diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15554 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998871968415116,
+  "eval_steps": 500,
+  "global_step": 2216,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004512126339537507,
+      "grad_norm": 11.364033095495042,
+      "learning_rate": 2.9850746268656714e-08,
+      "loss": 1.1259,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009024252679075014,
+      "grad_norm": 13.931051354363555,
+      "learning_rate": 5.970149253731343e-08,
+      "loss": 1.0695,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013536379018612521,
+      "grad_norm": 12.591108456147174,
+      "learning_rate": 8.955223880597014e-08,
+      "loss": 1.1413,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018048505358150028,
+      "grad_norm": 13.097679442569106,
+      "learning_rate": 1.1940298507462686e-07,
+      "loss": 1.1058,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022560631697687537,
+      "grad_norm": 10.498761541717155,
+      "learning_rate": 1.4925373134328355e-07,
+      "loss": 1.1735,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027072758037225042,
+      "grad_norm": 10.84063689789702,
+      "learning_rate": 1.7910447761194027e-07,
+      "loss": 1.1406,
+      "step": 6
+    },
+    {
+      "epoch": 0.003158488437676255,
+      "grad_norm": 11.189978725198145,
+      "learning_rate": 2.08955223880597e-07,
+      "loss": 1.1035,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036097010716300056,
+      "grad_norm": 11.74878304446126,
+      "learning_rate": 2.388059701492537e-07,
+      "loss": 1.1278,
+      "step": 8
+    },
+    {
+      "epoch": 0.0040609137055837565,
+      "grad_norm": 11.451667410942289,
+      "learning_rate": 2.686567164179104e-07,
+      "loss": 1.0887,
+      "step": 9
+    },
+    {
+      "epoch": 0.0045121263395375075,
+      "grad_norm": 13.400719338755062,
+      "learning_rate": 2.985074626865671e-07,
+      "loss": 0.9145,
+      "step": 10
+    },
+    {
+      "epoch": 0.0049633389734912575,
+      "grad_norm": 12.493648403772585,
+      "learning_rate": 3.2835820895522385e-07,
+      "loss": 0.9701,
+      "step": 11
+    },
+    {
+      "epoch": 0.0054145516074450084,
+      "grad_norm": 11.383629551476218,
+      "learning_rate": 3.5820895522388055e-07,
+      "loss": 1.0357,
+      "step": 12
+    },
+    {
+      "epoch": 0.005865764241398759,
+      "grad_norm": 10.68706994675752,
+      "learning_rate": 3.880597014925373e-07,
+      "loss": 1.0357,
+      "step": 13
+    },
+    {
+      "epoch": 0.00631697687535251,
+      "grad_norm": 12.110795882375692,
+      "learning_rate": 4.17910447761194e-07,
+      "loss": 1.1691,
+      "step": 14
+    },
+    {
+      "epoch": 0.00676818950930626,
+      "grad_norm": 11.01201459317177,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 1.0552,
+      "step": 15
+    },
+    {
+      "epoch": 0.007219402143260011,
+      "grad_norm": 11.621902494318055,
+      "learning_rate": 4.776119402985074e-07,
+      "loss": 1.1426,
+      "step": 16
+    },
+    {
+      "epoch": 0.007670614777213762,
+      "grad_norm": 12.473147801613736,
+      "learning_rate": 5.074626865671642e-07,
+      "loss": 1.0951,
+      "step": 17
+    },
+    {
+      "epoch": 0.008121827411167513,
+      "grad_norm": 9.182896427775415,
+      "learning_rate": 5.373134328358208e-07,
+      "loss": 1.101,
+      "step": 18
+    },
+    {
+      "epoch": 0.008573040045121263,
+      "grad_norm": 9.270909685736212,
+      "learning_rate": 5.671641791044775e-07,
+      "loss": 1.1811,
+      "step": 19
+    },
+    {
+      "epoch": 0.009024252679075015,
+      "grad_norm": 10.991256459745955,
+      "learning_rate": 5.970149253731342e-07,
+      "loss": 1.1503,
+      "step": 20
+    },
+    {
+      "epoch": 0.009475465313028765,
+      "grad_norm": 9.255537906212872,
+      "learning_rate": 6.26865671641791e-07,
+      "loss": 0.9254,
+      "step": 21
+    },
+    {
+      "epoch": 0.009926677946982515,
+      "grad_norm": 8.790099892645406,
+      "learning_rate": 6.567164179104477e-07,
+      "loss": 1.196,
+      "step": 22
+    },
+    {
+      "epoch": 0.010377890580936267,
+      "grad_norm": 8.636133146821528,
+      "learning_rate": 6.865671641791044e-07,
+      "loss": 1.0438,
+      "step": 23
+    },
+    {
+      "epoch": 0.010829103214890017,
+      "grad_norm": 7.949499233764216,
+      "learning_rate": 7.164179104477611e-07,
+      "loss": 1.0595,
+      "step": 24
+    },
+    {
+      "epoch": 0.011280315848843767,
+      "grad_norm": 8.40491249369533,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 0.9882,
+      "step": 25
+    },
+    {
+      "epoch": 0.011731528482797519,
+      "grad_norm": 7.041925202309207,
+      "learning_rate": 7.761194029850746e-07,
+      "loss": 0.8373,
+      "step": 26
+    },
+    {
+      "epoch": 0.012182741116751269,
+      "grad_norm": 5.936172454200004,
+      "learning_rate": 8.059701492537313e-07,
+      "loss": 1.0756,
+      "step": 27
+    },
+    {
+      "epoch": 0.01263395375070502,
+      "grad_norm": 5.908249319041824,
+      "learning_rate": 8.35820895522388e-07,
+      "loss": 1.0573,
+      "step": 28
+    },
+    {
+      "epoch": 0.01308516638465877,
+      "grad_norm": 4.780207987320098,
+      "learning_rate": 8.656716417910447e-07,
+      "loss": 0.9898,
+      "step": 29
+    },
+    {
+      "epoch": 0.01353637901861252,
+      "grad_norm": 4.772574247160276,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.7025,
+      "step": 30
+    },
+    {
+      "epoch": 0.013987591652566272,
+      "grad_norm": 4.950826482077291,
+      "learning_rate": 9.253731343283582e-07,
+      "loss": 0.935,
+      "step": 31
+    },
+    {
+      "epoch": 0.014438804286520023,
+      "grad_norm": 5.647161557749149,
+      "learning_rate": 9.552238805970149e-07,
+      "loss": 0.9128,
+      "step": 32
+    },
+    {
+      "epoch": 0.014890016920473773,
+      "grad_norm": 4.248897174075301,
+      "learning_rate": 9.850746268656714e-07,
+      "loss": 0.92,
+      "step": 33
+    },
+    {
+      "epoch": 0.015341229554427524,
+      "grad_norm": 4.260585778705935,
+      "learning_rate": 1.0149253731343285e-06,
+      "loss": 0.997,
+      "step": 34
+    },
+    {
+      "epoch": 0.015792442188381276,
+      "grad_norm": 4.712117773863939,
+      "learning_rate": 1.0447761194029848e-06,
+      "loss": 0.9365,
+      "step": 35
+    },
+    {
+      "epoch": 0.016243654822335026,
+      "grad_norm": 4.558718307740576,
+      "learning_rate": 1.0746268656716416e-06,
+      "loss": 0.9952,
+      "step": 36
+    },
+    {
+      "epoch": 0.016694867456288776,
+      "grad_norm": 4.536125586817622,
+      "learning_rate": 1.1044776119402984e-06,
+      "loss": 0.9118,
+      "step": 37
+    },
+    {
+      "epoch": 0.017146080090242526,
+      "grad_norm": 4.333053203593244,
+      "learning_rate": 1.134328358208955e-06,
+      "loss": 0.9406,
+      "step": 38
+    },
+    {
+      "epoch": 0.017597292724196276,
+      "grad_norm": 3.913517979851077,
+      "learning_rate": 1.1641791044776118e-06,
+      "loss": 0.68,
+      "step": 39
+    },
+    {
+      "epoch": 0.01804850535815003,
+      "grad_norm": 3.8887693821373746,
+      "learning_rate": 1.1940298507462684e-06,
+      "loss": 0.8987,
+      "step": 40
+    },
+    {
+      "epoch": 0.01849971799210378,
+      "grad_norm": 3.7966150058507626,
+      "learning_rate": 1.2238805970149252e-06,
+      "loss": 0.8783,
+      "step": 41
+    },
+    {
+      "epoch": 0.01895093062605753,
+      "grad_norm": 3.8413216293595465,
+      "learning_rate": 1.253731343283582e-06,
+      "loss": 1.0114,
+      "step": 42
+    },
+    {
+      "epoch": 0.01940214326001128,
+      "grad_norm": 4.031064280953991,
+      "learning_rate": 1.2835820895522386e-06,
+      "loss": 0.9885,
+      "step": 43
+    },
+    {
+      "epoch": 0.01985335589396503,
+      "grad_norm": 4.3421298053507025,
+      "learning_rate": 1.3134328358208954e-06,
+      "loss": 0.7812,
+      "step": 44
+    },
+    {
+      "epoch": 0.02030456852791878,
+      "grad_norm": 3.6498986771766377,
+      "learning_rate": 1.3432835820895522e-06,
+      "loss": 0.864,
+      "step": 45
+    },
+    {
+      "epoch": 0.020755781161872534,
+      "grad_norm": 3.959983938073483,
+      "learning_rate": 1.3731343283582088e-06,
+      "loss": 1.0879,
+      "step": 46
+    },
+    {
+      "epoch": 0.021206993795826284,
+      "grad_norm": 3.897851022301326,
+      "learning_rate": 1.4029850746268656e-06,
+      "loss": 0.8353,
+      "step": 47
+    },
+    {
+      "epoch": 0.021658206429780034,
+      "grad_norm": 3.9557985042022734,
+      "learning_rate": 1.4328358208955222e-06,
+      "loss": 1.0151,
+      "step": 48
+    },
+    {
+      "epoch": 0.022109419063733784,
+      "grad_norm": 3.954727075094809,
+      "learning_rate": 1.462686567164179e-06,
+      "loss": 0.8335,
+      "step": 49
+    },
+    {
+      "epoch": 0.022560631697687534,
+      "grad_norm": 4.085425811341359,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 0.8118,
+      "step": 50
+    },
+    {
+      "epoch": 0.023011844331641287,
+      "grad_norm": 3.8699428653568715,
+      "learning_rate": 1.5223880597014924e-06,
+      "loss": 0.8204,
+      "step": 51
+    },
+    {
+      "epoch": 0.023463056965595037,
+      "grad_norm": 3.9398844157805035,
+      "learning_rate": 1.5522388059701492e-06,
+      "loss": 0.8355,
+      "step": 52
+    },
+    {
+      "epoch": 0.023914269599548788,
+      "grad_norm": 3.9157701162909917,
+      "learning_rate": 1.5820895522388058e-06,
+      "loss": 0.7559,
+      "step": 53
+    },
+    {
+      "epoch": 0.024365482233502538,
+      "grad_norm": 3.2664411221295166,
+      "learning_rate": 1.6119402985074626e-06,
+      "loss": 0.6912,
+      "step": 54
+    },
+    {
+      "epoch": 0.024816694867456288,
+      "grad_norm": 4.102566095975955,
+      "learning_rate": 1.6417910447761194e-06,
+      "loss": 0.8001,
+      "step": 55
+    },
+    {
+      "epoch": 0.02526790750141004,
+      "grad_norm": 3.984516243751416,
+      "learning_rate": 1.671641791044776e-06,
+      "loss": 0.6703,
+      "step": 56
+    },
+    {
+      "epoch": 0.02571912013536379,
+      "grad_norm": 3.282165881291406,
+      "learning_rate": 1.7014925373134328e-06,
+      "loss": 0.6232,
+      "step": 57
+    },
+    {
+      "epoch": 0.02617033276931754,
+      "grad_norm": 3.5552685115913896,
+      "learning_rate": 1.7313432835820893e-06,
+      "loss": 0.8499,
+      "step": 58
+    },
+    {
+      "epoch": 0.02662154540327129,
+      "grad_norm": 3.963753035943293,
+      "learning_rate": 1.7611940298507461e-06,
+      "loss": 0.7578,
+      "step": 59
+    },
+    {
+      "epoch": 0.02707275803722504,
+      "grad_norm": 3.9201161090299235,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 0.7639,
+      "step": 60
+    },
+    {
+      "epoch": 0.02752397067117879,
+      "grad_norm": 4.525487683658723,
+      "learning_rate": 1.8208955223880595e-06,
+      "loss": 0.7993,
+      "step": 61
+    },
+    {
+      "epoch": 0.027975183305132545,
+      "grad_norm": 3.577108451322024,
+      "learning_rate": 1.8507462686567163e-06,
+      "loss": 0.712,
+      "step": 62
+    },
+    {
+      "epoch": 0.028426395939086295,
+      "grad_norm": 3.750651576099599,
+      "learning_rate": 1.8805970149253731e-06,
+      "loss": 0.8058,
+      "step": 63
+    },
+    {
+      "epoch": 0.028877608573040045,
+      "grad_norm": 4.016735732488834,
+      "learning_rate": 1.9104477611940297e-06,
+      "loss": 0.8695,
+      "step": 64
+    },
+    {
+      "epoch": 0.029328821206993795,
+      "grad_norm": 3.274144203992301,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 0.7518,
+      "step": 65
+    },
+    {
+      "epoch": 0.029780033840947545,
+      "grad_norm": 3.8331394793119284,
+      "learning_rate": 1.970149253731343e-06,
+      "loss": 0.7467,
+      "step": 66
+    },
+    {
+      "epoch": 0.0302312464749013,
+      "grad_norm": 3.7944270550638914,
+      "learning_rate": 2e-06,
+      "loss": 0.6613,
+      "step": 67
+    },
+    {
+      "epoch": 0.03068245910885505,
+      "grad_norm": 3.9078761683052377,
+      "learning_rate": 1.9999989314450967e-06,
+      "loss": 0.6253,
+      "step": 68
+    },
+    {
+      "epoch": 0.0311336717428088,
+      "grad_norm": 3.7685631166339966,
+      "learning_rate": 1.9999957257826715e-06,
+      "loss": 0.6976,
+      "step": 69
+    },
+    {
+      "epoch": 0.03158488437676255,
+      "grad_norm": 3.5388877908849423,
+      "learning_rate": 1.9999903830195744e-06,
+      "loss": 0.6982,
+      "step": 70
+    },
+    {
+      "epoch": 0.0320360970107163,
+      "grad_norm": 3.9557829292992435,
+      "learning_rate": 1.9999829031672236e-06,
+      "loss": 0.9304,
+      "step": 71
+    },
+    {
+      "epoch": 0.03248730964467005,
+      "grad_norm": 3.4688030496000697,
+      "learning_rate": 1.9999732862416053e-06,
+      "loss": 0.7131,
+      "step": 72
+    },
+    {
+      "epoch": 0.0329385222786238,
+      "grad_norm": 3.4382929899009707,
+      "learning_rate": 1.9999615322632707e-06,
+      "loss": 0.67,
+      "step": 73
+    },
+    {
+      "epoch": 0.03338973491257755,
+      "grad_norm": 3.68765625410286,
+      "learning_rate": 1.9999476412573397e-06,
+      "loss": 0.6871,
+      "step": 74
+    },
+    {
+      "epoch": 0.0338409475465313,
+      "grad_norm": 3.9034136569075444,
+      "learning_rate": 1.999931613253499e-06,
+      "loss": 0.7368,
+      "step": 75
+    },
+    {
+      "epoch": 0.03429216018048505,
+      "grad_norm": 3.4434580645266983,
+      "learning_rate": 1.9999134482860026e-06,
+      "loss": 0.7582,
+      "step": 76
+    },
+    {
+      "epoch": 0.0347433728144388,
+      "grad_norm": 3.4427590958689827,
+      "learning_rate": 1.9998931463936704e-06,
+      "loss": 0.6917,
+      "step": 77
+    },
+    {
+      "epoch": 0.03519458544839255,
+      "grad_norm": 3.5129243585420045,
+      "learning_rate": 1.9998707076198903e-06,
+      "loss": 0.7202,
+      "step": 78
+    },
+    {
+      "epoch": 0.0356457980823463,
+      "grad_norm": 3.3097035106926334,
+      "learning_rate": 1.999846132012616e-06,
+      "loss": 0.6518,
+      "step": 79
+    },
+    {
+      "epoch": 0.03609701071630006,
+      "grad_norm": 4.01365302973293,
+      "learning_rate": 1.9998194196243685e-06,
+      "loss": 0.8061,
+      "step": 80
+    },
+    {
+      "epoch": 0.03654822335025381,
+      "grad_norm": 3.936175515147698,
+      "learning_rate": 1.999790570512235e-06,
+      "loss": 0.6833,
+      "step": 81
+    },
+    {
+      "epoch": 0.03699943598420756,
+      "grad_norm": 4.047146080445624,
+      "learning_rate": 1.9997595847378693e-06,
+      "loss": 0.5754,
+      "step": 82
+    },
+    {
+      "epoch": 0.03745064861816131,
+      "grad_norm": 4.520964419884895,
+      "learning_rate": 1.9997264623674913e-06,
+      "loss": 0.7149,
+      "step": 83
+    },
+    {
+      "epoch": 0.03790186125211506,
+      "grad_norm": 4.192870395330733,
+      "learning_rate": 1.999691203471887e-06,
+      "loss": 0.8313,
+      "step": 84
+    },
+    {
+      "epoch": 0.03835307388606881,
+      "grad_norm": 3.9624103451987347,
+      "learning_rate": 1.9996538081264093e-06,
+      "loss": 0.6514,
+      "step": 85
+    },
+    {
+      "epoch": 0.03880428652002256,
+      "grad_norm": 3.520519118359727,
+      "learning_rate": 1.9996142764109753e-06,
+      "loss": 0.7045,
+      "step": 86
+    },
+    {
+      "epoch": 0.03925549915397631,
+      "grad_norm": 3.5083592386843914,
