.2783,
+      "step": 1600
+    },
+    {
+      "epoch": 1.800089948279739,
+      "grad_norm": 2.6240804621003444,
+      "learning_rate": 4.1274624083650414e-06,
+      "loss": 0.2709,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8012143017764786,
+      "grad_norm": 2.861955691362295,
+      "learning_rate": 4.1210186003282275e-06,
+      "loss": 0.3202,
+      "step": 1602
+    },
+    {
+      "epoch": 1.802338655273218,
+      "grad_norm": 2.483493861641376,
+      "learning_rate": 4.114576298402085e-06,
+      "loss": 0.2832,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8034630087699572,
+      "grad_norm": 3.0062671474462017,
+      "learning_rate": 4.108135513625317e-06,
+      "loss": 0.3422,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8045873622666968,
+      "grad_norm": 2.5106738042210237,
+      "learning_rate": 4.1016962570340375e-06,
+      "loss": 0.2627,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8057117157634361,
+      "grad_norm": 2.660049621299163,
+      "learning_rate": 4.095258539661736e-06,
+      "loss": 0.2754,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8068360692601755,
+      "grad_norm": 2.874103669266149,
+      "learning_rate": 4.088822372539263e-06,
+      "loss": 0.2796,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8079604227569148,
+      "grad_norm": 2.3330728863347465,
+      "learning_rate": 4.082387766694819e-06,
+      "loss": 0.2516,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8090847762536542,
+      "grad_norm": 3.874058241142867,
+      "learning_rate": 4.0759547331539224e-06,
+      "loss": 0.3435,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8102091297503935,
+      "grad_norm": 2.744056557544653,
+      "learning_rate": 4.0695232829394046e-06,
+      "loss": 0.3273,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8113334832471328,
+      "grad_norm": 2.539850606959489,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 0.2705,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8124578367438722,
+      "grad_norm": 2.4945230924544797,
+      "learning_rate": 4.056665176567225e-06,
+      "loss": 0.2858,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8135821902406115,
+      "grad_norm": 2.989945762234656,
+      "learning_rate": 4.050238542441578e-06,
+      "loss": 0.3107,
+      "step": 1613
+    },
+    {
+      "epoch": 1.814706543737351,
+      "grad_norm": 2.619984249495354,
+      "learning_rate": 4.043813535706299e-06,
+      "loss": 0.2741,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8158308972340904,
+      "grad_norm": 2.614772540185921,
+      "learning_rate": 4.037390167370464e-06,
+      "loss": 0.2806,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8169552507308298,
+      "grad_norm": 2.483525502745818,
+      "learning_rate": 4.030968448440337e-06,
+      "loss": 0.2586,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8180796042275693,
+      "grad_norm": 2.770860916126362,
+      "learning_rate": 4.02454838991936e-06,
+      "loss": 0.2644,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8192039577243087,
+      "grad_norm": 2.412304939246745,
+      "learning_rate": 4.018130002808127e-06,
+      "loss": 0.2492,
+      "step": 1618
+    },
+    {
+      "epoch": 1.820328311221048,
+      "grad_norm": 2.8507239281595895,
+      "learning_rate": 4.0117132981043695e-06,
+      "loss": 0.2807,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8214526647177873,
+      "grad_norm": 2.7902608558497013,
+      "learning_rate": 4.005298286802938e-06,
+      "loss": 0.3276,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8225770182145267,
+      "grad_norm": 2.3745334941830185,
+      "learning_rate": 3.998884979895777e-06,
+      "loss": 0.2745,
+      "step": 1621
+    },
+    {
+      "epoch": 1.823701371711266,
+      "grad_norm": 2.6538070968306937,
+      "learning_rate": 3.992473388371914e-06,
+      "loss": 0.265,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8248257252080053,