+      "learning_rate": 1.999572608410069e-06,
+      "loss": 0.5047,
+      "step": 87
+    },
+    {
+      "epoch": 0.03970671178793006,
+      "grad_norm": 3.620127744810978,
+      "learning_rate": 1.999528804212739e-06,
+      "loss": 0.6376,
+      "step": 88
+    },
+    {
+      "epoch": 0.04015792442188381,
+      "grad_norm": 3.0483993598270906,
+      "learning_rate": 1.9994828639126008e-06,
+      "loss": 0.6099,
+      "step": 89
+    },
+    {
+      "epoch": 0.04060913705583756,
+      "grad_norm": 3.7495513331273984,
+      "learning_rate": 1.999434787607833e-06,
+      "loss": 0.7896,
+      "step": 90
+    },
+    {
+      "epoch": 0.04106034968979132,
+      "grad_norm": 3.4700395599360796,
+      "learning_rate": 1.9993845754011797e-06,
+      "loss": 0.6285,
+      "step": 91
+    },
+    {
+      "epoch": 0.04151156232374507,
+      "grad_norm": 3.131175268232334,
+      "learning_rate": 1.9993322273999505e-06,
+      "loss": 0.6492,
+      "step": 92
+    },
+    {
+      "epoch": 0.04196277495769882,
+      "grad_norm": 3.5622755072245127,
+      "learning_rate": 1.9992777437160185e-06,
+      "loss": 0.8218,
+      "step": 93
+    },
+    {
+      "epoch": 0.04241398759165257,
+      "grad_norm": 3.3564760065634744,
+      "learning_rate": 1.9992211244658214e-06,
+      "loss": 0.6667,
+      "step": 94
+    },
+    {
+      "epoch": 0.04286520022560632,
+      "grad_norm": 3.687670006527879,
+      "learning_rate": 1.999162369770361e-06,
+      "loss": 0.6048,
+      "step": 95
+    },
+    {
+      "epoch": 0.04331641285956007,
+      "grad_norm": 3.7369728445839776,
+      "learning_rate": 1.9991014797552027e-06,
+      "loss": 0.8528,
+      "step": 96
+    },
+    {
+      "epoch": 0.04376762549351382,
+      "grad_norm": 4.326301281314035,
+      "learning_rate": 1.999038454550474e-06,
+      "loss": 0.6348,
+      "step": 97
+    },
+    {
+      "epoch": 0.04421883812746757,
+      "grad_norm": 3.330654698188114,
+      "learning_rate": 1.998973294290868e-06,
+      "loss": 0.6672,
+      "step": 98
+    },
+    {
+      "epoch": 0.04467005076142132,
+      "grad_norm": 3.4395950335828465,
+      "learning_rate": 1.998905999115639e-06,
+      "loss": 0.6424,
+      "step": 99
+    },
+    {
+      "epoch": 0.04512126339537507,
+      "grad_norm": 3.4714288997142835,
+      "learning_rate": 1.998836569168603e-06,
+      "loss": 0.6887,
+      "step": 100
+    },
+    {
+      "epoch": 0.04557247602932882,
+      "grad_norm": 3.739675822036197,
+      "learning_rate": 1.998765004598141e-06,
+      "loss": 0.744,
+      "step": 101
+    },
+    {
+      "epoch": 0.046023688663282575,
+      "grad_norm": 3.5376293890493167,
+      "learning_rate": 1.998691305557194e-06,
+      "loss": 0.75,
+      "step": 102
+    },
+    {
+      "epoch": 0.046474901297236325,
+      "grad_norm": 3.9480104427554745,
+      "learning_rate": 1.9986154722032646e-06,
+      "loss": 0.594,
+      "step": 103
+    },
+    {
+      "epoch": 0.046926113931190075,
+      "grad_norm": 3.9967536358361135,
+      "learning_rate": 1.9985375046984165e-06,
+      "loss": 0.7111,
+      "step": 104
+    },
+    {
+      "epoch": 0.047377326565143825,
+      "grad_norm": 4.404083699085148,
+      "learning_rate": 1.9984574032092758e-06,
+      "loss": 0.6737,
+      "step": 105
+    },
+    {
+      "epoch": 0.047828539199097575,
+      "grad_norm": 3.9446789384138943,
+      "learning_rate": 1.998375167907028e-06,
+      "loss": 0.5581,
+      "step": 106
+    },
+    {
+      "epoch": 0.048279751833051325,
+      "grad_norm": 3.2749627513940354,
+      "learning_rate": 1.9982907989674186e-06,
+      "loss": 0.6848,
+      "step": 107
+    },
+    {
+      "epoch": 0.048730964467005075,
+      "grad_norm": 3.1745752031824503,
+      "learning_rate": 1.9982042965707534e-06,
+      "loss": 0.6993,
+      "step": 108
+    },
+    {
+      "epoch": 0.049182177100958825,
+      "grad_norm": 2.918514031480819,
+      "learning_rate": 1.9981156609018974e-06,
+      "loss": 0.5622,
+      "step": 109
+    },
+    {
+      "epoch": 0.049633389734912575,
+      "grad_norm": 3.2967345302885773,
+      "learning_rate": 1.998024892150275e-06,
+      "loss": 0.7764,
+      "step": 110
+    },
+    {
+      "epoch": 0.050084602368866325,
+      "grad_norm": 4.096076864953819,
+      "learning_rate": 1.997931990509869e-06,
+      "loss": 0.8217,
+      "step": 111
+    },
+    {
+      "epoch": 0.05053581500282008,
+      "grad_norm": 3.5086960564534073,
+      "learning_rate": 1.9978369561792203e-06,
+      "loss": 0.7334,
+      "step": 112
+    },
+    {
+      "epoch": 0.05098702763677383,
+      "grad_norm": 4.957647662303255,
+      "learning_rate": 1.997739789361428e-06,
+      "loss": 0.7043,
+      "step": 113
+    },
+    {
+      "epoch": 0.05143824027072758,
+      "grad_norm": 3.108513749586831,
+      "learning_rate": 1.9976404902641475e-06,
+      "loss": 0.5838,
+      "step": 114
+    },
+    {
+      "epoch": 0.05188945290468133,
+      "grad_norm": 4.374801020766516,
+      "learning_rate": 1.9975390590995923e-06,
+      "loss": 0.6207,
+      "step": 115
+    },
+    {
+      "epoch": 0.05234066553863508,
+      "grad_norm": 3.5847626690371746,
+      "learning_rate": 1.9974354960845323e-06,
+      "loss": 0.7108,
+      "step": 116
+    },
+    {
+      "epoch": 0.05279187817258883,
+      "grad_norm": 3.3972794164820566,
+      "learning_rate": 1.9973298014402927e-06,
+      "loss": 0.6389,
+      "step": 117
+    },
+    {
+      "epoch": 0.05324309080654258,
+      "grad_norm": 3.4765618382257686,
+      "learning_rate": 1.9972219753927547e-06,
+      "loss": 0.7736,
+      "step": 118
+    },
+    {
+      "epoch": 0.05369430344049633,
+      "grad_norm": 3.983880838580834,
+      "learning_rate": 1.997112018172354e-06,
+      "loss": 0.6782,
+      "step": 119
+    },
+    {
+      "epoch": 0.05414551607445008,
+      "grad_norm": 3.475938176625616,
+      "learning_rate": 1.9969999300140816e-06,
+      "loss": 0.6435,
+      "step": 120
+    },
+    {
+      "epoch": 0.05459672870840383,
+      "grad_norm": 3.2723738223571908,
+      "learning_rate": 1.9968857111574823e-06,
+      "loss": 0.725,
+      "step": 121
+    },
+    {
+      "epoch": 0.05504794134235758,
+      "grad_norm": 2.7837662665336684,
+      "learning_rate": 1.9967693618466537e-06,
+      "loss": 0.5725,
+      "step": 122
+    },
+    {
+      "epoch": 0.05549915397631134,
+      "grad_norm": 3.6895192194143287,
+      "learning_rate": 1.996650882330248e-06,
+      "loss": 0.6869,
+      "step": 123
+    },
+    {
+      "epoch": 0.05595036661026509,
+      "grad_norm": 3.4109605481942475,
+      "learning_rate": 1.9965302728614685e-06,
+      "loss": 0.7059,
+      "step": 124
+    },
+    {
+      "epoch": 0.05640157924421884,
+      "grad_norm": 3.1175030442363023,
+      "learning_rate": 1.9964075336980705e-06,
+      "loss": 0.5166,
+      "step": 125
+    },
+    {
+      "epoch": 0.05685279187817259,
+      "grad_norm": 3.7149554868454207,
+      "learning_rate": 1.9962826651023618e-06,
+      "loss": 0.6477,
+      "step": 126
+    },
+    {
+      "epoch": 0.05730400451212634,
+      "grad_norm": 2.7779966288953006,
+      "learning_rate": 1.9961556673412e-06,
+      "loss": 0.5851,
+      "step": 127
+    },
+    {
+      "epoch": 0.05775521714608009,
+      "grad_norm": 3.6464661606212627,
+      "learning_rate": 1.9960265406859927e-06,
+      "loss": 0.7105,
+      "step": 128
+    },
+    {
+      "epoch": 0.05820642978003384,
+      "grad_norm": 4.222181937491618,
+      "learning_rate": 1.9958952854126986e-06,
+      "loss": 0.8397,
+      "step": 129
+    },
+    {
+      "epoch": 0.05865764241398759,
+      "grad_norm": 3.463205955817649,
+      "learning_rate": 1.995761901801824e-06,
+      "loss": 0.6813,
+      "step": 130
+    },
+    {
+      "epoch": 0.05910885504794134,
+      "grad_norm": 3.5951321527152214,
+      "learning_rate": 1.995626390138425e-06,
+      "loss": 0.6965,
+      "step": 131
+    },
+    {
+      "epoch": 0.05956006768189509,
+      "grad_norm": 3.321905615889881,
+      "learning_rate": 1.995488750712104e-06,
+      "loss": 0.7583,
+      "step": 132
+    },
+    {
+      "epoch": 0.06001128031584884,
+      "grad_norm": 3.5601557448366763,
+      "learning_rate": 1.995348983817012e-06,
+      "loss": 0.595,
+      "step": 133
+    },
+    {
+      "epoch": 0.0604624929498026,
+      "grad_norm": 3.734905545964483,
+      "learning_rate": 1.9952070897518465e-06,
+      "loss": 0.6582,
+      "step": 134
+    },
+    {
+      "epoch": 0.06091370558375635,
+      "grad_norm": 3.3634272520299824,
+      "learning_rate": 1.99506306881985e-06,
+      "loss": 0.635,
+      "step": 135
+    },
+    {
+      "epoch": 0.0613649182177101,
+      "grad_norm": 3.2571407250429774,
+      "learning_rate": 1.9949169213288123e-06,
+      "loss": 0.6879,
+      "step": 136
+    },
+    {
+      "epoch": 0.06181613085166385,
+      "grad_norm": 3.402056992142215,
+      "learning_rate": 1.9947686475910653e-06,
+      "loss": 0.66,
+      "step": 137
+    },
+    {
+      "epoch": 0.0622673434856176,
+      "grad_norm": 3.230669775565966,
+      "learning_rate": 1.9946182479234867e-06,
+      "loss": 0.6797,
+      "step": 138
+    },
+    {
+      "epoch": 0.06271855611957135,
+      "grad_norm": 3.2932122097098606,
+      "learning_rate": 1.9944657226474975e-06,
+      "loss": 0.6038,
+      "step": 139
+    },
+    {
+      "epoch": 0.0631697687535251,
+      "grad_norm": 3.453841593898078,
+      "learning_rate": 1.9943110720890605e-06,
+      "loss": 0.7284,
+      "step": 140
+    },
+    {
+      "epoch": 0.06362098138747885,
+      "grad_norm": 3.4659777218616945,
+      "learning_rate": 1.994154296578681e-06,
+      "loss": 0.7008,
+      "step": 141
+    },
+    {
+      "epoch": 0.0640721940214326,
+      "grad_norm": 3.2914543179139737,
+      "learning_rate": 1.993995396451406e-06,
+      "loss": 0.679,
+      "step": 142
+    },
+    {
+      "epoch": 0.06452340665538635,
+      "grad_norm": 3.5718717851262873,
+      "learning_rate": 1.9938343720468215e-06,
+      "loss": 0.7916,
+      "step": 143
+    },
+    {
+      "epoch": 0.0649746192893401,
+      "grad_norm": 3.822827292185195,
+      "learning_rate": 1.993671223709055e-06,
+      "loss": 0.7997,
+      "step": 144
+    },
+    {
+      "epoch": 0.06542583192329385,
+      "grad_norm": 3.636935399595887,
+      "learning_rate": 1.9935059517867726e-06,
+      "loss": 0.8537,
+      "step": 145
+    },
+    {
+      "epoch": 0.0658770445572476,
+      "grad_norm": 3.6710814044757787,
+      "learning_rate": 1.993338556633178e-06,
+      "loss": 0.7573,
+      "step": 146
+    },
+    {
+      "epoch": 0.06632825719120135,
+      "grad_norm": 2.7858354684349282,
+      "learning_rate": 1.993169038606014e-06,
+      "loss": 0.5819,
+      "step": 147
+    },
+    {
+      "epoch": 0.0667794698251551,
+      "grad_norm": 3.276036115967757,
+      "learning_rate": 1.992997398067558e-06,
+      "loss": 0.574,
+      "step": 148
+    },
+    {
+      "epoch": 0.06723068245910886,
+      "grad_norm": 3.0839012680951052,
+      "learning_rate": 1.992823635384625e-06,
+      "loss": 0.524,
+      "step": 149
+    },
+    {
+      "epoch": 0.0676818950930626,
+      "grad_norm": 3.506642162753119,
+      "learning_rate": 1.9926477509285654e-06,
+      "loss": 0.8239,
+      "step": 150
+    },
+    {
+      "epoch": 0.06813310772701636,
+      "grad_norm": 3.148995922283508,
+      "learning_rate": 1.9924697450752634e-06,
+      "loss": 0.6518,
+      "step": 151
+    },
+    {
+      "epoch": 0.0685843203609701,
+      "grad_norm": 3.42401846246941,
+      "learning_rate": 1.9922896182051368e-06,
+      "loss": 0.7514,
+      "step": 152
+    },
+    {
+      "epoch": 0.06903553299492386,
+      "grad_norm": 3.418930096776447,
+      "learning_rate": 1.9921073707031367e-06,
+      "loss": 0.7616,
+      "step": 153
+    },
+    {
+      "epoch": 0.0694867456288776,
+      "grad_norm": 3.4667255988772614,
+      "learning_rate": 1.9919230029587457e-06,
+      "loss": 0.7787,
+      "step": 154
+    },
+    {
+      "epoch": 0.06993795826283136,
+      "grad_norm": 3.4038766672764527,
+      "learning_rate": 1.991736515365979e-06,
+      "loss": 0.5716,
+      "step": 155
+    },
+    {
+      "epoch": 0.0703891708967851,
+      "grad_norm": 2.9809305533819743,
+      "learning_rate": 1.99154790832338e-06,
+      "loss": 0.6286,
+      "step": 156
+    },
+    {
+      "epoch": 0.07084038353073886,
+      "grad_norm": 3.4845882856327264,
+      "learning_rate": 1.9913571822340225e-06,
+      "loss": 0.6862,
+      "step": 157
+    },
+    {
+      "epoch": 0.0712915961646926,
+      "grad_norm": 3.5834795110879507,
+      "learning_rate": 1.9911643375055103e-06,
+      "loss": 0.8095,
+      "step": 158
+    },
+    {
+      "epoch": 0.07174280879864636,
+      "grad_norm": 3.4230242670945867,
+      "learning_rate": 1.9909693745499727e-06,
+      "loss": 0.7338,
+      "step": 159
+    },
+    {
+      "epoch": 0.07219402143260012,
+      "grad_norm": 3.423862511414622,
+      "learning_rate": 1.9907722937840673e-06,
+      "loss": 0.6397,
+      "step": 160
+    },
+    {
+      "epoch": 0.07264523406655386,
+      "grad_norm": 3.2464417100716476,
+      "learning_rate": 1.990573095628977e-06,
+      "loss": 0.6437,
+      "step": 161
+    },
+    {
+      "epoch": 0.07309644670050762,
+      "grad_norm": 3.2452080634188603,
+      "learning_rate": 1.990371780510411e-06,
+      "loss": 0.7933,
+      "step": 162
+    },
+    {
+      "epoch": 0.07354765933446136,
+      "grad_norm": 3.8701961071397193,
+      "learning_rate": 1.990168348858601e-06,
+      "loss": 0.6959,
+      "step": 163
+    },
+    {
+      "epoch": 0.07399887196841512,
+      "grad_norm": 3.346386820700462,
+      "learning_rate": 1.9899628011083025e-06,
+      "loss": 0.7028,
+      "step": 164
+    },
+    {
+      "epoch": 0.07445008460236886,
+      "grad_norm": 3.3695620544061917,
+      "learning_rate": 1.9897551376987948e-06,
+      "loss": 0.7207,
+      "step": 165
+    },
+    {
+      "epoch": 0.07490129723632262,
+      "grad_norm": 3.3459806929523928,
+      "learning_rate": 1.9895453590738766e-06,
+      "loss": 0.565,
+      "step": 166
+    },
+    {
+      "epoch": 0.07535250987027636,
+      "grad_norm": 3.5190095763107907,
+      "learning_rate": 1.9893334656818678e-06,
+      "loss": 0.5682,
+      "step": 167
+    },
+    {
+      "epoch": 0.07580372250423012,
+      "grad_norm": 3.4154315855452277,
+      "learning_rate": 1.989119457975608e-06,
+      "loss": 0.7248,
+      "step": 168
+    },
+    {
+      "epoch": 0.07625493513818386,
+      "grad_norm": 3.619554148268442,
+      "learning_rate": 1.988903336412455e-06,
+      "loss": 0.7166,
+      "step": 169
+    },
+    {
+      "epoch": 0.07670614777213762,
+      "grad_norm": 3.628782657060156,
+      "learning_rate": 1.988685101454285e-06,
+      "loss": 0.6744,
+      "step": 170
+    },
+    {
+      "epoch": 0.07715736040609138,
+      "grad_norm": 3.222286302019397,
+      "learning_rate": 1.9884647535674897e-06,
+      "loss": 0.6345,
+      "step": 171
+    },
+    {
+      "epoch": 0.07760857304004512,
+      "grad_norm": 3.3817029503603746,
+      "learning_rate": 1.988242293222976e-06,
+      "loss": 0.6987,
+      "step": 172
+    },
+    {
+      "epoch": 0.07805978567399888,
+      "grad_norm": 3.192434440346879,
+      "learning_rate": 1.9880177208961674e-06,
+      "loss": 0.6651,
+      "step": 173
+    },
+    {
+      "epoch": 0.07851099830795262,
+      "grad_norm": 3.459120242900092,
+      "learning_rate": 1.9877910370669984e-06,
+      "loss": 0.6704,
+      "step": 174
+    },
+    {
+      "epoch": 0.07896221094190638,
+      "grad_norm": 3.5176326381584473,
+      "learning_rate": 1.9875622422199184e-06,
+      "loss": 0.7315,
+      "step": 175
+    },
+    {
+      "epoch": 0.07941342357586012,
+      "grad_norm": 3.42358670836462,
+      "learning_rate": 1.9873313368438856e-06,
+      "loss": 0.709,
+      "step": 176
+    },
+    {
+      "epoch": 0.07986463620981388,
+      "grad_norm": 3.514397087656355,
+      "learning_rate": 1.987098321432372e-06,
+      "loss": 0.6897,
+      "step": 177
+    },
+    {
+      "epoch": 0.08031584884376762,
+      "grad_norm": 3.9791349562876364,
+      "learning_rate": 1.9868631964833554e-06,
+      "loss": 0.6663,
+      "step": 178
+    },
+    {
+      "epoch": 0.08076706147772138,
+      "grad_norm": 3.511691191882589,
+      "learning_rate": 1.9866259624993243e-06,
+      "loss": 0.7678,
+      "step": 179
+    },
+    {
+      "epoch": 0.08121827411167512,
+      "grad_norm": 3.6035462780242082,
+      "learning_rate": 1.9863866199872745e-06,
+      "loss": 0.6732,
+      "step": 180
+    },
+    {
+      "epoch": 0.08166948674562888,
+      "grad_norm": 3.3265262825383384,
+      "learning_rate": 1.986145169458706e-06,
+      "loss": 0.6459,
+      "step": 181
+    },
+    {
+      "epoch": 0.08212069937958263,
+      "grad_norm": 3.070737136256683,
+      "learning_rate": 1.9859016114296256e-06,
+      "loss": 0.6672,
+      "step": 182
+    },
+    {
+      "epoch": 0.08257191201353638,
+      "grad_norm": 3.1293576967982006,
+      "learning_rate": 1.985655946420544e-06,
+      "loss": 0.6482,
+      "step": 183
+    },
+    {
+      "epoch": 0.08302312464749013,
+      "grad_norm": 3.2300733698316,
+      "learning_rate": 1.9854081749564737e-06,
+      "loss": 0.5774,
+      "step": 184
+    },
+    {
+      "epoch": 0.08347433728144388,
+      "grad_norm": 3.1911556869196525,
+      "learning_rate": 1.98515829756693e-06,
+      "loss": 0.5046,
+      "step": 185
+    },
+    {
+      "epoch": 0.08392554991539763,
+      "grad_norm": 3.463324413729419,
+      "learning_rate": 1.984906314785928e-06,
+      "loss": 0.6925,
+      "step": 186
+    },
+    {
+      "epoch": 0.08437676254935138,
+      "grad_norm": 3.4082698478159053,
+      "learning_rate": 1.984652227151982e-06,
+      "loss": 0.6869,
+      "step": 187
+    },
+    {
+      "epoch": 0.08482797518330513,
+      "grad_norm": 3.0232384513032526,
+      "learning_rate": 1.984396035208107e-06,
+      "loss": 0.6854,
+      "step": 188
+    },
+    {
+      "epoch": 0.08527918781725888,
+      "grad_norm": 3.405093474595666,
+      "learning_rate": 1.984137739501811e-06,
+      "loss": 0.8003,
+      "step": 189
+    },
+    {
+      "epoch": 0.08573040045121263,
+      "grad_norm": 2.8650239504287827,
+      "learning_rate": 1.983877340585102e-06,
+      "loss": 0.6247,
+      "step": 190
+    },
+    {
+      "epoch": 0.08618161308516638,
+      "grad_norm": 3.1843877890785053,
+      "learning_rate": 1.98361483901448e-06,
+      "loss": 0.6268,
+      "step": 191
+    },
+    {
+      "epoch": 0.08663282571912014,
+      "grad_norm": 3.2905100987787574,
+      "learning_rate": 1.983350235350941e-06,
+      "loss": 0.6307,
+      "step": 192
+    },
+    {
+      "epoch": 0.08708403835307389,
+      "grad_norm": 3.6238689522296896,
+      "learning_rate": 1.9830835301599705e-06,
+      "loss": 0.6972,
+      "step": 193
+    },
+    {
+      "epoch": 0.08753525098702764,
+      "grad_norm": 3.6407089349858994,
+      "learning_rate": 1.982814724011548e-06,
+      "loss": 0.5048,
+      "step": 194
+    },
+    {
+      "epoch": 0.08798646362098139,
+      "grad_norm": 3.414185881113126,
+      "learning_rate": 1.982543817480141e-06,
+      "loss": 0.6903,
+      "step": 195
+    },
+    {
+      "epoch": 0.08843767625493514,
+      "grad_norm": 3.5617662518759423,
+      "learning_rate": 1.9822708111447073e-06,
+      "loss": 0.7267,
+      "step": 196
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 3.260537631153434,
+      "learning_rate": 1.9819957055886903e-06,
+      "loss": 0.6019,
+      "step": 197
+    },
+    {
+      "epoch": 0.08934010152284264,
+      "grad_norm": 3.9131056691786648,
+      "learning_rate": 1.9817185014000216e-06,
+      "loss": 0.7248,
+      "step": 198
+    },
+    {
+      "epoch": 0.08979131415679639,
+      "grad_norm": 3.2158505559140087,
+      "learning_rate": 1.981439199171117e-06,
+      "loss": 0.4917,
+      "step": 199
+    },
+    {
+      "epoch": 0.09024252679075014,
+      "grad_norm": 3.3742251974916306,
+      "learning_rate": 1.9811577994988754e-06,
+      "loss": 0.6419,
+      "step": 200
+    },
+    {
+      "epoch": 0.09069373942470389,
+      "grad_norm": 3.960767715416477,
+      "learning_rate": 1.9808743029846793e-06,
+      "loss": 0.6485,
+      "step": 201
+    },
+    {
+      "epoch": 0.09114495205865764,
+      "grad_norm": 3.2106160146208556,
+      "learning_rate": 1.980588710234392e-06,
+      "loss": 0.6513,
+      "step": 202
+    },
+    {
+      "epoch": 0.09159616469261139,
+      "grad_norm": 3.3168520724652435,
+      "learning_rate": 1.980301021858356e-06,
+      "loss": 0.6639,
+      "step": 203
+    },
+    {
+      "epoch": 0.09204737732656515,
+      "grad_norm": 3.3182631173626,
+      "learning_rate": 1.9800112384713937e-06,
+      "loss": 0.6178,
+      "step": 204
+    },
+    {
+      "epoch": 0.09249858996051889,
+      "grad_norm": 3.162501767023835,
+      "learning_rate": 1.9797193606928037e-06,
+      "loss": 0.6681,
+      "step": 205
+    },
+    {
+      "epoch": 0.09294980259447265,
+      "grad_norm": 3.2385347157247955,
+      "learning_rate": 1.9794253891463602e-06,
+      "loss": 0.5039,
+      "step": 206
+    },
+    {
+      "epoch": 0.09340101522842639,
+      "grad_norm": 3.573031152116351,
+      "learning_rate": 1.979129324460314e-06,
+      "loss": 0.6903,
+      "step": 207
+    },
+    {
+      "epoch": 0.09385222786238015,
+      "grad_norm": 3.601364277364461,
+      "learning_rate": 1.978831167267387e-06,
+      "loss": 0.5897,
+      "step": 208
+    },
+    {
+      "epoch": 0.09430344049633389,
+      "grad_norm": 3.2791769809356537,
+      "learning_rate": 1.9785309182047735e-06,
+      "loss": 0.5525,
+      "step": 209
+    },
+    {
+      "epoch": 0.09475465313028765,
+      "grad_norm": 4.31342038746554,
+      "learning_rate": 1.9782285779141393e-06,
+      "loss": 0.5973,
+      "step": 210
+    },
+    {
+      "epoch": 0.09520586576424139,
+      "grad_norm": 3.196166578673568,
+      "learning_rate": 1.977924147041619e-06,
+      "loss": 0.662,
+      "step": 211
+    },
+    {
+      "epoch": 0.09565707839819515,
+      "grad_norm": 3.5134072643179586,
+      "learning_rate": 1.9776176262378144e-06,
+      "loss": 0.6167,
+      "step": 212
+    },
+    {
+      "epoch": 0.09610829103214891,
+      "grad_norm": 3.4425320121595977,
+      "learning_rate": 1.977309016157794e-06,
+      "loss": 0.595,
+      "step": 213
+    },
+    {
+      "epoch": 0.09655950366610265,
+      "grad_norm": 3.374754780954467,
+      "learning_rate": 1.9769983174610917e-06,
+      "loss": 0.5101,
+      "step": 214
+    },
+    {
+      "epoch": 0.09701071630005641,
+      "grad_norm": 3.107648208296178,
+      "learning_rate": 1.9766855308117048e-06,
+      "loss": 0.705,
+      "step": 215
+    },
+    {
+      "epoch": 0.09746192893401015,
+      "grad_norm": 3.507381249990775,
+      "learning_rate": 1.9763706568780925e-06,
+      "loss": 0.6518,
+      "step": 216
+    },
+    {
+      "epoch": 0.09791314156796391,
+      "grad_norm": 3.178375092239186,
+      "learning_rate": 1.9760536963331747e-06,
+      "loss": 0.8446,
+      "step": 217
+    },
+    {
+      "epoch": 0.09836435420191765,
+      "grad_norm": 3.361779236537557,
+      "learning_rate": 1.9757346498543316e-06,
+      "loss": 0.5807,
+      "step": 218
+    },
+    {
+      "epoch": 0.09881556683587141,
+      "grad_norm": 3.859132201884954,
+      "learning_rate": 1.9754135181234003e-06,
+      "loss": 0.6968,
+      "step": 219
+    },
+    {
+      "epoch": 0.09926677946982515,
+      "grad_norm": 3.300316326276715,
+      "learning_rate": 1.9750903018266743e-06,
+      "loss": 0.7015,
+      "step": 220
+    },
+    {
+      "epoch": 0.09971799210377891,
+      "grad_norm": 3.149000744047261,
+      "learning_rate": 1.9747650016549027e-06,
+      "loss": 0.6008,
+      "step": 221
+    },
+    {
+      "epoch": 0.10016920473773265,
+      "grad_norm": 3.4892072583658695,
+      "learning_rate": 1.9744376183032873e-06,
+      "loss": 0.6077,
+      "step": 222
+    },
+    {
+      "epoch": 0.10062041737168641,
+      "grad_norm": 3.6046339473790323,
+      "learning_rate": 1.9741081524714825e-06,
+      "loss": 0.6846,
+      "step": 223
+    },
+    {
+      "epoch": 0.10107163000564016,
+      "grad_norm": 3.4520154967728085,
+      "learning_rate": 1.9737766048635928e-06,
+      "loss": 0.6671,
+      "step": 224
+    },
+    {
+      "epoch": 0.10152284263959391,
+      "grad_norm": 3.2952487774101797,
+      "learning_rate": 1.973442976188172e-06,
+      "loss": 0.6594,
+      "step": 225
+    },
+    {
+      "epoch": 0.10197405527354766,
+      "grad_norm": 3.2514684073236584,
+      "learning_rate": 1.973107267158221e-06,
+      "loss": 0.6518,
+      "step": 226
+    },
+    {
+      "epoch": 0.10242526790750141,
+      "grad_norm": 3.266600653278959,
+      "learning_rate": 1.9727694784911876e-06,
+      "loss": 0.6731,
+      "step": 227
+    },
+    {
+      "epoch": 0.10287648054145516,
+      "grad_norm": 3.1924596316986533,
+      "learning_rate": 1.972429610908962e-06,
+      "loss": 0.6636,
+      "step": 228
+    },
+    {
+      "epoch": 0.10332769317540891,
+      "grad_norm": 3.2104025132138623,
+      "learning_rate": 1.9720876651378794e-06,
+      "loss": 0.6779,
+      "step": 229
+    },
+    {
+      "epoch": 0.10377890580936266,
+      "grad_norm": 3.2963718496191294,
+      "learning_rate": 1.9717436419087155e-06,
+      "loss": 0.634,
+      "step": 230
+    },
+    {
+      "epoch": 0.10423011844331641,
+      "grad_norm": 3.2073822291013254,
+      "learning_rate": 1.9713975419566858e-06,
+      "loss": 0.5812,
+      "step": 231
+    },
+    {
+      "epoch": 0.10468133107727016,
+      "grad_norm": 3.4189730159980884,
+      "learning_rate": 1.971049366021443e-06,
+      "loss": 0.7316,
+      "step": 232
+    },
+    {
+      "epoch": 0.10513254371122391,
+      "grad_norm": 3.23394924091578,
+      "learning_rate": 1.9706991148470783e-06,
+      "loss": 0.6624,
+      "step": 233
+    },
+    {
+      "epoch": 0.10558375634517767,
+      "grad_norm": 3.1317986475999375,
+      "learning_rate": 1.970346789182116e-06,
+      "loss": 0.5124,
+      "step": 234
+    },
+    {
+      "epoch": 0.10603496897913142,
+      "grad_norm": 3.385294259471175,
+      "learning_rate": 1.969992389779516e-06,
+      "loss": 0.5569,
+      "step": 235
+    },
+    {
+      "epoch": 0.10648618161308517,
+      "grad_norm": 4.118491129332631,
+      "learning_rate": 1.9696359173966676e-06,
+      "loss": 0.6456,
+      "step": 236
+    },
+    {
+      "epoch": 0.10693739424703892,
+      "grad_norm": 3.385938350325588,
+      "learning_rate": 1.9692773727953923e-06,
+      "loss": 0.6794,
+      "step": 237
+    },
+    {
+      "epoch": 0.10738860688099267,
+      "grad_norm": 3.401423714020342,
+      "learning_rate": 1.9689167567419383e-06,
+      "loss": 0.6555,
+      "step": 238
+    },
+    {
+      "epoch": 0.10783981951494642,
+      "grad_norm": 3.401519447295867,
+      "learning_rate": 1.9685540700069827e-06,
+      "loss": 0.578,
+      "step": 239
+    },
+    {
+      "epoch": 0.10829103214890017,
+      "grad_norm": 4.10708011535945,
+      "learning_rate": 1.9681893133656257e-06,
+      "loss": 0.7618,
+      "step": 240
+    },
+    {
+      "epoch": 0.10874224478285392,
+      "grad_norm": 3.5017274039945163,
+      "learning_rate": 1.9678224875973932e-06,
+      "loss": 0.592,
+      "step": 241
+    },
+    {
+      "epoch": 0.10919345741680767,
+      "grad_norm": 3.1688095237487017,
+      "learning_rate": 1.9674535934862324e-06,
+      "loss": 0.6643,
+      "step": 242
+    },
+    {
+      "epoch": 0.10964467005076142,
+      "grad_norm": 3.1685960960934336,
+      "learning_rate": 1.9670826318205098e-06,
+      "loss": 0.6199,
+      "step": 243
+    },
+    {
+      "epoch": 0.11009588268471517,
+      "grad_norm": 3.4775317850948104,
+      "learning_rate": 1.9667096033930114e-06,
+      "loss": 0.6675,
+      "step": 244
+    },
+    {
+      "epoch": 0.11054709531866892,
+      "grad_norm": 3.643632342788612,
+      "learning_rate": 1.96633450900094e-06,
+      "loss": 0.711,
+      "step": 245
+    },
+    {
+      "epoch": 0.11099830795262268,
+      "grad_norm": 2.9819737147125003,
+      "learning_rate": 1.965957349445914e-06,
+      "loss": 0.6962,
+      "step": 246
+    },
+    {
+      "epoch": 0.11144952058657642,
+      "grad_norm": 2.9768216888435477,
+      "learning_rate": 1.9655781255339632e-06,
+      "loss": 0.65,
+      "step": 247
+    },
+    {
+      "epoch": 0.11190073322053018,
+      "grad_norm": 3.174224366971097,
+      "learning_rate": 1.965196838075533e-06,
+      "loss": 0.5309,
+      "step": 248
+    },
+    {
+      "epoch": 0.11235194585448392,
+      "grad_norm": 3.0582374952193985,
+      "learning_rate": 1.9648134878854744e-06,
+      "loss": 0.5635,
+      "step": 249
+    },
+    {
+      "epoch": 0.11280315848843768,
+      "grad_norm": 3.8277212668333114,
+      "learning_rate": 1.9644280757830507e-06,
+      "loss": 0.7526,
+      "step": 250
+    },
+    {
+      "epoch": 0.11325437112239142,
+      "grad_norm": 3.8709397850662945,
+      "learning_rate": 1.9640406025919285e-06,
+      "loss": 0.6035,
+      "step": 251
+    },
+    {
+      "epoch": 0.11370558375634518,
+      "grad_norm": 3.3137132906235203,
+      "learning_rate": 1.963651069140181e-06,
+      "loss": 0.6656,
+      "step": 252
+    },
+    {
+      "epoch": 0.11415679639029892,
+      "grad_norm": 3.574058474566791,
+      "learning_rate": 1.963259476260284e-06,
+      "loss": 0.7095,
+      "step": 253
+    },
+    {
+      "epoch": 0.11460800902425268,
+      "grad_norm": 3.6146347899303115,
+      "learning_rate": 1.962865824789115e-06,
+      "loss": 0.6267,
+      "step": 254
+    },
+    {
+      "epoch": 0.11505922165820642,
+      "grad_norm": 4.082630583384243,
+      "learning_rate": 1.96247011556795e-06,
+      "loss": 0.7022,
+      "step": 255
+    },
+    {
+      "epoch": 0.11551043429216018,
+      "grad_norm": 3.291509482076301,
+      "learning_rate": 1.9620723494424623e-06,
+      "loss": 0.5721,
+      "step": 256
+    },
+    {
+      "epoch": 0.11596164692611394,
+      "grad_norm": 3.58368271064381,
+      "learning_rate": 1.961672527262723e-06,
+      "loss": 0.6236,
+      "step": 257
+    },
+    {
+      "epoch": 0.11641285956006768,
+      "grad_norm": 3.1224675347563275,
+      "learning_rate": 1.9612706498831957e-06,
+      "loss": 0.5134,
+      "step": 258
+    },
+    {
+      "epoch": 0.11686407219402144,
+      "grad_norm": 3.5167852362721654,
+      "learning_rate": 1.9608667181627357e-06,
+      "loss": 0.6986,
+      "step": 259
+    },
+    {
+      "epoch": 0.11731528482797518,
+      "grad_norm": 3.5809347550436277,
+      "learning_rate": 1.96046073296459e-06,
+      "loss": 0.6251,
+      "step": 260
+    },
+    {
+      "epoch": 0.11776649746192894,
+      "grad_norm": 2.8641754562099244,
+      "learning_rate": 1.9600526951563937e-06,
+      "loss": 0.6369,