+      "grad_norm": 2.4469590443541294,
+      "learning_rate": 3.986063523217439e-06,
+      "loss": 0.2336,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8259500787047447,
+      "grad_norm": 2.6898473649762185,
+      "learning_rate": 3.979655395415476e-06,
+      "loss": 0.24,
+      "step": 1624
+    },
+    {
+      "epoch": 1.827074432201484,
+      "grad_norm": 2.4451909101778355,
+      "learning_rate": 3.973249015946182e-06,
+      "loss": 0.2456,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8281987856982234,
+      "grad_norm": 2.306175822165644,
+      "learning_rate": 3.966844395786709e-06,
+      "loss": 0.2308,
+      "step": 1626
+    },
+    {
+      "epoch": 1.829323139194963,
+      "grad_norm": 2.6549369900604614,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.2655,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8304474926917023,
+      "grad_norm": 3.1959557998057178,
+      "learning_rate": 3.954040477290771e-06,
+      "loss": 0.3866,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8315718461884416,
+      "grad_norm": 2.920125337800839,
+      "learning_rate": 3.947641200893473e-06,
+      "loss": 0.3361,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8326961996851812,
+      "grad_norm": 2.7052129567786127,
+      "learning_rate": 3.9412437276842895e-06,
+      "loss": 0.2904,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8338205531819205,
+      "grad_norm": 2.6336016117434413,
+      "learning_rate": 3.934848068625117e-06,
+      "loss": 0.2586,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8349449066786598,
+      "grad_norm": 2.6207679987150105,
+      "learning_rate": 3.928454234674748e-06,
+      "loss": 0.235,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8360692601753992,
+      "grad_norm": 3.0045452855430956,
+      "learning_rate": 3.922062236788836e-06,
+      "loss": 0.2783,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8371936136721385,
+      "grad_norm": 2.5837589409340906,
+      "learning_rate": 3.9156720859199006e-06,
+      "loss": 0.299,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8383179671688779,
+      "grad_norm": 2.7590098927703632,
+      "learning_rate": 3.909283793017289e-06,
+      "loss": 0.2605,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8394423206656172,
+      "grad_norm": 3.0223285414273007,
+      "learning_rate": 3.902897369027165e-06,
+      "loss": 0.3279,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8405666741623565,
+      "grad_norm": 2.50966318816316,
+      "learning_rate": 3.8965128248924956e-06,
+      "loss": 0.2769,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8416910276590959,
+      "grad_norm": 2.1304566090503934,
+      "learning_rate": 3.890130171553021e-06,
+      "loss": 0.1999,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8428153811558354,
+      "grad_norm": 3.0199315839158807,
+      "learning_rate": 3.883749419945244e-06,
+      "loss": 0.2957,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8439397346525748,
+      "grad_norm": 2.6608896743771346,
+      "learning_rate": 3.877370581002408e-06,
+      "loss": 0.2902,
+      "step": 1640
+    },
+    {
+      "epoch": 1.845064088149314,
+      "grad_norm": 3.388297466396155,
+      "learning_rate": 3.870993665654482e-06,
+      "loss": 0.3074,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8461884416460537,
+      "grad_norm": 2.4245799424032852,
+      "learning_rate": 3.864618684828135e-06,
+      "loss": 0.2357,
+      "step": 1642
+    },
+    {
+      "epoch": 1.847312795142793,
+      "grad_norm": 3.0577867562106427,
+      "learning_rate": 3.8582456494467214e-06,
+      "loss": 0.3085,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8484371486395323,
+      "grad_norm": 2.7442331230711026,
+      "learning_rate": 3.851874570430266e-06,
+      "loss": 0.2893,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8495615021362717,
+      "grad_norm": 2.701161301643132,
+      "learning_rate": 3.845505458695438e-06,