+      "step": 261
+    },
+    {
+      "epoch": 0.11821771009588268,
+      "grad_norm": 3.613739867174843,
+      "learning_rate": 1.9596426056101684e-06,
+      "loss": 0.786,
+      "step": 262
+    },
+    {
+      "epoch": 0.11866892272983644,
+      "grad_norm": 3.580296089646273,
+      "learning_rate": 1.9592304652023203e-06,
+      "loss": 0.6038,
+      "step": 263
+    },
+    {
+      "epoch": 0.11912013536379018,
+      "grad_norm": 3.15842598534144,
+      "learning_rate": 1.958816274813639e-06,
+      "loss": 0.5811,
+      "step": 264
+    },
+    {
+      "epoch": 0.11957134799774394,
+      "grad_norm": 3.0081075803485247,
+      "learning_rate": 1.958400035329294e-06,
+      "loss": 0.528,
+      "step": 265
+    },
+    {
+      "epoch": 0.12002256063169768,
+      "grad_norm": 2.8554740174867606,
+      "learning_rate": 1.9579817476388357e-06,
+      "loss": 0.625,
+      "step": 266
+    },
+    {
+      "epoch": 0.12047377326565144,
+      "grad_norm": 3.3849924046760447,
+      "learning_rate": 1.9575614126361907e-06,
+      "loss": 0.5894,
+      "step": 267
+    },
+    {
+      "epoch": 0.1209249858996052,
+      "grad_norm": 2.8658227520314994,
+      "learning_rate": 1.9571390312196607e-06,
+      "loss": 0.6311,
+      "step": 268
+    },
+    {
+      "epoch": 0.12137619853355894,
+      "grad_norm": 3.22918688434795,
+      "learning_rate": 1.9567146042919213e-06,
+      "loss": 0.6711,
+      "step": 269
+    },
+    {
+      "epoch": 0.1218274111675127,
+      "grad_norm": 3.102667861242874,
+      "learning_rate": 1.9562881327600197e-06,
+      "loss": 0.5661,
+      "step": 270
+    },
+    {
+      "epoch": 0.12227862380146644,
+      "grad_norm": 2.9951718210029843,
+      "learning_rate": 1.955859617535372e-06,
+      "loss": 0.7687,
+      "step": 271
+    },
+    {
+      "epoch": 0.1227298364354202,
+      "grad_norm": 3.1684228477861884,
+      "learning_rate": 1.9554290595337625e-06,
+      "loss": 0.6228,
+      "step": 272
+    },
+    {
+      "epoch": 0.12318104906937394,
+      "grad_norm": 3.393968292035563,
+      "learning_rate": 1.954996459675341e-06,
+      "loss": 0.5848,
+      "step": 273
+    },
+    {
+      "epoch": 0.1236322617033277,
+      "grad_norm": 2.852079323030113,
+      "learning_rate": 1.9545618188846205e-06,
+      "loss": 0.5769,
+      "step": 274
+    },
+    {
+      "epoch": 0.12408347433728144,
+      "grad_norm": 2.9641384401198074,
+      "learning_rate": 1.9541251380904762e-06,
+      "loss": 0.5864,
+      "step": 275
+    },
+    {
+      "epoch": 0.1245346869712352,
+      "grad_norm": 3.3554476588870568,
+      "learning_rate": 1.9536864182261435e-06,
+      "loss": 0.657,
+      "step": 276
+    },
+    {
+      "epoch": 0.12498589960518895,
+      "grad_norm": 3.281618043698293,
+      "learning_rate": 1.9532456602292146e-06,
+      "loss": 0.6213,
+      "step": 277
+    },
+    {
+      "epoch": 0.1254371122391427,
+      "grad_norm": 3.3143546652497644,
+      "learning_rate": 1.9528028650416375e-06,
+      "loss": 0.7062,
+      "step": 278
+    },
+    {
+      "epoch": 0.12588832487309645,
+      "grad_norm": 3.3561372317938134,
+      "learning_rate": 1.9523580336097145e-06,
+      "loss": 0.4876,
+      "step": 279
+    },
+    {
+      "epoch": 0.1263395375070502,
+      "grad_norm": 3.0109523300134264,
+      "learning_rate": 1.9519111668840982e-06,
+      "loss": 0.5256,
+      "step": 280
+    },
+    {
+      "epoch": 0.12679075014100394,
+      "grad_norm": 3.578806428790207,
+      "learning_rate": 1.9514622658197933e-06,
+      "loss": 0.5089,
+      "step": 281
+    },
+    {
+      "epoch": 0.1272419627749577,
+      "grad_norm": 3.074575948566442,
+      "learning_rate": 1.95101133137615e-06,
+      "loss": 0.695,
+      "step": 282
+    },
+    {
+      "epoch": 0.12769317540891145,
+      "grad_norm": 3.1027160235638984,
+      "learning_rate": 1.950558364516865e-06,
+      "loss": 0.5853,
+      "step": 283
+    },
+    {
+      "epoch": 0.1281443880428652,
+      "grad_norm": 3.3099951191328563,
+      "learning_rate": 1.9501033662099777e-06,
+      "loss": 0.5303,
+      "step": 284
+    },
+    {
+      "epoch": 0.12859560067681894,
+      "grad_norm": 3.446711214585093,
+      "learning_rate": 1.9496463374278696e-06,
+      "loss": 0.5828,
+      "step": 285
+    },
+    {
+      "epoch": 0.1290468133107727,
+      "grad_norm": 3.263902821177112,
+      "learning_rate": 1.949187279147262e-06,
+      "loss": 0.6326,
+      "step": 286
+    },
+    {
+      "epoch": 0.12949802594472645,
+      "grad_norm": 3.571231287506934,
+      "learning_rate": 1.9487261923492116e-06,
+      "loss": 0.6723,
+      "step": 287
+    },
+    {
+      "epoch": 0.1299492385786802,
+      "grad_norm": 3.001117794130769,
+      "learning_rate": 1.9482630780191126e-06,
+      "loss": 0.5732,
+      "step": 288
+    },
+    {
+      "epoch": 0.13040045121263397,
+      "grad_norm": 3.474015517628523,
+      "learning_rate": 1.947797937146691e-06,
+      "loss": 0.6214,
+      "step": 289
+    },
+    {
+      "epoch": 0.1308516638465877,
+      "grad_norm": 3.2034938171083955,
+      "learning_rate": 1.947330770726004e-06,
+      "loss": 0.7137,
+      "step": 290
+    },
+    {
+      "epoch": 0.13130287648054145,
+      "grad_norm": 4.0987214667589935,
+      "learning_rate": 1.946861579755437e-06,
+      "loss": 0.6807,
+      "step": 291
+    },
+    {
+      "epoch": 0.1317540891144952,
+      "grad_norm": 2.9958320731805355,
+      "learning_rate": 1.9463903652377026e-06,
+      "loss": 0.648,
+      "step": 292
+    },
+    {
+      "epoch": 0.13220530174844897,
+      "grad_norm": 3.5326747287868274,
+      "learning_rate": 1.945917128179839e-06,
+      "loss": 0.5305,
+      "step": 293
+    },
+    {
+      "epoch": 0.1326565143824027,
+      "grad_norm": 3.087656746813269,
+      "learning_rate": 1.9454418695932045e-06,
+      "loss": 0.6877,
+      "step": 294
+    },
+    {
+      "epoch": 0.13310772701635645,
+      "grad_norm": 3.161562230483419,
+      "learning_rate": 1.94496459049348e-06,
+      "loss": 0.7082,
+      "step": 295
+    },
+    {
+      "epoch": 0.1335589396503102,
+      "grad_norm": 3.1405430418827236,
+      "learning_rate": 1.9444852919006623e-06,
+      "loss": 0.6361,
+      "step": 296
+    },
+    {
+      "epoch": 0.13401015228426397,
+      "grad_norm": 3.2609308388363165,
+      "learning_rate": 1.944003974839066e-06,
+      "loss": 0.4959,
+      "step": 297
+    },
+    {
+      "epoch": 0.13446136491821772,
+      "grad_norm": 3.136461389328533,
+      "learning_rate": 1.9435206403373178e-06,
+      "loss": 0.7346,
+      "step": 298
+    },
+    {
+      "epoch": 0.13491257755217145,
+      "grad_norm": 3.5274417741499198,
+      "learning_rate": 1.9430352894283567e-06,
+      "loss": 0.5936,
+      "step": 299
+    },
+    {
+      "epoch": 0.1353637901861252,
+      "grad_norm": 3.7645608637306647,
+      "learning_rate": 1.9425479231494318e-06,
+      "loss": 0.6398,
+      "step": 300
+    },
+    {
+      "epoch": 0.13581500282007897,
+      "grad_norm": 3.1396546903591718,
+      "learning_rate": 1.942058542542097e-06,
+      "loss": 0.5478,
+      "step": 301
+    },
+    {
+      "epoch": 0.13626621545403272,
+      "grad_norm": 3.594716193105851,
+      "learning_rate": 1.9415671486522137e-06,
+      "loss": 0.7303,
+      "step": 302
+    },
+    {
+      "epoch": 0.13671742808798645,
+      "grad_norm": 3.507738272824508,
+      "learning_rate": 1.9410737425299434e-06,
+      "loss": 0.6789,
+      "step": 303
+    },
+    {
+      "epoch": 0.1371686407219402,
+      "grad_norm": 3.118749146822311,
+      "learning_rate": 1.94057832522975e-06,
+      "loss": 0.6171,
+      "step": 304
+    },
+    {
+      "epoch": 0.13761985335589397,
+      "grad_norm": 3.210316055229788,
+      "learning_rate": 1.9400808978103944e-06,
+      "loss": 0.5672,
+      "step": 305
+    },
+    {
+      "epoch": 0.13807106598984772,
+      "grad_norm": 3.0272573921584374,
+      "learning_rate": 1.9395814613349338e-06,
+      "loss": 0.6,
+      "step": 306
+    },
+    {
+      "epoch": 0.13852227862380145,
+      "grad_norm": 2.8900194965740487,
+      "learning_rate": 1.9390800168707182e-06,
+      "loss": 0.6387,
+      "step": 307
+    },
+    {
+      "epoch": 0.1389734912577552,
+      "grad_norm": 3.0767885306822738,
+      "learning_rate": 1.93857656548939e-06,
+      "loss": 0.6114,
+      "step": 308
+    },
+    {
+      "epoch": 0.13942470389170897,
+      "grad_norm": 3.1091627287910777,
+      "learning_rate": 1.93807110826688e-06,
+      "loss": 0.5469,
+      "step": 309
+    },
+    {
+      "epoch": 0.13987591652566272,
+      "grad_norm": 3.5058309066962976,
+      "learning_rate": 1.937563646283406e-06,
+      "loss": 0.5231,
+      "step": 310
+    },
+    {
+      "epoch": 0.14032712915961648,
+      "grad_norm": 3.381960145855753,
+      "learning_rate": 1.93705418062347e-06,
+      "loss": 0.6071,
+      "step": 311
+    },
+    {
+      "epoch": 0.1407783417935702,
+      "grad_norm": 3.3900203190115707,
+      "learning_rate": 1.9365427123758547e-06,
+      "loss": 0.6137,
+      "step": 312
+    },
+    {
+      "epoch": 0.14122955442752397,
+      "grad_norm": 3.1063379022563007,
+      "learning_rate": 1.936029242633626e-06,
+      "loss": 0.5899,
+      "step": 313
+    },
+    {
+      "epoch": 0.14168076706147772,
+      "grad_norm": 3.6211320331970343,
+      "learning_rate": 1.9355137724941234e-06,
+      "loss": 0.6263,
+      "step": 314
+    },
+    {
+      "epoch": 0.14213197969543148,
+      "grad_norm": 3.1680939515915902,
+      "learning_rate": 1.9349963030589644e-06,
+      "loss": 0.6648,
+      "step": 315
+    },
+    {
+      "epoch": 0.1425831923293852,
+      "grad_norm": 3.5195879110562562,
+      "learning_rate": 1.9344768354340377e-06,
+      "loss": 0.6575,
+      "step": 316
+    },
+    {
+      "epoch": 0.14303440496333897,
+      "grad_norm": 3.1488331602409176,
+      "learning_rate": 1.9339553707295018e-06,
+      "loss": 0.5313,
+      "step": 317
+    },
+    {
+      "epoch": 0.14348561759729273,
+      "grad_norm": 3.2151330395191557,
+      "learning_rate": 1.933431910059785e-06,
+      "loss": 0.528,
+      "step": 318
+    },
+    {
+      "epoch": 0.14393683023124648,
+      "grad_norm": 3.132260724021693,
+      "learning_rate": 1.93290645454358e-06,
+      "loss": 0.5815,
+      "step": 319
+    },
+    {
+      "epoch": 0.14438804286520024,
+      "grad_norm": 3.622222391906822,
+      "learning_rate": 1.9323790053038433e-06,
+      "loss": 0.7457,
+      "step": 320
+    },
+    {
+      "epoch": 0.14483925549915397,
+      "grad_norm": 4.223750697040322,
+      "learning_rate": 1.9318495634677907e-06,
+      "loss": 0.5711,
+      "step": 321
+    },
+    {
+      "epoch": 0.14529046813310773,
+      "grad_norm": 3.3650649933245376,
+      "learning_rate": 1.9313181301668985e-06,
+      "loss": 0.65,
+      "step": 322
+    },
+    {
+      "epoch": 0.14574168076706148,
+      "grad_norm": 3.7805123756041934,
+      "learning_rate": 1.9307847065368978e-06,
+      "loss": 0.4404,
+      "step": 323
+    },
+    {
+      "epoch": 0.14619289340101524,
+      "grad_norm": 3.6691078891308466,
+      "learning_rate": 1.9302492937177733e-06,
+      "loss": 0.5287,
+      "step": 324
+    },
+    {
+      "epoch": 0.14664410603496897,
+      "grad_norm": 2.9203910805909925,
+      "learning_rate": 1.9297118928537616e-06,
+      "loss": 0.5946,
+      "step": 325
+    },
+    {
+      "epoch": 0.14709531866892273,
+      "grad_norm": 3.4322784585242094,
+      "learning_rate": 1.9291725050933466e-06,
+      "loss": 0.5945,
+      "step": 326
+    },
+    {
+      "epoch": 0.14754653130287648,
+      "grad_norm": 3.4949175306707487,
+      "learning_rate": 1.9286311315892592e-06,
+      "loss": 0.6565,
+      "step": 327
+    },
+    {
+      "epoch": 0.14799774393683024,
+      "grad_norm": 3.3116625435966855,
+      "learning_rate": 1.9280877734984745e-06,
+      "loss": 0.5587,
+      "step": 328
+    },
+    {
+      "epoch": 0.14844895657078397,
+      "grad_norm": 3.9017874847497596,
+      "learning_rate": 1.9275424319822084e-06,
+      "loss": 0.6092,
+      "step": 329
+    },
+    {
+      "epoch": 0.14890016920473773,
+      "grad_norm": 3.483226347899875,
+      "learning_rate": 1.926995108205915e-06,
+      "loss": 0.661,
+      "step": 330
+    },
+    {
+      "epoch": 0.14935138183869148,
+      "grad_norm": 3.358115279938035,
+      "learning_rate": 1.926445803339286e-06,
+      "loss": 0.588,
+      "step": 331
+    },
+    {
+      "epoch": 0.14980259447264524,
+      "grad_norm": 3.6543876138753446,
+      "learning_rate": 1.925894518556246e-06,
+      "loss": 0.6363,
+      "step": 332
+    },
+    {
+      "epoch": 0.150253807106599,
+      "grad_norm": 3.5803877393350514,
+      "learning_rate": 1.9253412550349505e-06,
+      "loss": 0.6382,
+      "step": 333
+    },
+    {
+      "epoch": 0.15070501974055273,
+      "grad_norm": 3.514105036732028,
+      "learning_rate": 1.9247860139577852e-06,
+      "loss": 0.531,
+      "step": 334
+    },
+    {
+      "epoch": 0.15115623237450648,
+      "grad_norm": 3.0800678193550617,
+      "learning_rate": 1.924228796511361e-06,
+      "loss": 0.5065,
+      "step": 335
+    },
+    {
+      "epoch": 0.15160744500846024,
+      "grad_norm": 3.431976731385007,
+      "learning_rate": 1.923669603886513e-06,
+      "loss": 0.6681,
+      "step": 336
+    },
+    {
+      "epoch": 0.152058657642414,
+      "grad_norm": 3.1591532382990493,
+      "learning_rate": 1.9231084372782968e-06,
+      "loss": 0.5994,
+      "step": 337
+    },
+    {
+      "epoch": 0.15250987027636773,
+      "grad_norm": 2.8404865058044613,
+      "learning_rate": 1.9225452978859873e-06,
+      "loss": 0.5612,
+      "step": 338
+    },
+    {
+      "epoch": 0.15296108291032148,
+      "grad_norm": 3.326070609318879,
+      "learning_rate": 1.921980186913075e-06,
+      "loss": 0.5316,
+      "step": 339
+    },
+    {
+      "epoch": 0.15341229554427524,
+      "grad_norm": 3.5187885378329886,
+      "learning_rate": 1.9214131055672642e-06,
+      "loss": 0.7119,
+      "step": 340
+    },
+    {
+      "epoch": 0.153863508178229,
+      "grad_norm": 3.544854067555435,
+      "learning_rate": 1.9208440550604702e-06,
+      "loss": 0.7238,
+      "step": 341
+    },
+    {
+      "epoch": 0.15431472081218275,
+      "grad_norm": 3.2845637387346267,
+      "learning_rate": 1.9202730366088164e-06,
+      "loss": 0.6078,
+      "step": 342
+    },
+    {
+      "epoch": 0.15476593344613648,
+      "grad_norm": 3.3136242560590294,
+      "learning_rate": 1.9197000514326317e-06,
+      "loss": 0.5787,
+      "step": 343
+    },
+    {
+      "epoch": 0.15521714608009024,
+      "grad_norm": 3.284411489524786,
+      "learning_rate": 1.9191251007564487e-06,
+      "loss": 0.5574,
+      "step": 344
+    },
+    {
+      "epoch": 0.155668358714044,
+      "grad_norm": 3.6355673133340747,
+      "learning_rate": 1.9185481858089996e-06,
+      "loss": 0.7101,
+      "step": 345
+    },
+    {
+      "epoch": 0.15611957134799775,
+      "grad_norm": 4.128016030092157,
+      "learning_rate": 1.9179693078232155e-06,
+      "loss": 0.5079,
+      "step": 346
+    },
+    {
+      "epoch": 0.15657078398195148,
+      "grad_norm": 3.0530977440321405,
+      "learning_rate": 1.917388468036222e-06,
+      "loss": 0.5866,
+      "step": 347
+    },
+    {
+      "epoch": 0.15702199661590524,
+      "grad_norm": 3.165514172841862,
+      "learning_rate": 1.9168056676893374e-06,
+      "loss": 0.6284,
+      "step": 348