+      "loss": 0.2607,
+      "step": 1645
+    },
+    {
+      "epoch": 1.850685855633011,
+      "grad_norm": 2.8843983258559054,
+      "learning_rate": 3.8391383251555375e-06,
+      "loss": 0.2677,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8518102091297504,
+      "grad_norm": 2.7120202473055524,
+      "learning_rate": 3.832773180720475e-06,
+      "loss": 0.2849,
+      "step": 1647
+    },
+    {
+      "epoch": 1.8529345626264897,
+      "grad_norm": 2.523673410788536,
+      "learning_rate": 3.826410036296747e-06,
+      "loss": 0.257,
+      "step": 1648
+    },
+    {
+      "epoch": 1.854058916123229,
+      "grad_norm": 2.741577191836433,
+      "learning_rate": 3.820048902787435e-06,
+      "loss": 0.2994,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8551832696199684,
+      "grad_norm": 2.854632590370757,
+      "learning_rate": 3.813689791092161e-06,
+      "loss": 0.3305,
+      "step": 1650
+    },
+    {
+      "epoch": 1.856307623116708,
+      "grad_norm": 2.5963320422144065,
+      "learning_rate": 3.8073327121070968e-06,
+      "loss": 0.2922,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8574319766134473,
+      "grad_norm": 2.48945902864117,
+      "learning_rate": 3.800977676724919e-06,
+      "loss": 0.2561,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8585563301101866,
+      "grad_norm": 2.855449522407969,
+      "learning_rate": 3.7946246958348077e-06,
+      "loss": 0.2797,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8596806836069262,
+      "grad_norm": 2.5544320482307605,
+      "learning_rate": 3.7882737803224243e-06,
+      "loss": 0.2139,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8608050371036655,
+      "grad_norm": 2.739558018531337,
+      "learning_rate": 3.7819249410698877e-06,
+      "loss": 0.3172,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8619293906004049,
+      "grad_norm": 2.737712706354599,
+      "learning_rate": 3.775578188955763e-06,
+      "loss": 0.2458,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8630537440971442,
+      "grad_norm": 2.88691265578672,
+      "learning_rate": 3.769233534855035e-06,
+      "loss": 0.3529,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8641780975938835,
+      "grad_norm": 2.9006063103271567,
+      "learning_rate": 3.7628909896390976e-06,
+      "loss": 0.3182,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8653024510906229,
+      "grad_norm": 2.546915339293291,
+      "learning_rate": 3.756550564175727e-06,
+      "loss": 0.2502,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8664268045873622,
+      "grad_norm": 2.9174184220550776,
+      "learning_rate": 3.75021226932907e-06,
+      "loss": 0.309,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8675511580841015,
+      "grad_norm": 2.8668347915671952,
+      "learning_rate": 3.743876115959623e-06,
+      "loss": 0.2841,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8686755115808409,
+      "grad_norm": 2.6283271029122144,
+      "learning_rate": 3.7375421149242102e-06,
+      "loss": 0.3157,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8697998650775802,
+      "grad_norm": 2.6520864718165766,
+      "learning_rate": 3.7312102770759724e-06,
+      "loss": 0.2375,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8709242185743198,
+      "grad_norm": 2.6980835259796496,
+      "learning_rate": 3.7248806132643395e-06,
+      "loss": 0.2238,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8720485720710591,
+      "grad_norm": 2.816847721745182,
+      "learning_rate": 3.7185531343350167e-06,
+      "loss": 0.2769,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8731729255677985,
+      "grad_norm": 2.210774257669556,
+      "learning_rate": 3.7122278511299697e-06,
+      "loss": 0.23,
+      "step": 1666
+    },
+    {
+      "epoch": 1.874297279064538,
+      "grad_norm": 2.9880356676857365,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.2905,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8754216325612774,
+      "grad_norm": 2.5671514954139405,
+      "learning_rate": 3.6995839152417173e-06,
+      "loss": 0.2118,
+      "step": 1668
+    },
+    {
+      "epoch": 1.8765459860580167,
+      "grad_norm": 2.525570273285523,
+      "learning_rate": 3.6932652842235537e-06,
+      "loss": 0.2626,
+      "step": 1669
+    },
+    {
+      "epoch": 1.877670339554756,
+      "grad_norm": 2.589594036843031,
+      "learning_rate": 3.686948892259703e-06,
+      "loss": 0.2896,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8787946930514954,
+      "grad_norm": 2.7623486740264775,
+      "learning_rate": 3.680634750173137e-06,
+      "loss": 0.264,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8799190465482347,
+      "grad_norm": 2.656088064850942,
+      "learning_rate": 3.6743228687829596e-06,
+      "loss": 0.2546,
+      "step": 1672
+    },
+    {
+      "epoch": 1.881043400044974,
+      "grad_norm": 2.4404517895382547,
+      "learning_rate": 3.668013258904414e-06,
+      "loss": 0.2463,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8821677535417134,
+      "grad_norm": 3.471314384866373,
+      "learning_rate": 3.661705931348838e-06,
+      "loss": 0.309,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8832921070384527,
+      "grad_norm": 2.5738321091755827,
+      "learning_rate": 3.655400896923672e-06,
+      "loss": 0.2907,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8844164605351923,
+      "grad_norm": 2.7031642057841108,
+      "learning_rate": 3.649098166432417e-06,
+      "loss": 0.2616,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8855408140319316,
+      "grad_norm": 2.2930167024440475,
+      "learning_rate": 3.6427977506746293e-06,
+      "loss": 0.2306,
+      "step": 1677
+    },
+    {
+      "epoch": 1.886665167528671,
+      "grad_norm": 2.995712865716739,
+      "learning_rate": 3.6364996604459017e-06,
+      "loss": 0.2941,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8877895210254105,
+      "grad_norm": 2.457277060355579,
+      "learning_rate": 3.630203906537838e-06,
+      "loss": 0.2497,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8889138745221499,
+      "grad_norm": 2.42376310050752,
+      "learning_rate": 3.623910499738043e-06,
+      "loss": 0.2421,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8900382280188892,
+      "grad_norm": 2.8961704497733516,
+      "learning_rate": 3.6176194508300973e-06,
+      "loss": 0.3162,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8911625815156285,
+      "grad_norm": 2.533404149126167,
+      "learning_rate": 3.6113307705935398e-06,
+      "loss": 0.2369,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8922869350123679,
+      "grad_norm": 3.301588628650935,
+      "learning_rate": 3.6050444698038547e-06,
+      "loss": 0.34,
+      "step": 1683
+    },
+    {
+      "epoch": 1.8934112885091072,
+      "grad_norm": 2.528210605397471,
+      "learning_rate": 3.5987605592324453e-06,
+      "loss": 0.2804,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8945356420058466,
+      "grad_norm": 2.894638456220528,
+      "learning_rate": 3.5924790496466233e-06,
+      "loss": 0.2742,
+      "step": 1685
+    },
+    {
+      "epoch": 1.895659995502586,
+      "grad_norm": 2.589454754109356,
+      "learning_rate": 3.5861999518095827e-06,
+      "loss": 0.263,
+      "step": 1686
+    },
+    {
+      "epoch": 1.8967843489993252,
+      "grad_norm": 3.250683971551812,
+      "learning_rate": 3.579923276480387e-06,
+      "loss": 0.3489,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8979087024960648,
+      "grad_norm": 2.7439824720887485,
+      "learning_rate": 3.5736490344139494e-06,
+      "loss": 0.3192,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8990330559928041,
+      "grad_norm": 2.2827715002390327,
+      "learning_rate": 3.5673772363610083e-06,
+      "loss": 0.2684,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9001574094895435,
+      "grad_norm": 2.7033239286573245,