+    },
+    {
+      "epoch": 0.157473209249859,
+      "grad_norm": 2.995277805594194,
+      "learning_rate": 1.91622090802807e-06,
+      "loss": 0.6062,
+      "step": 349
+    },
+    {
+      "epoch": 0.15792442188381275,
+      "grad_norm": 3.005909595695992,
+      "learning_rate": 1.9156341903021155e-06,
+      "loss": 0.6028,
+      "step": 350
+    },
+    {
+      "epoch": 0.15837563451776648,
+      "grad_norm": 3.721941326683315,
+      "learning_rate": 1.9150455157653543e-06,
+      "loss": 0.5475,
+      "step": 351
+    },
+    {
+      "epoch": 0.15882684715172024,
+      "grad_norm": 3.5649796357457313,
+      "learning_rate": 1.9144548856758486e-06,
+      "loss": 0.6411,
+      "step": 352
+    },
+    {
+      "epoch": 0.159278059785674,
+      "grad_norm": 3.2716596578005066,
+      "learning_rate": 1.9138623012958393e-06,
+      "loss": 0.5344,
+      "step": 353
+    },
+    {
+      "epoch": 0.15972927241962775,
+      "grad_norm": 3.218727183746927,
+      "learning_rate": 1.913267763891745e-06,
+      "loss": 0.6509,
+      "step": 354
+    },
+    {
+      "epoch": 0.1601804850535815,
+      "grad_norm": 3.7971566293616035,
+      "learning_rate": 1.912671274734156e-06,
+      "loss": 0.5323,
+      "step": 355
+    },
+    {
+      "epoch": 0.16063169768753524,
+      "grad_norm": 3.201203458372683,
+      "learning_rate": 1.9120728350978367e-06,
+      "loss": 0.6743,
+      "step": 356
+    },
+    {
+      "epoch": 0.161082910321489,
+      "grad_norm": 3.3662683266894255,
+      "learning_rate": 1.9114724462617175e-06,
+      "loss": 0.6395,
+      "step": 357
+    },
+    {
+      "epoch": 0.16153412295544275,
+      "grad_norm": 3.4368545644579274,
+      "learning_rate": 1.910870109508896e-06,
+      "loss": 0.5264,
+      "step": 358
+    },
+    {
+      "epoch": 0.1619853355893965,
+      "grad_norm": 3.0271768583204435,
+      "learning_rate": 1.9102658261266306e-06,
+      "loss": 0.6154,
+      "step": 359
+    },
+    {
+      "epoch": 0.16243654822335024,
+      "grad_norm": 3.4935218886533757,
+      "learning_rate": 1.9096595974063424e-06,
+      "loss": 0.5963,
+      "step": 360
+    },
+    {
+      "epoch": 0.162887760857304,
+      "grad_norm": 3.64027486586054,
+      "learning_rate": 1.9090514246436083e-06,
+      "loss": 0.6211,
+      "step": 361
+    },
+    {
+      "epoch": 0.16333897349125776,
+      "grad_norm": 3.396850249691266,
+      "learning_rate": 1.908441309138161e-06,
+      "loss": 0.5142,
+      "step": 362
+    },
+    {
+      "epoch": 0.1637901861252115,
+      "grad_norm": 3.3870576147192053,
+      "learning_rate": 1.907829252193883e-06,
+      "loss": 0.6988,
+      "step": 363
+    },
+    {
+      "epoch": 0.16424139875916527,
+      "grad_norm": 3.788954000277857,
+      "learning_rate": 1.9072152551188081e-06,
+      "loss": 0.5813,
+      "step": 364
+    },
+    {
+      "epoch": 0.164692611393119,
+      "grad_norm": 3.1429732682111933,
+      "learning_rate": 1.9065993192251156e-06,
+      "loss": 0.6034,
+      "step": 365
+    },
+    {
+      "epoch": 0.16514382402707276,
+      "grad_norm": 3.678596714953962,
+      "learning_rate": 1.9059814458291275e-06,
+      "loss": 0.6919,
+      "step": 366
+    },
+    {
+      "epoch": 0.1655950366610265,
+      "grad_norm": 3.062164866400341,
+      "learning_rate": 1.9053616362513076e-06,
+      "loss": 0.4805,
+      "step": 367
+    },
+    {
+      "epoch": 0.16604624929498027,
+      "grad_norm": 3.3886077949868985,
+      "learning_rate": 1.904739891816257e-06,
+      "loss": 0.6668,
+      "step": 368
+    },
+    {
+      "epoch": 0.166497461928934,
+      "grad_norm": 3.521006425014518,
+      "learning_rate": 1.9041162138527112e-06,
+      "loss": 0.7878,
+      "step": 369
+    },
+    {
+      "epoch": 0.16694867456288776,
+      "grad_norm": 3.666265951900786,
+      "learning_rate": 1.9034906036935391e-06,
+      "loss": 0.6963,
+      "step": 370
+    },
+    {
+      "epoch": 0.1673998871968415,
+      "grad_norm": 3.645045850919915,
+      "learning_rate": 1.9028630626757386e-06,
+      "loss": 0.5457,
+      "step": 371
+    },
+    {
+      "epoch": 0.16785109983079527,
+      "grad_norm": 3.4414096877643097,
+      "learning_rate": 1.902233592140433e-06,
+      "loss": 0.6719,
+      "step": 372
+    },
+    {
+      "epoch": 0.16830231246474903,
+      "grad_norm": 3.265461782165135,
+      "learning_rate": 1.9016021934328706e-06,
+      "loss": 0.7438,
+      "step": 373
+    },
+    {
+      "epoch": 0.16875352509870276,
+      "grad_norm": 3.5380564324518415,
+      "learning_rate": 1.9009688679024189e-06,
+      "loss": 0.5506,
+      "step": 374
+    },
+    {
+      "epoch": 0.1692047377326565,
+      "grad_norm": 3.8268327357230345,
+      "learning_rate": 1.9003336169025653e-06,
+      "loss": 0.6524,
+      "step": 375
+    },
+    {
+      "epoch": 0.16965595036661027,
+      "grad_norm": 3.017258959670839,
+      "learning_rate": 1.89969644179091e-06,
+      "loss": 0.5096,
+      "step": 376
+    },
+    {
+      "epoch": 0.17010716300056403,
+      "grad_norm": 3.624267899480704,
+      "learning_rate": 1.8990573439291665e-06,
+      "loss": 0.6393,
+      "step": 377
+    },
+    {
+      "epoch": 0.17055837563451776,
+      "grad_norm": 3.3240315516106405,
+      "learning_rate": 1.8984163246831569e-06,
+      "loss": 0.5305,
+      "step": 378
+    },
+    {
+      "epoch": 0.1710095882684715,
+      "grad_norm": 3.096119100843986,
+      "learning_rate": 1.89777338542281e-06,
+      "loss": 0.5599,
+      "step": 379
+    },
+    {
+      "epoch": 0.17146080090242527,
+      "grad_norm": 3.4243899877650787,
+      "learning_rate": 1.8971285275221577e-06,
+      "loss": 0.5102,
+      "step": 380
+    },
+    {
+      "epoch": 0.17191201353637903,
+      "grad_norm": 3.500631529144404,
+      "learning_rate": 1.8964817523593318e-06,
+      "loss": 0.5937,
+      "step": 381
+    },
+    {
+      "epoch": 0.17236322617033276,
+      "grad_norm": 3.214480819264598,
+      "learning_rate": 1.8958330613165621e-06,
+      "loss": 0.7037,
+      "step": 382
+    },
+    {
+      "epoch": 0.1728144388042865,
+      "grad_norm": 3.2032488316353307,
+      "learning_rate": 1.8951824557801723e-06,
+      "loss": 0.5986,
+      "step": 383
+    },
+    {
+      "epoch": 0.17326565143824027,
+      "grad_norm": 2.773335792708175,
+      "learning_rate": 1.8945299371405783e-06,
+      "loss": 0.4929,
+      "step": 384
+    },
+    {
+      "epoch": 0.17371686407219403,
+      "grad_norm": 3.5616413240552283,
+      "learning_rate": 1.8938755067922836e-06,
+      "loss": 0.6427,
+      "step": 385
+    },
+    {
+      "epoch": 0.17416807670614778,
+      "grad_norm": 3.3452575187465423,
+      "learning_rate": 1.893219166133878e-06,
+      "loss": 0.6748,
+      "step": 386
+    },
+    {
+      "epoch": 0.1746192893401015,
+      "grad_norm": 3.329395667620974,
+      "learning_rate": 1.8925609165680336e-06,
+      "loss": 0.5985,
+      "step": 387
+    },
+    {
+      "epoch": 0.17507050197405527,
+      "grad_norm": 3.1397011497408323,
+      "learning_rate": 1.8919007595015017e-06,
+      "loss": 0.5812,
+      "step": 388
+    },
+    {
+      "epoch": 0.17552171460800903,
+      "grad_norm": 3.1476385473200343,
+      "learning_rate": 1.891238696345111e-06,
+      "loss": 0.5604,
+      "step": 389
+    },
+    {
+      "epoch": 0.17597292724196278,
+      "grad_norm": 3.1908486340362754,
+      "learning_rate": 1.8905747285137625e-06,
+      "loss": 0.6034,
+      "step": 390
+    },
+    {
+      "epoch": 0.1764241398759165,
+      "grad_norm": 3.2455622054904376,
+      "learning_rate": 1.889908857426429e-06,
+      "loss": 0.6501,
+      "step": 391
+    },
+    {
+      "epoch": 0.17687535250987027,
+      "grad_norm": 3.1460953645857925,
+      "learning_rate": 1.8892410845061496e-06,
+      "loss": 0.6749,
+      "step": 392
+    },
+    {
+      "epoch": 0.17732656514382403,
+      "grad_norm": 3.5804988998358818,
+      "learning_rate": 1.8885714111800286e-06,
+      "loss": 0.6729,
+      "step": 393
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.8850245833802806,
+      "learning_rate": 1.8878998388792312e-06,
+      "loss": 0.5124,
+      "step": 394
+    },
+    {
+      "epoch": 0.17822899041173154,
+      "grad_norm": 3.3696439645865772,
+      "learning_rate": 1.8872263690389817e-06,
+      "loss": 0.6009,
+      "step": 395
+    },
+    {
+      "epoch": 0.17868020304568527,
+      "grad_norm": 4.50248696358533,
+      "learning_rate": 1.8865510030985585e-06,
+      "loss": 0.8121,
+      "step": 396
+    },
+    {
+      "epoch": 0.17913141567963903,
+      "grad_norm": 3.2758219521283856,
+      "learning_rate": 1.8858737425012932e-06,
+      "loss": 0.6326,
+      "step": 397
+    },
+    {
+      "epoch": 0.17958262831359278,
+      "grad_norm": 3.0876038288374974,
+      "learning_rate": 1.8851945886945658e-06,
+      "loss": 0.5091,
+      "step": 398
+    },
+    {
+      "epoch": 0.18003384094754654,
+      "grad_norm": 3.161733247624105,
+      "learning_rate": 1.8845135431298025e-06,
+      "loss": 0.5075,
+      "step": 399
+    },
+    {
+      "epoch": 0.18048505358150027,
+      "grad_norm": 3.3744189725502096,
+      "learning_rate": 1.8838306072624729e-06,
+      "loss": 0.6991,
+      "step": 400
+    },
+    {
+      "epoch": 0.18093626621545403,
+      "grad_norm": 3.297593259153372,
+      "learning_rate": 1.8831457825520851e-06,
+      "loss": 0.6062,
+      "step": 401
+    },
+    {
+      "epoch": 0.18138747884940778,
+      "grad_norm": 3.366058459902841,
+      "learning_rate": 1.8824590704621856e-06,
+      "loss": 0.6666,
+      "step": 402
+    },
+    {
+      "epoch": 0.18183869148336154,
+      "grad_norm": 3.7126042071152976,
+      "learning_rate": 1.8817704724603533e-06,
+      "loss": 0.6618,
+      "step": 403
+    },
+    {
+      "epoch": 0.18228990411731527,
+      "grad_norm": 2.9848703959766825,
+      "learning_rate": 1.8810799900181976e-06,
+      "loss": 0.609,
+      "step": 404
+    },
+    {
+      "epoch": 0.18274111675126903,
+      "grad_norm": 3.649965565236517,
+      "learning_rate": 1.8803876246113552e-06,
+      "loss": 0.5993,
+      "step": 405
+    },
+    {
+      "epoch": 0.18319232938522279,
+      "grad_norm": 2.89156057373746,
+      "learning_rate": 1.8796933777194871e-06,
+      "loss": 0.6068,
+      "step": 406
+    },
+    {
+      "epoch": 0.18364354201917654,
+      "grad_norm": 3.0603328302400197,
+      "learning_rate": 1.8789972508262752e-06,
+      "loss": 0.5925,
+      "step": 407
+    },
+    {
+      "epoch": 0.1840947546531303,
+      "grad_norm": 3.2452938065021484,
+      "learning_rate": 1.8782992454194192e-06,
+      "loss": 0.4789,
+      "step": 408
+    },
+    {
+      "epoch": 0.18454596728708403,
+      "grad_norm": 2.8228292796413963,
+      "learning_rate": 1.877599362990633e-06,
+      "loss": 0.5509,
+      "step": 409
+    },
+    {
+      "epoch": 0.18499717992103779,
+      "grad_norm": 3.556457875267904,
+      "learning_rate": 1.8768976050356424e-06,
+      "loss": 0.5953,
+      "step": 410
+    },
+    {
+      "epoch": 0.18544839255499154,
+      "grad_norm": 3.792739279541135,
+      "learning_rate": 1.876193973054181e-06,
+      "loss": 0.7014,
+      "step": 411
+    },
+    {
+      "epoch": 0.1858996051889453,
+      "grad_norm": 3.2493298778303354,
+      "learning_rate": 1.8754884685499884e-06,
+      "loss": 0.4082,
+      "step": 412
+    },
+    {
+      "epoch": 0.18635081782289903,
+      "grad_norm": 3.6609624714373457,
+      "learning_rate": 1.874781093030804e-06,
+      "loss": 0.6845,
+      "step": 413
+    },
+    {
+      "epoch": 0.18680203045685279,
+      "grad_norm": 3.2137855218137594,
+      "learning_rate": 1.8740718480083678e-06,
+      "loss": 0.6125,
+      "step": 414
+    },
+    {
+      "epoch": 0.18725324309080654,
+      "grad_norm": 3.499410379932756,
+      "learning_rate": 1.8733607349984138e-06,
+      "loss": 0.6784,
+      "step": 415
+    },
+    {
+      "epoch": 0.1877044557247603,
+      "grad_norm": 3.2539894662199673,
+      "learning_rate": 1.8726477555206688e-06,
+      "loss": 0.6767,
+      "step": 416
+    },
+    {
+      "epoch": 0.18815566835871406,
+      "grad_norm": 3.7584814555532913,
+      "learning_rate": 1.8719329110988484e-06,
+      "loss": 0.5895,
+      "step": 417
+    },
+    {
+      "epoch": 0.18860688099266779,
+      "grad_norm": 2.7529038721428045,
+      "learning_rate": 1.8712162032606536e-06,
+      "loss": 0.5273,
+      "step": 418
+    },
+    {
+      "epoch": 0.18905809362662154,
+      "grad_norm": 3.033230260922539,
+      "learning_rate": 1.8704976335377676e-06,
+      "loss": 0.4975,
+      "step": 419
+    },
+    {
+      "epoch": 0.1895093062605753,
+      "grad_norm": 3.383380728150317,
+      "learning_rate": 1.8697772034658525e-06,
+      "loss": 0.7518,
+      "step": 420
+    },
+    {
+      "epoch": 0.18996051889452906,
+      "grad_norm": 3.4073642749167408,
+      "learning_rate": 1.8690549145845473e-06,
+      "loss": 0.6152,
+      "step": 421
+    },
+    {
+      "epoch": 0.19041173152848279,
+      "grad_norm": 2.9297215005432586,
+      "learning_rate": 1.8683307684374618e-06,
+      "loss": 0.5265,
+      "step": 422
+    },
+    {
+      "epoch": 0.19086294416243654,
+      "grad_norm": 3.167963569614919,
+      "learning_rate": 1.8676047665721763e-06,
+      "loss": 0.5796,
+      "step": 423
+    },
+    {
+      "epoch": 0.1913141567963903,
+      "grad_norm": 3.1693108021010654,
+      "learning_rate": 1.8668769105402365e-06,
+      "loss": 0.5744,
+      "step": 424
+    },
+    {
+      "epoch": 0.19176536943034406,
+      "grad_norm": 3.0256228579439046,
+      "learning_rate": 1.8661472018971502e-06,
+      "loss": 0.6616,
+      "step": 425
+    },
+    {
+      "epoch": 0.19221658206429781,
+      "grad_norm": 3.2078343663005486,
+      "learning_rate": 1.8654156422023858e-06,
+      "loss": 0.5892,
+      "step": 426
+    },
+    {
+      "epoch": 0.19266779469825154,
+      "grad_norm": 3.227710950393321,
+      "learning_rate": 1.8646822330193657e-06,
+      "loss": 0.6133,
+      "step": 427
+    },
+    {
+      "epoch": 0.1931190073322053,
+      "grad_norm": 3.941318300397022,
+      "learning_rate": 1.8639469759154665e-06,
+      "loss": 0.6455,
+      "step": 428
+    },
+    {
+      "epoch": 0.19357021996615906,
+      "grad_norm": 3.698927481165211,
+      "learning_rate": 1.863209872462013e-06,
+      "loss": 0.7186,
+      "step": 429
+    },
+    {
+      "epoch": 0.19402143260011281,
+      "grad_norm": 3.3302251096427833,
+      "learning_rate": 1.8624709242342763e-06,
+      "loss": 0.6038,
+      "step": 430
+    },
+    {
+      "epoch": 0.19447264523406654,
+      "grad_norm": 3.290865194019538,
+      "learning_rate": 1.8617301328114702e-06,
+      "loss": 0.5836,
+      "step": 431
+    },
+    {
+      "epoch": 0.1949238578680203,
+      "grad_norm": 3.310231926134767,
+      "learning_rate": 1.8609874997767471e-06,
+      "loss": 0.5143,
+      "step": 432
+    },
+    {
+      "epoch": 0.19537507050197406,
+      "grad_norm": 3.458415221251539,
+      "learning_rate": 1.8602430267171953e-06,
+      "loss": 0.5223,
+      "step": 433
+    },
+    {
+      "epoch": 0.19582628313592781,
+      "grad_norm": 3.5702337410450005,
+      "learning_rate": 1.8594967152238356e-06,
+      "loss": 0.6378,
+      "step": 434
+    },
+    {
+      "epoch": 0.19627749576988154,
+      "grad_norm": 3.3767941834158477,
+      "learning_rate": 1.8587485668916175e-06,
+      "loss": 0.7065,
+      "step": 435
+    },
+    {