+      "learning_rate": 3.5611078930681254e-06,
+      "loss": 0.3031,
+      "step": 1690
+    },
+    {
+      "epoch": 1.901281762986283,
+      "grad_norm": 2.985414669995917,
+      "learning_rate": 3.5548410152776414e-06,
+      "loss": 0.3086,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9024061164830224,
+      "grad_norm": 2.4657030093811145,
+      "learning_rate": 3.5485766137276894e-06,
+      "loss": 0.2657,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9035304699797617,
+      "grad_norm": 2.7074346641794187,
+      "learning_rate": 3.542314699152145e-06,
+      "loss": 0.2552,
+      "step": 1693
+    },
+    {
+      "epoch": 1.904654823476501,
+      "grad_norm": 2.6436851612557746,
+      "learning_rate": 3.536055282280627e-06,
+      "loss": 0.3014,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9057791769732404,
+      "grad_norm": 2.705758171938762,
+      "learning_rate": 3.5297983738384813e-06,
+      "loss": 0.2935,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9069035304699797,
+      "grad_norm": 2.7462376533288455,
+      "learning_rate": 3.523543984546744e-06,
+      "loss": 0.2644,
+      "step": 1696
+    },
+    {
+      "epoch": 1.908027883966719,
+      "grad_norm": 2.6171768186405426,
+      "learning_rate": 3.517292125122146e-06,
+      "loss": 0.273,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9091522374634584,
+      "grad_norm": 2.893327946444027,
+      "learning_rate": 3.511042806277075e-06,
+      "loss": 0.3537,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9102765909601978,
+      "grad_norm": 2.9667625873802437,
+      "learning_rate": 3.5047960387195673e-06,
+      "loss": 0.2438,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9114009444569373,
+      "grad_norm": 2.5438025939149407,
+      "learning_rate": 3.498551833153292e-06,
+      "loss": 0.2787,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9125252979536767,
+      "grad_norm": 2.7218123470147075,
+      "learning_rate": 3.492310200277522e-06,
+      "loss": 0.2682,
+      "step": 1701
+    },
+    {
+      "epoch": 1.913649651450416,
+      "grad_norm": 2.9100555244635857,
+      "learning_rate": 3.486071150787128e-06,
+      "loss": 0.2763,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9147740049471556,
+      "grad_norm": 2.7262064275284126,
+      "learning_rate": 3.4798346953725487e-06,
+      "loss": 0.2616,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9158983584438949,
+      "grad_norm": 2.348060966350837,
+      "learning_rate": 3.473600844719783e-06,
+      "loss": 0.2466,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9170227119406342,
+      "grad_norm": 2.5494485515833114,
+      "learning_rate": 3.4673696095103626e-06,
+      "loss": 0.2453,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9181470654373736,
+      "grad_norm": 3.045513294124238,
+      "learning_rate": 3.4611410004213373e-06,
+      "loss": 0.3546,
+      "step": 1706
+    },
+    {
+      "epoch": 1.919271418934113,
+      "grad_norm": 2.762271862676777,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.2794,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9203957724308522,
+      "grad_norm": 3.057460000640105,
+      "learning_rate": 3.4486917032901712e-06,
+      "loss": 0.3475,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9215201259275916,
+      "grad_norm": 3.531187534356175,
+      "learning_rate": 3.44247103657956e-06,
+      "loss": 0.3147,
+      "step": 1709
+    },
+    {
+      "epoch": 1.922644479424331,
+      "grad_norm": 2.779520411649804,
+      "learning_rate": 3.436253038652373e-06,
+      "loss": 0.2905,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9237688329210703,
+      "grad_norm": 2.462555770771326,
+      "learning_rate": 3.4300377201629753e-06,
+      "loss": 0.2408,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9248931864178096,
+      "grad_norm": 2.869669683555865,
+      "learning_rate": 3.4238250917611533e-06,