+      "epoch": 0.1967287084038353,
+      "grad_norm": 3.376132837313131,
+      "learning_rate": 1.857998583319416e-06,
+      "loss": 0.6704,
+      "step": 436
+    },
+    {
+      "epoch": 0.19717992103778906,
+      "grad_norm": 3.7269212236357614,
+      "learning_rate": 1.8572467661100285e-06,
+      "loss": 0.6655,
+      "step": 437
+    },
+    {
+      "epoch": 0.19763113367174281,
+      "grad_norm": 2.885465129936691,
+      "learning_rate": 1.856493116870171e-06,
+      "loss": 0.5092,
+      "step": 438
+    },
+    {
+      "epoch": 0.19808234630569657,
+      "grad_norm": 3.2219073896145374,
+      "learning_rate": 1.855737637210475e-06,
+      "loss": 0.6651,
+      "step": 439
+    },
+    {
+      "epoch": 0.1985335589396503,
+      "grad_norm": 3.020821738843593,
+      "learning_rate": 1.8549803287454828e-06,
+      "loss": 0.5339,
+      "step": 440
+    },
+    {
+      "epoch": 0.19898477157360406,
+      "grad_norm": 3.3074095664255814,
+      "learning_rate": 1.8542211930936461e-06,
+      "loss": 0.6784,
+      "step": 441
+    },
+    {
+      "epoch": 0.19943598420755781,
+      "grad_norm": 3.0185027225876344,
+      "learning_rate": 1.8534602318773211e-06,
+      "loss": 0.4799,
+      "step": 442
+    },
+    {
+      "epoch": 0.19988719684151157,
+      "grad_norm": 3.3658699683573587,
+      "learning_rate": 1.8526974467227657e-06,
+      "loss": 0.5748,
+      "step": 443
+    },
+    {
+      "epoch": 0.2003384094754653,
+      "grad_norm": 3.3754909363924925,
+      "learning_rate": 1.8519328392601348e-06,
+      "loss": 0.6109,
+      "step": 444
+    },
+    {
+      "epoch": 0.20078962210941906,
+      "grad_norm": 3.3689745519480665,
+      "learning_rate": 1.8511664111234796e-06,
+      "loss": 0.6079,
+      "step": 445
+    },
+    {
+      "epoch": 0.20124083474337282,
+      "grad_norm": 3.843717566439344,
+      "learning_rate": 1.8503981639507402e-06,
+      "loss": 0.6312,
+      "step": 446
+    },
+    {
+      "epoch": 0.20169204737732657,
+      "grad_norm": 3.2274835544812848,
+      "learning_rate": 1.8496280993837457e-06,
+      "loss": 0.5936,
+      "step": 447
+    },
+    {
+      "epoch": 0.20214326001128033,
+      "grad_norm": 3.311187513569298,
+      "learning_rate": 1.8488562190682087e-06,
+      "loss": 0.5793,
+      "step": 448
+    },
+    {
+      "epoch": 0.20259447264523406,
+      "grad_norm": 2.9215114080546214,
+      "learning_rate": 1.8480825246537217e-06,
+      "loss": 0.5959,
+      "step": 449
+    },
+    {
+      "epoch": 0.20304568527918782,
+      "grad_norm": 3.404182604070715,
+      "learning_rate": 1.847307017793755e-06,
+      "loss": 0.5429,
+      "step": 450
+    },
+    {
+      "epoch": 0.20349689791314157,
+      "grad_norm": 3.2363535948970856,
+      "learning_rate": 1.846529700145652e-06,
+      "loss": 0.6569,
+      "step": 451
+    },
+    {
+      "epoch": 0.20394811054709533,
+      "grad_norm": 3.511507526063595,
+      "learning_rate": 1.8457505733706258e-06,
+      "loss": 0.4839,
+      "step": 452
+    },
+    {
+      "epoch": 0.20439932318104906,
+      "grad_norm": 3.37043417245356,
+      "learning_rate": 1.8449696391337554e-06,
+      "loss": 0.669,
+      "step": 453
+    },
+    {
+      "epoch": 0.20485053581500282,
+      "grad_norm": 3.9931457464636155,
+      "learning_rate": 1.8441868991039837e-06,
+      "loss": 0.5805,
+      "step": 454
+    },
+    {
+      "epoch": 0.20530174844895657,
+      "grad_norm": 2.9942829712689605,
+      "learning_rate": 1.8434023549541115e-06,
+      "loss": 0.5925,
+      "step": 455
+    },
+    {
+      "epoch": 0.20575296108291033,
+      "grad_norm": 3.1956405252290594,
+      "learning_rate": 1.8426160083607964e-06,
+      "loss": 0.5746,
+      "step": 456
+    },
+    {
+      "epoch": 0.20620417371686406,
+      "grad_norm": 2.916883450705641,
+      "learning_rate": 1.841827861004547e-06,
+      "loss": 0.5662,
+      "step": 457
+    },
+    {
+      "epoch": 0.20665538635081782,
+      "grad_norm": 3.431394584303988,
+      "learning_rate": 1.8410379145697208e-06,
+      "loss": 0.6516,
+      "step": 458
+    },
+    {
+      "epoch": 0.20710659898477157,
+      "grad_norm": 3.165177547700952,
+      "learning_rate": 1.8402461707445203e-06,
+      "loss": 0.5172,
+      "step": 459
+    },
+    {
+      "epoch": 0.20755781161872533,
+      "grad_norm": 3.7977061154161302,
+      "learning_rate": 1.8394526312209885e-06,
+      "loss": 0.6817,
+      "step": 460
+    },
+    {
+      "epoch": 0.2080090242526791,
+      "grad_norm": 2.924039062504755,
+      "learning_rate": 1.838657297695007e-06,
+      "loss": 0.7493,
+      "step": 461
+    },
+    {
+      "epoch": 0.20846023688663282,
+      "grad_norm": 3.120504427001902,
+      "learning_rate": 1.8378601718662905e-06,
+      "loss": 0.5429,
+      "step": 462
+    },
+    {
+      "epoch": 0.20891144952058657,
+      "grad_norm": 3.45109119066591,
+      "learning_rate": 1.8370612554383848e-06,
+      "loss": 0.6886,
+      "step": 463
+    },
+    {
+      "epoch": 0.20936266215454033,
+      "grad_norm": 3.9533453151193005,
+      "learning_rate": 1.8362605501186618e-06,
+      "loss": 0.6993,
+      "step": 464
+    },
+    {
+      "epoch": 0.2098138747884941,
+      "grad_norm": 3.5868823288557463,
+      "learning_rate": 1.8354580576183167e-06,
+      "loss": 0.7202,
+      "step": 465
+    },
+    {
+      "epoch": 0.21026508742244782,
+      "grad_norm": 3.360808407157414,
+      "learning_rate": 1.8346537796523642e-06,
+      "loss": 0.7506,
+      "step": 466
+    },
+    {
+      "epoch": 0.21071630005640157,
+      "grad_norm": 3.5999786014669453,
+      "learning_rate": 1.8338477179396347e-06,
+      "loss": 0.6462,
+      "step": 467
+    },
+    {
+      "epoch": 0.21116751269035533,
+      "grad_norm": 3.089511771354358,
+      "learning_rate": 1.8330398742027702e-06,
+      "loss": 0.5945,
+      "step": 468
+    },
+    {
+      "epoch": 0.2116187253243091,
+      "grad_norm": 3.1839344612610376,
+      "learning_rate": 1.8322302501682216e-06,
+      "loss": 0.5521,
+      "step": 469
+    },
+    {
+      "epoch": 0.21206993795826284,
+      "grad_norm": 3.2784855658730883,
+      "learning_rate": 1.831418847566245e-06,
+      "loss": 0.5371,
+      "step": 470
+    },
+    {
+      "epoch": 0.21252115059221657,
+      "grad_norm": 3.346783346886957,
+      "learning_rate": 1.8306056681308957e-06,
+      "loss": 0.6287,
+      "step": 471
+    },
+    {
+      "epoch": 0.21297236322617033,
+      "grad_norm": 3.2663627359220646,
+      "learning_rate": 1.8297907136000283e-06,
+      "loss": 0.6486,
+      "step": 472
+    },
+    {
+      "epoch": 0.2134235758601241,
+      "grad_norm": 3.035708873461981,
+      "learning_rate": 1.8289739857152903e-06,
+      "loss": 0.5249,
+      "step": 473
+    },
+    {
+      "epoch": 0.21387478849407784,
+      "grad_norm": 3.14810881896666,
+      "learning_rate": 1.8281554862221179e-06,
+      "loss": 0.6863,
+      "step": 474
+    },
+    {
+      "epoch": 0.21432600112803157,
+      "grad_norm": 3.606635360142977,
+      "learning_rate": 1.827335216869735e-06,
+      "loss": 0.5768,
+      "step": 475
+    },
+    {
+      "epoch": 0.21477721376198533,
+      "grad_norm": 3.420453255878494,
+      "learning_rate": 1.8265131794111477e-06,
+      "loss": 0.5868,
+      "step": 476
+    },
+    {
+      "epoch": 0.2152284263959391,
+      "grad_norm": 3.5479674298971213,
+      "learning_rate": 1.8256893756031396e-06,
+      "loss": 0.5529,
+      "step": 477
+    },
+    {
+      "epoch": 0.21567963902989284,
+      "grad_norm": 3.5672449049409143,
+      "learning_rate": 1.82486380720627e-06,
+      "loss": 0.5414,
+      "step": 478
+    },
+    {
+      "epoch": 0.2161308516638466,
+      "grad_norm": 3.4755060623971157,
+      "learning_rate": 1.8240364759848697e-06,
+      "loss": 0.6827,
+      "step": 479
+    },
+    {
+      "epoch": 0.21658206429780033,
+      "grad_norm": 3.5075155658583532,
+      "learning_rate": 1.823207383707036e-06,
+      "loss": 0.6522,
+      "step": 480
+    },
+    {
+      "epoch": 0.2170332769317541,
+      "grad_norm": 3.615334770032639,
+      "learning_rate": 1.82237653214463e-06,
+      "loss": 0.6112,
+      "step": 481
+    },
+    {
+      "epoch": 0.21748448956570784,
+      "grad_norm": 2.930096209443241,
+      "learning_rate": 1.8215439230732728e-06,
+      "loss": 0.4904,
+      "step": 482
+    },
+    {
+      "epoch": 0.2179357021996616,
+      "grad_norm": 3.045321232724037,
+      "learning_rate": 1.8207095582723416e-06,
+      "loss": 0.4742,
+      "step": 483
+    },
+    {
+      "epoch": 0.21838691483361533,
+      "grad_norm": 3.318218916061034,
+      "learning_rate": 1.8198734395249654e-06,
+      "loss": 0.5816,
+      "step": 484
+    },
+    {
+      "epoch": 0.2188381274675691,
+      "grad_norm": 3.238638276761325,
+      "learning_rate": 1.8190355686180218e-06,
+      "loss": 0.7086,
+      "step": 485
+    },
+    {
+      "epoch": 0.21928934010152284,
+      "grad_norm": 3.088879882296647,
+      "learning_rate": 1.8181959473421334e-06,
+      "loss": 0.5967,
+      "step": 486
+    },
+    {
+      "epoch": 0.2197405527354766,
+      "grad_norm": 3.3716940597716585,
+      "learning_rate": 1.8173545774916626e-06,
+      "loss": 0.7092,
+      "step": 487
+    },
+    {
+      "epoch": 0.22019176536943033,
+      "grad_norm": 3.7448981916189825,
+      "learning_rate": 1.816511460864709e-06,
+      "loss": 0.6427,
+      "step": 488
+    },
+    {
+      "epoch": 0.2206429780033841,
+      "grad_norm": 2.9887789178064548,
+      "learning_rate": 1.8156665992631057e-06,
+      "loss": 0.5822,
+      "step": 489
+    },
+    {
+      "epoch": 0.22109419063733785,
+      "grad_norm": 3.3414881969228714,
+      "learning_rate": 1.8148199944924146e-06,
+      "loss": 0.5946,
+      "step": 490
+    },
+    {
+      "epoch": 0.2215454032712916,
+      "grad_norm": 3.4422877305450763,
+      "learning_rate": 1.8139716483619232e-06,
+      "loss": 0.6066,
+      "step": 491
+    },
+    {
+      "epoch": 0.22199661590524536,
+      "grad_norm": 3.4139774261820093,
+      "learning_rate": 1.8131215626846403e-06,
+      "loss": 0.6513,
+      "step": 492
+    },
+    {
+      "epoch": 0.2224478285391991,
+      "grad_norm": 2.9492703190535465,
+      "learning_rate": 1.8122697392772923e-06,
+      "loss": 0.512,
+      "step": 493
+    },
+    {
+      "epoch": 0.22289904117315285,
+      "grad_norm": 3.392493989272701,
+      "learning_rate": 1.8114161799603192e-06,
+      "loss": 0.6111,
+      "step": 494
+    },
+    {
+      "epoch": 0.2233502538071066,
+      "grad_norm": 3.4162235015689095,
+      "learning_rate": 1.8105608865578712e-06,
+      "loss": 0.7261,
+      "step": 495
+    },
+    {
+      "epoch": 0.22380146644106036,
+      "grad_norm": 3.3485739796813645,
+      "learning_rate": 1.809703860897804e-06,
+      "loss": 0.5977,
+      "step": 496
+    },
+    {
+      "epoch": 0.2242526790750141,
+      "grad_norm": 3.3359232989076917,
+      "learning_rate": 1.808845104811676e-06,
+      "loss": 0.4957,
+      "step": 497
+    },
+    {
+      "epoch": 0.22470389170896785,
+      "grad_norm": 3.1119819516264706,
+      "learning_rate": 1.8079846201347426e-06,
+      "loss": 0.566,
+      "step": 498
+    },
+    {
+      "epoch": 0.2251551043429216,
+      "grad_norm": 3.33013160940601,
+      "learning_rate": 1.8071224087059545e-06,
+      "loss": 0.6161,
+      "step": 499
+    },
+    {
+      "epoch": 0.22560631697687536,
+      "grad_norm": 3.726562728905523,
+      "learning_rate": 1.806258472367952e-06,
+      "loss": 0.6273,
+      "step": 500
+    },
+    {
+      "epoch": 0.22605752961082912,
+      "grad_norm": 3.112643238486848,
+      "learning_rate": 1.805392812967062e-06,
+      "loss": 0.524,
+      "step": 501
+    },
+    {
+      "epoch": 0.22650874224478285,
+      "grad_norm": 3.078076912846401,
+      "learning_rate": 1.8045254323532938e-06,
+      "loss": 0.5933,
+      "step": 502
+    },
+    {
+      "epoch": 0.2269599548787366,
+      "grad_norm": 3.3007163587543293,
+      "learning_rate": 1.803656332380335e-06,
+      "loss": 0.6058,
+      "step": 503
+    },
+    {
+      "epoch": 0.22741116751269036,
+      "grad_norm": 3.350696137526816,
+      "learning_rate": 1.8027855149055476e-06,
+      "loss": 0.6761,
+      "step": 504
+    },
+    {
+      "epoch": 0.22786238014664412,
+      "grad_norm": 3.056386670780574,
+      "learning_rate": 1.8019129817899641e-06,
+      "loss": 0.5403,
+      "step": 505
+    },
+    {
+      "epoch": 0.22831359278059785,
+      "grad_norm": 3.3200840574141703,
+      "learning_rate": 1.8010387348982834e-06,
+      "loss": 0.4723,
+      "step": 506
+    },
+    {
+      "epoch": 0.2287648054145516,
+      "grad_norm": 2.9031378599313964,
+      "learning_rate": 1.8001627760988676e-06,
+      "loss": 0.5124,
+      "step": 507
+    },
+    {
+      "epoch": 0.22921601804850536,
+      "grad_norm": 3.0170169190296425,
+      "learning_rate": 1.7992851072637364e-06,
+      "loss": 0.5241,
+      "step": 508
+    },
+    {
+      "epoch": 0.22966723068245912,
+      "grad_norm": 3.260274778241787,
+      "learning_rate": 1.7984057302685645e-06,
+      "loss": 0.612,
+      "step": 509
+    },
+    {
+      "epoch": 0.23011844331641285,
+      "grad_norm": 2.957767077162712,
+      "learning_rate": 1.7975246469926773e-06,
+      "loss": 0.5856,
+      "step": 510
+    },
+    {
+      "epoch": 0.2305696559503666,
+      "grad_norm": 3.077346118529033,
+      "learning_rate": 1.7966418593190466e-06,
+      "loss": 0.6476,
+      "step": 511
+    },
+    {
+      "epoch": 0.23102086858432036,
+      "grad_norm": 3.40658654324856,
+      "learning_rate": 1.7957573691342863e-06,
+      "loss": 0.6496,
+      "step": 512
+    },
+    {
+      "epoch": 0.23147208121827412,
+      "grad_norm": 3.1537347899768187,
+      "learning_rate": 1.7948711783286494e-06,
+      "loss": 0.4991,
+      "step": 513
+    },
+    {
+      "epoch": 0.23192329385222787,
+      "grad_norm": 3.3406426536701064,
+      "learning_rate": 1.7939832887960228e-06,
+      "loss": 0.6539,
+      "step": 514
+    },
+    {
+      "epoch": 0.2323745064861816,
+      "grad_norm": 3.6637558500642617,
+      "learning_rate": 1.7930937024339236e-06,
+      "loss": 0.5154,
+      "step": 515
+    },
+    {
+      "epoch": 0.23282571912013536,
+      "grad_norm": 3.528943118739639,
+      "learning_rate": 1.7922024211434958e-06,
+      "loss": 0.4997,
+      "step": 516
+    },
+    {
+      "epoch": 0.23327693175408912,
+      "grad_norm": 2.8648086426744386,
+      "learning_rate": 1.7913094468295056e-06,
+      "loss": 0.4715,
+      "step": 517
+    },
+    {
+      "epoch": 0.23372814438804287,
+      "grad_norm": 3.39893956335253,
+      "learning_rate": 1.790414781400337e-06,
+      "loss": 0.6078,
+      "step": 518
+    },
+    {
+      "epoch": 0.2341793570219966,
+      "grad_norm": 3.3165376220099976,
+      "learning_rate": 1.7895184267679882e-06,
+      "loss": 0.5808,
+      "step": 519
+    },
+    {
+      "epoch": 0.23463056965595036,
+      "grad_norm": 3.220094259664805,
+      "learning_rate": 1.7886203848480671e-06,
+      "loss": 0.6231,
+      "step": 520
+    },
+    {
+      "epoch": 0.23508178228990412,
+      "grad_norm": 3.1138985690928886,
+      "learning_rate": 1.7877206575597887e-06,
+      "loss": 0.6523,
+      "step": 521
+    },
+    {
+      "epoch": 0.23553299492385787,
+      "grad_norm": 3.1979428387224935,
+      "learning_rate": 1.7868192468259684e-06,
+      "loss": 0.598,
+      "step": 522
+    },
+    {