+      "loss": 0.3001,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9260175399145492,
+      "grad_norm": 2.6061622086119125,
+      "learning_rate": 3.4176151640920696e-06,
+      "loss": 0.256,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9271418934112885,
+      "grad_norm": 2.308647836331266,
+      "learning_rate": 3.4114079477962713e-06,
+      "loss": 0.2964,
+      "step": 1714
+    },
+    {
+      "epoch": 1.9282662469080278,
+      "grad_norm": 2.570872427051156,
+      "learning_rate": 3.40520345350965e-06,
+      "loss": 0.2773,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9293906004047674,
+      "grad_norm": 2.872485493360307,
+      "learning_rate": 3.3990016918634415e-06,
+      "loss": 0.318,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9305149539015067,
+      "grad_norm": 2.522290784553334,
+      "learning_rate": 3.3928026734841935e-06,
+      "loss": 0.2544,
+      "step": 1717
+    },
+    {
+      "epoch": 1.931639307398246,
+      "grad_norm": 2.582478822575269,
+      "learning_rate": 3.3866064089937533e-06,
+      "loss": 0.2937,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9327636608949854,
+      "grad_norm": 2.674401117550367,
+      "learning_rate": 3.3804129090092542e-06,
+      "loss": 0.2779,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9338880143917248,
+      "grad_norm": 2.9168178562439535,
+      "learning_rate": 3.3742221841430866e-06,
+      "loss": 0.3598,
+      "step": 1720
+    },
+    {
+      "epoch": 1.935012367888464,
+      "grad_norm": 2.596900966711773,
+      "learning_rate": 3.368034245002892e-06,
+      "loss": 0.2346,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9361367213852034,
+      "grad_norm": 3.0002933712833277,
+      "learning_rate": 3.3618491021915334e-06,
+      "loss": 0.3696,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9372610748819428,
+      "grad_norm": 2.745293521088584,
+      "learning_rate": 3.355666766307084e-06,
+      "loss": 0.3117,
+      "step": 1723
+    },
+    {
+      "epoch": 1.938385428378682,
+      "grad_norm": 2.832756271170832,
+      "learning_rate": 3.34948724794281e-06,
+      "loss": 0.2834,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9395097818754217,
+      "grad_norm": 2.658928094412263,
+      "learning_rate": 3.3433105576871448e-06,
+      "loss": 0.2646,
+      "step": 1725
+    },
+    {
+      "epoch": 1.940634135372161,
+      "grad_norm": 2.8528087984403667,
+      "learning_rate": 3.3371367061236826e-06,
+      "loss": 0.3296,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9417584888689003,
+      "grad_norm": 2.411561337350954,
+      "learning_rate": 3.330965703831146e-06,
+      "loss": 0.2228,
+      "step": 1727
+    },
+    {
+      "epoch": 1.94288284236564,
+      "grad_norm": 2.3429669526762793,
+      "learning_rate": 3.3247975613833805e-06,
+      "loss": 0.1993,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9440071958623792,
+      "grad_norm": 2.164421724592276,
+      "learning_rate": 3.3186322893493316e-06,
+      "loss": 0.2266,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9451315493591186,
+      "grad_norm": 2.4911110019693483,
+      "learning_rate": 3.312469898293022e-06,
+      "loss": 0.2519,
+      "step": 1730
+    },
+    {
+      "epoch": 1.946255902855858,
+      "grad_norm": 2.771675387573257,
+      "learning_rate": 3.3063103987735433e-06,
+      "loss": 0.3203,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9473802563525973,
+      "grad_norm": 3.157556993384277,
+      "learning_rate": 3.3001538013450285e-06,
+      "loss": 0.314,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9485046098493366,
+      "grad_norm": 2.6644788282271907,
+      "learning_rate": 3.294000116556641e-06,
+      "loss": 0.2955,
+      "step": 1733
+    },
+    {
+      "epoch": 1.949628963346076,
+      "grad_norm": 2.700657768040809,
+      "learning_rate": 3.287849354952552e-06,
+      "loss": 0.2378,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9507533168428153,