+      "epoch": 0.23598420755781163,
+      "grad_norm": 3.135032930567459,
+      "learning_rate": 1.7859161545730204e-06,
+      "loss": 0.7253,
+      "step": 523
+    },
+    {
+      "epoch": 0.23643542019176536,
+      "grad_norm": 2.85623015158301,
+      "learning_rate": 1.7850113827309516e-06,
+      "loss": 0.485,
+      "step": 524
+    },
+    {
+      "epoch": 0.23688663282571912,
+      "grad_norm": 3.126524923453851,
+      "learning_rate": 1.7841049332333588e-06,
+      "loss": 0.6822,
+      "step": 525
+    },
+    {
+      "epoch": 0.23733784545967287,
+      "grad_norm": 3.2260751063220416,
+      "learning_rate": 1.7831968080174245e-06,
+      "loss": 0.5968,
+      "step": 526
+    },
+    {
+      "epoch": 0.23778905809362663,
+      "grad_norm": 3.2163409363779656,
+      "learning_rate": 1.7822870090239116e-06,
+      "loss": 0.6282,
+      "step": 527
+    },
+    {
+      "epoch": 0.23824027072758036,
+      "grad_norm": 3.223131291822765,
+      "learning_rate": 1.7813755381971603e-06,
+      "loss": 0.6361,
+      "step": 528
+    },
+    {
+      "epoch": 0.23869148336153412,
+      "grad_norm": 3.5266582311851367,
+      "learning_rate": 1.7804623974850843e-06,
+      "loss": 0.5917,
+      "step": 529
+    },
+    {
+      "epoch": 0.23914269599548788,
+      "grad_norm": 3.756212649445671,
+      "learning_rate": 1.7795475888391654e-06,
+      "loss": 0.5342,
+      "step": 530
+    },
+    {
+      "epoch": 0.23959390862944163,
+      "grad_norm": 3.0598171380651658,
+      "learning_rate": 1.7786311142144501e-06,
+      "loss": 0.7084,
+      "step": 531
+    },
+    {
+      "epoch": 0.24004512126339536,
+      "grad_norm": 3.3421256359633476,
+      "learning_rate": 1.7777129755695453e-06,
+      "loss": 0.5933,
+      "step": 532
+    },
+    {
+      "epoch": 0.24049633389734912,
+      "grad_norm": 3.5281023446435564,
+      "learning_rate": 1.7767931748666143e-06,
+      "loss": 0.6578,
+      "step": 533
+    },
+    {
+      "epoch": 0.24094754653130288,
+      "grad_norm": 3.0273532574157613,
+      "learning_rate": 1.7758717140713717e-06,
+      "loss": 0.5256,
+      "step": 534
+    },
+    {
+      "epoch": 0.24139875916525663,
+      "grad_norm": 3.197844985849922,
+      "learning_rate": 1.7749485951530812e-06,
+      "loss": 0.6199,
+      "step": 535
+    },
+    {
+      "epoch": 0.2418499717992104,
+      "grad_norm": 3.380058493493815,
+      "learning_rate": 1.7740238200845484e-06,
+      "loss": 0.5581,
+      "step": 536
+    },
+    {
+      "epoch": 0.24230118443316412,
+      "grad_norm": 2.8794589169982405,
+      "learning_rate": 1.7730973908421196e-06,
+      "loss": 0.5515,
+      "step": 537
+    },
+    {
+      "epoch": 0.24275239706711788,
+      "grad_norm": 3.303088479063469,
+      "learning_rate": 1.772169309405676e-06,
+      "loss": 0.6068,
+      "step": 538
+    },
+    {
+      "epoch": 0.24320360970107163,
+      "grad_norm": 3.5894154844174775,
+      "learning_rate": 1.7712395777586294e-06,
+      "loss": 0.569,
+      "step": 539
+    },
+    {
+      "epoch": 0.2436548223350254,
+      "grad_norm": 3.412578792730928,
+      "learning_rate": 1.770308197887918e-06,
+      "loss": 0.5934,
+      "step": 540
+    },
+    {
+      "epoch": 0.24410603496897912,
+      "grad_norm": 3.4253075329490814,
+      "learning_rate": 1.7693751717840035e-06,
+      "loss": 0.6087,
+      "step": 541
+    },
+    {
+      "epoch": 0.24455724760293288,
+      "grad_norm": 3.499633565706812,
+      "learning_rate": 1.7684405014408647e-06,
+      "loss": 0.588,
+      "step": 542
+    },
+    {
+      "epoch": 0.24500846023688663,
+      "grad_norm": 2.814806236683677,
+      "learning_rate": 1.767504188855995e-06,
+      "loss": 0.482,
+      "step": 543
+    },
+    {
+      "epoch": 0.2454596728708404,
+      "grad_norm": 3.098881310366283,
+      "learning_rate": 1.766566236030397e-06,
+      "loss": 0.5435,
+      "step": 544
+    },
+    {
+      "epoch": 0.24591088550479415,
+      "grad_norm": 3.288381281791134,
+      "learning_rate": 1.7656266449685792e-06,
+      "loss": 0.5992,
+      "step": 545
+    },
+    {
+      "epoch": 0.24636209813874788,
+      "grad_norm": 3.4041539654568425,
+      "learning_rate": 1.7646854176785506e-06,
+      "loss": 0.5966,
+      "step": 546
+    },
+    {
+      "epoch": 0.24681331077270163,
+      "grad_norm": 3.4008922515143345,
+      "learning_rate": 1.7637425561718174e-06,
+      "loss": 0.7074,
+      "step": 547
+    },
+    {
+      "epoch": 0.2472645234066554,
+      "grad_norm": 3.552760007080402,
+      "learning_rate": 1.762798062463378e-06,
+      "loss": 0.7019,
+      "step": 548
+    },
+    {
+      "epoch": 0.24771573604060915,
+      "grad_norm": 3.6537377564708255,
+      "learning_rate": 1.7618519385717192e-06,
+      "loss": 0.6482,
+      "step": 549
+    },
+    {
+      "epoch": 0.24816694867456288,
+      "grad_norm": 3.11958006185935,
+      "learning_rate": 1.7609041865188118e-06,
+      "loss": 0.5418,
+      "step": 550
+    },
+    {
+      "epoch": 0.24861816130851663,
+      "grad_norm": 3.336252618186167,
+      "learning_rate": 1.759954808330106e-06,
+      "loss": 0.6389,
+      "step": 551
+    },
+    {
+      "epoch": 0.2490693739424704,
+      "grad_norm": 2.921189003705698,
+      "learning_rate": 1.7590038060345274e-06,
+      "loss": 0.4537,
+      "step": 552
+    },
+    {
+      "epoch": 0.24952058657642415,
+      "grad_norm": 3.33855785234176,
+      "learning_rate": 1.7580511816644715e-06,
+      "loss": 0.5591,
+      "step": 553
+    },
+    {
+      "epoch": 0.2499717992103779,
+      "grad_norm": 4.125492421237276,
+      "learning_rate": 1.7570969372558021e-06,
+      "loss": 0.5379,
+      "step": 554
+    },
+    {
+      "epoch": 0.25042301184433163,
+      "grad_norm": 3.0575497100367683,
+      "learning_rate": 1.7561410748478441e-06,
+      "loss": 0.5602,
+      "step": 555
+    },
+    {
+      "epoch": 0.2508742244782854,
+      "grad_norm": 3.2758898433144075,
+      "learning_rate": 1.75518359648338e-06,
+      "loss": 0.4833,
+      "step": 556
+    },
+    {
+      "epoch": 0.25132543711223915,
+      "grad_norm": 3.2778885737843777,
+      "learning_rate": 1.7542245042086467e-06,
+      "loss": 0.5926,
+      "step": 557
+    },
+    {
+      "epoch": 0.2517766497461929,
+      "grad_norm": 3.2596395219328085,
+      "learning_rate": 1.7532638000733293e-06,
+      "loss": 0.555,
+      "step": 558
+    },
+    {
+      "epoch": 0.25222786238014666,
+      "grad_norm": 2.9901086229008227,
+      "learning_rate": 1.7523014861305585e-06,
+      "loss": 0.4896,
+      "step": 559
+    },
+    {
+      "epoch": 0.2526790750141004,
+      "grad_norm": 3.244315223694814,
+      "learning_rate": 1.7513375644369046e-06,
+      "loss": 0.585,
+      "step": 560
+    },
+    {
+      "epoch": 0.2531302876480541,
+      "grad_norm": 3.198365136580784,
+      "learning_rate": 1.750372037052374e-06,
+      "loss": 0.528,
+      "step": 561
+    },
+    {
+      "epoch": 0.2535815002820079,
+      "grad_norm": 3.2771916408813406,
+      "learning_rate": 1.7494049060404047e-06,
+      "loss": 0.7574,
+      "step": 562
+    },
+    {
+      "epoch": 0.25403271291596163,
+      "grad_norm": 2.8807336033407487,
+      "learning_rate": 1.7484361734678621e-06,
+      "loss": 0.5565,
+      "step": 563
+    },
+    {
+      "epoch": 0.2544839255499154,
+      "grad_norm": 3.6211865966847494,
+      "learning_rate": 1.747465841405034e-06,
+      "loss": 0.5855,
+      "step": 564
+    },
+    {
+      "epoch": 0.25493513818386915,
+      "grad_norm": 3.1436849609005204,
+      "learning_rate": 1.7464939119256266e-06,
+      "loss": 0.6694,
+      "step": 565
+    },
+    {
+      "epoch": 0.2553863508178229,
+      "grad_norm": 2.7285316400886113,
+      "learning_rate": 1.7455203871067596e-06,
+      "loss": 0.5036,
+      "step": 566
+    },
+    {
+      "epoch": 0.25583756345177666,
+      "grad_norm": 3.186433555908316,
+      "learning_rate": 1.744545269028963e-06,
+      "loss": 0.553,
+      "step": 567
+    },
+    {
+      "epoch": 0.2562887760857304,
+      "grad_norm": 2.8827236037421033,
+      "learning_rate": 1.7435685597761707e-06,
+      "loss": 0.5289,
+      "step": 568
+    },
+    {
+      "epoch": 0.2567399887196842,
+      "grad_norm": 3.181781322960324,
+      "learning_rate": 1.742590261435718e-06,
+      "loss": 0.6055,
+      "step": 569
+    },
+    {
+      "epoch": 0.2571912013536379,
+      "grad_norm": 3.4460098001320443,
+      "learning_rate": 1.7416103760983356e-06,
+      "loss": 0.5721,
+      "step": 570
+    },
+    {
+      "epoch": 0.25764241398759163,
+      "grad_norm": 3.2888809340579033,
+      "learning_rate": 1.7406289058581463e-06,
+      "loss": 0.5706,
+      "step": 571
+    },
+    {
+      "epoch": 0.2580936266215454,
+      "grad_norm": 3.039055552956028,
+      "learning_rate": 1.7396458528126594e-06,
+      "loss": 0.6311,
+      "step": 572
+    },
+    {
+      "epoch": 0.25854483925549915,
+      "grad_norm": 3.1817781660973012,
+      "learning_rate": 1.738661219062768e-06,
+      "loss": 0.5731,
+      "step": 573
+    },
+    {
+      "epoch": 0.2589960518894529,
+      "grad_norm": 3.2166470180702276,
+      "learning_rate": 1.7376750067127412e-06,
+      "loss": 0.6233,
+      "step": 574
+    },
+    {
+      "epoch": 0.25944726452340666,
+      "grad_norm": 2.933567700518428,
+      "learning_rate": 1.7366872178702246e-06,
+      "loss": 0.4681,
+      "step": 575
+    },
+    {
+      "epoch": 0.2598984771573604,
+      "grad_norm": 3.339854630057023,
+      "learning_rate": 1.7356978546462305e-06,
+      "loss": 0.5115,
+      "step": 576
+    },
+    {
+      "epoch": 0.2603496897913142,
+      "grad_norm": 3.042537582951306,
+      "learning_rate": 1.7347069191551367e-06,
+      "loss": 0.4721,
+      "step": 577
+    },
+    {
+      "epoch": 0.26080090242526793,
+      "grad_norm": 3.4742498601316583,
+      "learning_rate": 1.7337144135146815e-06,
+      "loss": 0.6191,
+      "step": 578
+    },
+    {
+      "epoch": 0.26125211505922163,
+      "grad_norm": 3.338002903883107,
+      "learning_rate": 1.7327203398459584e-06,
+      "loss": 0.5757,
+      "step": 579
+    },
+    {
+      "epoch": 0.2617033276931754,
+      "grad_norm": 3.582565654268005,
+      "learning_rate": 1.731724700273412e-06,
+      "loss": 0.5778,
+      "step": 580
+    },
+    {
+      "epoch": 0.26215454032712915,
+      "grad_norm": 3.3503897003901773,
+      "learning_rate": 1.7307274969248334e-06,
+      "loss": 0.6079,
+      "step": 581
+    },
+    {
+      "epoch": 0.2626057529610829,
+      "grad_norm": 3.6953227055441644,
+      "learning_rate": 1.7297287319313552e-06,
+      "loss": 0.6023,
+      "step": 582
+    },
+    {
+      "epoch": 0.26305696559503666,
+      "grad_norm": 3.2964361664081143,
+      "learning_rate": 1.7287284074274485e-06,
+      "loss": 0.6274,
+      "step": 583
+    },
+    {
+      "epoch": 0.2635081782289904,
+      "grad_norm": 3.347719926763811,
+      "learning_rate": 1.7277265255509163e-06,
+      "loss": 0.5585,
+      "step": 584
+    },
+    {
+      "epoch": 0.2639593908629442,
+      "grad_norm": 2.9228072338236237,
+      "learning_rate": 1.7267230884428903e-06,
+      "loss": 0.5438,
+      "step": 585
+    },
+    {
+      "epoch": 0.26441060349689793,
+      "grad_norm": 2.9981318621426545,
+      "learning_rate": 1.7257180982478254e-06,
+      "loss": 0.4731,
+      "step": 586
+    },
+    {
+      "epoch": 0.2648618161308517,
+      "grad_norm": 2.7051775991961073,
+      "learning_rate": 1.7247115571134968e-06,
+      "loss": 0.5209,
+      "step": 587
+    },
+    {
+      "epoch": 0.2653130287648054,
+      "grad_norm": 3.201881004333221,
+      "learning_rate": 1.7237034671909927e-06,
+      "loss": 0.5185,
+      "step": 588
+    },
+    {
+      "epoch": 0.26576424139875915,
+      "grad_norm": 3.0674421520184803,
+      "learning_rate": 1.7226938306347122e-06,
+      "loss": 0.4585,
+      "step": 589
+    },
+    {
+      "epoch": 0.2662154540327129,
+      "grad_norm": 2.9375079500042136,
+      "learning_rate": 1.7216826496023592e-06,
+      "loss": 0.5657,
+      "step": 590
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 3.526546662395944,
+      "learning_rate": 1.7206699262549394e-06,
+      "loss": 0.6383,
+      "step": 591
+    },
+    {
+      "epoch": 0.2671178793006204,
+      "grad_norm": 3.1723564070744406,
+      "learning_rate": 1.719655662756753e-06,
+      "loss": 0.4663,
+      "step": 592
+    },
+    {
+      "epoch": 0.2675690919345742,
+      "grad_norm": 3.2019446199311017,
+      "learning_rate": 1.7186398612753927e-06,
+      "loss": 0.5142,
+      "step": 593
+    },
+    {
+      "epoch": 0.26802030456852793,
+      "grad_norm": 3.041038824521175,
+      "learning_rate": 1.7176225239817378e-06,
+      "loss": 0.5848,
+      "step": 594
+    },
+    {
+      "epoch": 0.2684715172024817,
+      "grad_norm": 2.9674872867935385,
+      "learning_rate": 1.7166036530499502e-06,
+      "loss": 0.5316,
+      "step": 595
+    },
+    {
+      "epoch": 0.26892272983643545,
+      "grad_norm": 3.127580522901588,
+      "learning_rate": 1.7155832506574686e-06,
+      "loss": 0.6055,
+      "step": 596
+    },
+    {
+      "epoch": 0.26937394247038915,
+      "grad_norm": 3.135126305745904,
+      "learning_rate": 1.7145613189850048e-06,
+      "loss": 0.6033,
+      "step": 597
+    },
+    {
+      "epoch": 0.2698251551043429,
+      "grad_norm": 3.3477656942247425,
+      "learning_rate": 1.713537860216539e-06,
+      "loss": 0.5876,
+      "step": 598
+    },
+    {
+      "epoch": 0.27027636773829666,
+      "grad_norm": 3.2848248111575455,
+      "learning_rate": 1.7125128765393152e-06,
+      "loss": 0.5931,
+      "step": 599
+    },
+    {
+      "epoch": 0.2707275803722504,
+      "grad_norm": 3.4561387987225367,
+      "learning_rate": 1.7114863701438363e-06,
+      "loss": 0.6233,
+      "step": 600
+    },
+    {
+      "epoch": 0.2711787930062042,
+      "grad_norm": 3.4262466490817234,
+      "learning_rate": 1.7104583432238587e-06,
+      "loss": 0.6797,
+      "step": 601
+    },
+    {
+      "epoch": 0.27163000564015793,
+      "grad_norm": 3.333713435824011,
+      "learning_rate": 1.7094287979763891e-06,
+      "loss": 0.6204,
+      "step": 602
+    },
+    {
+      "epoch": 0.2720812182741117,
+      "grad_norm": 3.351871668499492,
+      "learning_rate": 1.7083977366016785e-06,
+      "loss": 0.6415,
+      "step": 603
+    },
+    {
+      "epoch": 0.27253243090806545,
+      "grad_norm": 3.2603491068445587,
+      "learning_rate": 1.7073651613032184e-06,
+      "loss": 0.5484,
+      "step": 604
+    },
+    {
+      "epoch": 0.27298364354201915,
+      "grad_norm": 3.024159212795425,
+      "learning_rate": 1.706331074287736e-06,
+      "loss": 0.6172,
+      "step": 605
+    },
+    {
+      "epoch": 0.2734348561759729,
+      "grad_norm": 2.8744869206183368,
+      "learning_rate": 1.705295477765188e-06,
+      "loss": 0.5581,
+      "step": 606
+    },
+    {
+      "epoch": 0.27388606880992666,
+      "grad_norm": 3.3867448368720483,
+      "learning_rate": 1.7042583739487584e-06,
+      "loss": 0.7357,
+      "step": 607
+    },
+    {
+      "epoch": 0.2743372814438804,
+      "grad_norm": 2.935426978961505,
+      "learning_rate": 1.703219765054852e-06,
+      "loss": 0.5314,
+      "step": 608
+    },
+    {
+      "epoch": 0.2747884940778342,
+      "grad_norm": 3.0973120617238004,
+      "learning_rate": 1.70217965330309e-06,
+      "loss": 0.598,
+      "step": 609
+    },
+    {
+      "epoch": 0.27523970671178793,