+      "grad_norm": 2.1779517815129643,
+      "learning_rate": 3.28170152707192e-06,
+      "loss": 0.2266,
+      "step": 1735
+    },
+    {
+      "epoch": 1.9518776703395546,
+      "grad_norm": 2.5917148467711772,
+      "learning_rate": 3.2755566434488855e-06,
+      "loss": 0.2944,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9530020238362942,
+      "grad_norm": 2.6357297296167514,
+      "learning_rate": 3.269414714612534e-06,
+      "loss": 0.2891,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9541263773330335,
+      "grad_norm": 2.646252824330066,
+      "learning_rate": 3.2632757510869005e-06,
+      "loss": 0.2852,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9552507308297729,
+      "grad_norm": 3.166130149444911,
+      "learning_rate": 3.2571397633909252e-06,
+      "loss": 0.3661,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9563750843265124,
+      "grad_norm": 2.3608228975581755,
+      "learning_rate": 3.2510067620384566e-06,
+      "loss": 0.2375,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9574994378232518,
+      "grad_norm": 2.783714973592718,
+      "learning_rate": 3.2448767575382284e-06,
+      "loss": 0.2629,
+      "step": 1741
+    },
+    {
+      "epoch": 1.958623791319991,
+      "grad_norm": 2.6564215470733132,
+      "learning_rate": 3.2387497603938327e-06,
+      "loss": 0.2386,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9597481448167304,
+      "grad_norm": 2.9954929451583054,
+      "learning_rate": 3.2326257811037154e-06,
+      "loss": 0.2704,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9608724983134698,
+      "grad_norm": 2.587229761450412,
+      "learning_rate": 3.2265048301611458e-06,
+      "loss": 0.3009,
+      "step": 1744
+    },
+    {
+      "epoch": 1.961996851810209,
+      "grad_norm": 3.220431184426228,
+      "learning_rate": 3.220386918054206e-06,
+      "loss": 0.3186,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9631212053069484,
+      "grad_norm": 2.865393178706932,
+      "learning_rate": 3.2142720552657746e-06,
+      "loss": 0.2713,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9642455588036878,
+      "grad_norm": 2.8586987418301955,
+      "learning_rate": 3.2081602522734987e-06,
+      "loss": 0.2872,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9653699123004271,
+      "grad_norm": 3.394807620117732,
+      "learning_rate": 3.2020515195497894e-06,
+      "loss": 0.2867,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9664942657971665,
+      "grad_norm": 3.4651814292929792,
+      "learning_rate": 3.195945867561791e-06,
+      "loss": 0.3323,
+      "step": 1749
+    },
+    {
+      "epoch": 1.967618619293906,
+      "grad_norm": 2.6863480425231754,
+      "learning_rate": 3.189843306771374e-06,
+      "loss": 0.228,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9687429727906454,
+      "grad_norm": 2.778711581193494,
+      "learning_rate": 3.183743847635109e-06,
+      "loss": 0.2271,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9698673262873847,
+      "grad_norm": 2.8276351112983007,
+      "learning_rate": 3.177647500604252e-06,
+      "loss": 0.2715,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9709916797841243,
+      "grad_norm": 2.639033561490985,
+      "learning_rate": 3.1715542761247286e-06,
+      "loss": 0.2899,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9721160332808636,
+      "grad_norm": 2.451229045244266,
+      "learning_rate": 3.1654641846371125e-06,
+      "loss": 0.2305,
+      "step": 1754
+    },
+    {
+      "epoch": 1.973240386777603,
+      "grad_norm": 2.168111289157082,
+      "learning_rate": 3.1593772365766107e-06,
+      "loss": 0.181,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9743647402743423,
+      "grad_norm": 2.759433965892206,
+      "learning_rate": 3.1532934423730423e-06,
+      "loss": 0.2887,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9754890937710816,
+      "grad_norm": 2.800039279931356,
+      "learning_rate": 3.147212812450819e-06,
+      "loss": 0.2382,
+      "step": 1757
+    },
+    {
+      "epoch": 1.976613447267821,
+      "grad_norm": 2.7959718698923566,
+      "learning_rate": 3.1411353572289404e-06,
+      "loss": 0.3185,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9777378007645603,
+      "grad_norm": 2.590360574912748,
+      "learning_rate": 3.1350610871209553e-06,
+      "loss": 0.2651,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9788621542612996,
+      "grad_norm": 2.7069779596014287,
+      "learning_rate": 3.1289900125349646e-06,
+      "loss": 0.2753,
+      "step": 1760
+    },
+    {
+      "epoch": 1.979986507758039,
+      "grad_norm": 2.385961135101505,
+      "learning_rate": 3.122922143873584e-06,
+      "loss": 0.2485,
+      "step": 1761
+    },
+    {
+      "epoch": 1.9811108612547785,
+      "grad_norm": 3.164404030256705,
+      "learning_rate": 3.1168574915339465e-06,
+      "loss": 0.2644,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9822352147515179,
+      "grad_norm": 2.6966795196740203,
+      "learning_rate": 3.110796065907665e-06,
+      "loss": 0.2408,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9833595682482572,
+      "grad_norm": 2.976674057602844,
+      "learning_rate": 3.104737877380828e-06,
+      "loss": 0.3295,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9844839217449968,
+      "grad_norm": 2.638742096274046,
+      "learning_rate": 3.0986829363339766e-06,
+      "loss": 0.2432,
+      "step": 1765
+    },
+    {
+      "epoch": 1.985608275241736,
+      "grad_norm": 2.916352157127519,
+      "learning_rate": 3.0926312531420856e-06,
+      "loss": 0.312,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9867326287384754,
+      "grad_norm": 2.699773165705141,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.3042,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9878569822352148,
+      "grad_norm": 3.0209833912564013,
+      "learning_rate": 3.080537701795168e-06,
+      "loss": 0.3453,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9889813357319541,
+      "grad_norm": 2.641512059381042,
+      "learning_rate": 3.07449585436211e-06,
+      "loss": 0.2783,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9901056892286935,
+      "grad_norm": 2.4072661700059323,
+      "learning_rate": 3.068457306227921e-06,
+      "loss": 0.2468,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9912300427254328,
+      "grad_norm": 2.662730472010233,
+      "learning_rate": 3.0624220677394854e-06,
+      "loss": 0.2639,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9923543962221721,
+      "grad_norm": 3.129772999752526,
+      "learning_rate": 3.056390149238022e-06,
+      "loss": 0.2896,
+      "step": 1772
+    },
+    {
+      "epoch": 1.9934787497189115,
+      "grad_norm": 2.0763523761744596,
+      "learning_rate": 3.0503615610590605e-06,
+      "loss": 0.1875,
+      "step": 1773
+    },
+    {
+      "epoch": 1.994603103215651,
+      "grad_norm": 2.5603871146968413,
+      "learning_rate": 3.0443363135324167e-06,
+      "loss": 0.2483,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9957274567123904,
+      "grad_norm": 2.6840872227575834,
+      "learning_rate": 3.0383144169821944e-06,
+      "loss": 0.2953,
+      "step": 1775
+    },
+    {
+      "epoch": 1.9968518102091297,
+      "grad_norm": 2.8953745634362025,
+      "learning_rate": 3.0322958817267428e-06,
+      "loss": 0.3871,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9979761637058693,
+      "grad_norm": 2.8510250141822535,
+      "learning_rate": 3.0262807180786647e-06,
+      "loss": 0.3524,
+      "step": 1777
+    },
+    {
+      "epoch": 1.9991005172026086,
+      "grad_norm": 2.8324968047004546,
+      "learning_rate": 3.0202689363447734e-06,
+      "loss": 0.2372,
+      "step": 1778
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2667,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 596577496596480.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}