+      "grad_norm": 3.028317877474072,
+      "learning_rate": 1.701138040916305e-06,
+      "loss": 0.5377,
+      "step": 610
+    },
+    {
+      "epoch": 0.2756909193457417,
+      "grad_norm": 3.1165117895424723,
+      "learning_rate": 1.7000949301205373e-06,
+      "loss": 0.6877,
+      "step": 611
+    },
+    {
+      "epoch": 0.27614213197969545,
+      "grad_norm": 3.7422765877651516,
+      "learning_rate": 1.6990503231450297e-06,
+      "loss": 0.5404,
+      "step": 612
+    },
+    {
+      "epoch": 0.2765933446136492,
+      "grad_norm": 3.229450690131071,
+      "learning_rate": 1.6980042222222216e-06,
+      "loss": 0.6637,
+      "step": 613
+    },
+    {
+      "epoch": 0.2770445572476029,
+      "grad_norm": 2.9411077856787355,
+      "learning_rate": 1.696956629587745e-06,
+      "loss": 0.6576,
+      "step": 614
+    },
+    {
+      "epoch": 0.27749576988155666,
+      "grad_norm": 3.1585991420412554,
+      "learning_rate": 1.6959075474804215e-06,
+      "loss": 0.5854,
+      "step": 615
+    },
+    {
+      "epoch": 0.2779469825155104,
+      "grad_norm": 3.5738811250095197,
+      "learning_rate": 1.6948569781422538e-06,
+      "loss": 0.6342,
+      "step": 616
+    },
+    {
+      "epoch": 0.2783981951494642,
+      "grad_norm": 3.2043997534775635,
+      "learning_rate": 1.6938049238184244e-06,
+      "loss": 0.6077,
+      "step": 617
+    },
+    {
+      "epoch": 0.27884940778341794,
+      "grad_norm": 3.009802670893003,
+      "learning_rate": 1.6927513867572887e-06,
+      "loss": 0.5461,
+      "step": 618
+    },
+    {
+      "epoch": 0.2793006204173717,
+      "grad_norm": 3.735900498450287,
+      "learning_rate": 1.6916963692103713e-06,
+      "loss": 0.61,
+      "step": 619
+    },
+    {
+      "epoch": 0.27975183305132545,
+      "grad_norm": 3.3631424989752525,
+      "learning_rate": 1.6906398734323606e-06,
+      "loss": 0.7085,
+      "step": 620
+    },
+    {
+      "epoch": 0.2802030456852792,
+      "grad_norm": 3.217240383492006,
+      "learning_rate": 1.6895819016811038e-06,
+      "loss": 0.6147,
+      "step": 621
+    },
+    {
+      "epoch": 0.28065425831923296,
+      "grad_norm": 3.1433256959207885,
+      "learning_rate": 1.6885224562176031e-06,
+      "loss": 0.6808,
+      "step": 622
+    },
+    {
+      "epoch": 0.28110547095318666,
+      "grad_norm": 3.6645785813982856,
+      "learning_rate": 1.6874615393060091e-06,
+      "loss": 0.493,
+      "step": 623
+    },
+    {
+      "epoch": 0.2815566835871404,
+      "grad_norm": 2.9879222448827014,
+      "learning_rate": 1.6863991532136184e-06,
+      "loss": 0.4975,
+      "step": 624
+    },
+    {
+      "epoch": 0.2820078962210942,
+      "grad_norm": 3.373734037289056,
+      "learning_rate": 1.6853353002108667e-06,
+      "loss": 0.5831,
+      "step": 625
+    },
+    {
+      "epoch": 0.28245910885504794,
+      "grad_norm": 3.3186458821895872,
+      "learning_rate": 1.6842699825713242e-06,
+      "loss": 0.5788,
+      "step": 626
+    },
+    {
+      "epoch": 0.2829103214890017,
+      "grad_norm": 3.240295183925887,
+      "learning_rate": 1.683203202571692e-06,
+      "loss": 0.5798,
+      "step": 627
+    },
+    {
+      "epoch": 0.28336153412295545,
+      "grad_norm": 2.8461310264884188,
+      "learning_rate": 1.682134962491796e-06,
+      "loss": 0.624,
+      "step": 628
+    },
+    {
+      "epoch": 0.2838127467569092,
+      "grad_norm": 3.182637425551016,
+      "learning_rate": 1.6810652646145828e-06,
+      "loss": 0.5762,
+      "step": 629
+    },
+    {
+      "epoch": 0.28426395939086296,
+      "grad_norm": 3.21165606325256,
+      "learning_rate": 1.679994111226114e-06,
+      "loss": 0.5685,
+      "step": 630
+    },
+    {
+      "epoch": 0.2847151720248167,
+      "grad_norm": 3.065992459300649,
+      "learning_rate": 1.678921504615562e-06,
+      "loss": 0.5511,
+      "step": 631
+    },
+    {
+      "epoch": 0.2851663846587704,
+      "grad_norm": 4.0157081970954405,
+      "learning_rate": 1.677847447075205e-06,
+      "loss": 0.6992,
+      "step": 632
+    },
+    {
+      "epoch": 0.2856175972927242,
+      "grad_norm": 3.656189168497442,
+      "learning_rate": 1.676771940900422e-06,
+      "loss": 0.6023,
+      "step": 633
+    },
+    {
+      "epoch": 0.28606880992667794,
+      "grad_norm": 3.641835731718804,
+      "learning_rate": 1.6756949883896874e-06,
+      "loss": 0.657,
+      "step": 634
+    },
+    {
+      "epoch": 0.2865200225606317,
+      "grad_norm": 3.631195546639338,
+      "learning_rate": 1.6746165918445672e-06,
+      "loss": 0.5844,
+      "step": 635
+    },
+    {
+      "epoch": 0.28697123519458545,
+      "grad_norm": 3.305073818481503,
+      "learning_rate": 1.6735367535697133e-06,
+      "loss": 0.691,
+      "step": 636
+    },
+    {
+      "epoch": 0.2874224478285392,
+      "grad_norm": 3.6875236525441935,
+      "learning_rate": 1.6724554758728586e-06,
+      "loss": 0.4427,
+      "step": 637
+    },
+    {
+      "epoch": 0.28787366046249296,
+      "grad_norm": 3.270744869677528,
+      "learning_rate": 1.6713727610648122e-06,
+      "loss": 0.6261,
+      "step": 638
+    },
+    {
+      "epoch": 0.2883248730964467,
+      "grad_norm": 3.294150755730515,
+      "learning_rate": 1.670288611459455e-06,
+      "loss": 0.6095,
+      "step": 639
+    },
+    {
+      "epoch": 0.2887760857304005,
+      "grad_norm": 3.418142081910065,
+      "learning_rate": 1.669203029373733e-06,
+      "loss": 0.7189,
+      "step": 640
+    },
+    {
+      "epoch": 0.2892272983643542,
+      "grad_norm": 3.114231217877707,
+      "learning_rate": 1.6681160171276548e-06,
+      "loss": 0.5874,
+      "step": 641
+    },
+    {
+      "epoch": 0.28967851099830794,
+      "grad_norm": 3.3583157977609233,
+      "learning_rate": 1.6670275770442849e-06,
+      "loss": 0.6205,
+      "step": 642
+    },
+    {
+      "epoch": 0.2901297236322617,
+      "grad_norm": 2.946320692741977,
+      "learning_rate": 1.665937711449739e-06,
+      "loss": 0.532,
+      "step": 643
+    },
+    {
+      "epoch": 0.29058093626621545,
+      "grad_norm": 3.4190265435248155,
+      "learning_rate": 1.66484642267318e-06,
+      "loss": 0.5944,
+      "step": 644
+    },
+    {
+      "epoch": 0.2910321489001692,
+      "grad_norm": 2.8472323370866612,
+      "learning_rate": 1.6637537130468113e-06,
+      "loss": 0.6206,
+      "step": 645
+    },
+    {
+      "epoch": 0.29148336153412296,
+      "grad_norm": 3.2713722111872245,
+      "learning_rate": 1.662659584905874e-06,
+      "loss": 0.5323,
+      "step": 646
+    },
+    {
+      "epoch": 0.2919345741680767,
+      "grad_norm": 2.922995795945147,
+      "learning_rate": 1.6615640405886395e-06,
+      "loss": 0.5084,
+      "step": 647
+    },
+    {
+      "epoch": 0.2923857868020305,
+      "grad_norm": 3.5339931194135077,
+      "learning_rate": 1.6604670824364067e-06,
+      "loss": 0.6357,
+      "step": 648
+    },
+    {
+      "epoch": 0.2928369994359842,
+      "grad_norm": 3.317497856029728,
+      "learning_rate": 1.659368712793495e-06,
+      "loss": 0.6874,
+      "step": 649
+    },
+    {
+      "epoch": 0.29328821206993794,
+      "grad_norm": 3.7424311686840577,
+      "learning_rate": 1.6582689340072417e-06,
+      "loss": 0.7115,
+      "step": 650
+    },
+    {
+      "epoch": 0.2937394247038917,
+      "grad_norm": 3.1345128131384383,
+      "learning_rate": 1.6571677484279948e-06,
+      "loss": 0.578,
+      "step": 651
+    },
+    {
+      "epoch": 0.29419063733784545,
+      "grad_norm": 3.911109938851912,
+      "learning_rate": 1.6560651584091082e-06,
+      "loss": 0.7279,
+      "step": 652
+    },
+    {
+      "epoch": 0.2946418499717992,
+      "grad_norm": 3.3288650410464253,
+      "learning_rate": 1.6549611663069383e-06,
+      "loss": 0.5636,
+      "step": 653
+    },
+    {
+      "epoch": 0.29509306260575296,
+      "grad_norm": 2.863150590413283,
+      "learning_rate": 1.6538557744808371e-06,
+      "loss": 0.6478,
+      "step": 654
+    },
+    {
+      "epoch": 0.2955442752397067,
+      "grad_norm": 3.006213042305221,
+      "learning_rate": 1.6527489852931486e-06,
+      "loss": 0.562,
+      "step": 655
+    },
+    {
+      "epoch": 0.2959954878736605,
+      "grad_norm": 3.638111266983312,
+      "learning_rate": 1.6516408011092027e-06,
+      "loss": 0.5496,
+      "step": 656
+    },
+    {
+      "epoch": 0.29644670050761424,
+      "grad_norm": 3.316824374679788,
+      "learning_rate": 1.6505312242973108e-06,
+      "loss": 0.665,
+      "step": 657
+    },
+    {
+      "epoch": 0.29689791314156794,
+      "grad_norm": 3.0712540463962914,
+      "learning_rate": 1.6494202572287605e-06,
+      "loss": 0.5783,
+      "step": 658
+    },
+    {
+      "epoch": 0.2973491257755217,
+      "grad_norm": 3.792785889031542,
+      "learning_rate": 1.64830790227781e-06,
+      "loss": 0.6901,
+      "step": 659
+    },
+    {
+      "epoch": 0.29780033840947545,
+      "grad_norm": 3.420839278819223,
+      "learning_rate": 1.6471941618216842e-06,
+      "loss": 0.617,
+      "step": 660
+    },
+    {
+      "epoch": 0.2982515510434292,
+      "grad_norm": 3.310228590338887,
+      "learning_rate": 1.6460790382405688e-06,
+      "loss": 0.693,
+      "step": 661
+    },
+    {
+      "epoch": 0.29870276367738297,
+      "grad_norm": 3.4270595757910964,
+      "learning_rate": 1.6449625339176053e-06,
+      "loss": 0.5201,
+      "step": 662
+    },
+    {
+      "epoch": 0.2991539763113367,
+      "grad_norm": 3.1683236124814274,
+      "learning_rate": 1.643844651238886e-06,
+      "loss": 0.5696,
+      "step": 663
+    },
+    {
+      "epoch": 0.2996051889452905,
+      "grad_norm": 3.7401623649500473,
+      "learning_rate": 1.6427253925934492e-06,
+      "loss": 0.7044,
+      "step": 664
+    },
+    {
+      "epoch": 0.30005640157924424,
+      "grad_norm": 3.4634585244712124,
+      "learning_rate": 1.641604760373273e-06,
+      "loss": 0.5573,
+      "step": 665
+    },
+    {
+      "epoch": 0.300507614213198,
+      "grad_norm": 3.3656674389925443,
+      "learning_rate": 1.640482756973272e-06,
+      "loss": 0.5742,
+      "step": 666
+    },
+    {
+      "epoch": 0.3009588268471517,
+      "grad_norm": 3.5089959297906654,
+      "learning_rate": 1.6393593847912903e-06,
+      "loss": 0.625,
+      "step": 667
+    },
+    {
+      "epoch": 0.30141003948110545,
+      "grad_norm": 2.8305729096605194,
+      "learning_rate": 1.6382346462280979e-06,
+      "loss": 0.4971,
+      "step": 668
+    },
+    {
+      "epoch": 0.3018612521150592,
+      "grad_norm": 3.2836377236231926,
+      "learning_rate": 1.6371085436873843e-06,
+      "loss": 0.573,
+      "step": 669
+    },
+    {
+      "epoch": 0.30231246474901297,
+      "grad_norm": 2.839883087141276,
+      "learning_rate": 1.635981079575755e-06,
+      "loss": 0.4281,
+      "step": 670
+    },
+    {
+      "epoch": 0.3027636773829667,
+      "grad_norm": 3.0551538272584913,
+      "learning_rate": 1.6348522563027235e-06,
+      "loss": 0.5763,
+      "step": 671
+    },
+    {
+      "epoch": 0.3032148900169205,
+      "grad_norm": 3.1227628457864633,
+      "learning_rate": 1.6337220762807098e-06,
+      "loss": 0.5347,
+      "step": 672
+    },
+    {
+      "epoch": 0.30366610265087424,
+      "grad_norm": 3.0644130436580297,
+      "learning_rate": 1.6325905419250325e-06,
+      "loss": 0.5723,
+      "step": 673
+    },
+    {
+      "epoch": 0.304117315284828,
+      "grad_norm": 3.369765645299152,
+      "learning_rate": 1.631457655653905e-06,
+      "loss": 0.5128,
+      "step": 674
+    },
+    {
+      "epoch": 0.30456852791878175,
+      "grad_norm": 3.130938520045816,
+      "learning_rate": 1.6303234198884294e-06,
+      "loss": 0.5637,
+      "step": 675
+    },
+    {
+      "epoch": 0.30501974055273545,
+      "grad_norm": 2.8738261799739755,
+      "learning_rate": 1.6291878370525925e-06,
+      "loss": 0.5167,
+      "step": 676
+    },
+    {
+      "epoch": 0.3054709531866892,
+      "grad_norm": 3.566044327017721,
+      "learning_rate": 1.6280509095732588e-06,
+      "loss": 0.5657,
+      "step": 677
+    },
+    {
+      "epoch": 0.30592216582064297,
+      "grad_norm": 3.4988087210997554,
+      "learning_rate": 1.6269126398801679e-06,
+      "loss": 0.5211,
+      "step": 678
+    },
+    {
+      "epoch": 0.3063733784545967,
+      "grad_norm": 3.159865034013114,
+      "learning_rate": 1.6257730304059263e-06,
+      "loss": 0.5702,
+      "step": 679
+    },
+    {
+      "epoch": 0.3068245910885505,
+      "grad_norm": 2.765987657389912,
+      "learning_rate": 1.6246320835860052e-06,
+      "loss": 0.5558,
+      "step": 680
+    },
+    {
+      "epoch": 0.30727580372250424,
+      "grad_norm": 3.2675556171262277,
+      "learning_rate": 1.6234898018587336e-06,
+      "loss": 0.5354,
+      "step": 681
+    },
+    {
+      "epoch": 0.307727016356458,
+      "grad_norm": 3.0700923951988,
+      "learning_rate": 1.622346187665292e-06,
+      "loss": 0.4686,
+      "step": 682
+    },
+    {
+      "epoch": 0.30817822899041175,
+      "grad_norm": 2.7395711139916425,
+      "learning_rate": 1.6212012434497101e-06,
+      "loss": 0.5564,
+      "step": 683
+    },
+    {
+      "epoch": 0.3086294416243655,
+      "grad_norm": 2.9605613921627616,
+      "learning_rate": 1.6200549716588595e-06,
+      "loss": 0.4968,
+      "step": 684
+    },
+    {
+      "epoch": 0.3090806542583192,
+      "grad_norm": 3.055443187363974,
+      "learning_rate": 1.6189073747424482e-06,
+      "loss": 0.5546,
+      "step": 685
+    },
+    {
+      "epoch": 0.30953186689227297,
+      "grad_norm": 3.304091566042142,
+      "learning_rate": 1.6177584551530177e-06,
+      "loss": 0.5676,
+      "step": 686
+    },
+    {
+      "epoch": 0.3099830795262267,
+      "grad_norm": 3.127344665872412,
+      "learning_rate": 1.6166082153459346e-06,
+      "loss": 0.5305,
+      "step": 687
+    },
+    {
+      "epoch": 0.3104342921601805,
+      "grad_norm": 3.313157696795387,
+      "learning_rate": 1.6154566577793885e-06,
+      "loss": 0.5159,
+      "step": 688
+    },
+    {
+      "epoch": 0.31088550479413424,
+      "grad_norm": 3.4602680074652548,
+      "learning_rate": 1.6143037849143832e-06,
+      "loss": 0.5878,
+      "step": 689
+    },
+    {
+      "epoch": 0.311336717428088,
+      "grad_norm": 3.47196454330974,
+      "learning_rate": 1.6131495992147359e-06,
+      "loss": 0.6752,
+      "step": 690
+    },
+    {
+      "epoch": 0.31178793006204175,
+      "grad_norm": 2.926369050074452,
+      "learning_rate": 1.6119941031470675e-06,
+      "loss": 0.4326,
+      "step": 691
+    },
+    {
+      "epoch": 0.3122391426959955,
+      "grad_norm": 2.8753404832004272,
+      "learning_rate": 1.6108372991807996e-06,
+      "loss": 0.6389,
+      "step": 692
+    },
+    {
+      "epoch": 0.31269035532994927,
+      "grad_norm": 3.4818672558945845,
+      "learning_rate": 1.6096791897881498e-06,
+      "loss": 0.5769,
+      "step": 693
+    },
+    {
+      "epoch": 0.31314156796390297,
+      "grad_norm": 3.045331557479258,
+      "learning_rate": 1.608519777444125e-06,
+      "loss": 0.5813,
+      "step": 694
+    },
+    {
+      "epoch": 0.3135927805978567,
+      "grad_norm": 3.1173014954702665,
+      "learning_rate": 1.607359064626517e-06,
+      "loss": 0.609,
+      "step": 695
+    },
+    {
+      "epoch": 0.3140439932318105,
+      "grad_norm": 3.1211308380487486,
+      "learning_rate": 1.6061970538158958e-06,
+      "loss": 0.5857,
+      "step": 696
+    },
+    {
+      "epoch": 0.31449520586576424,
+      "grad_norm": 3.6380292500889255,
+      "learning_rate": 1.6050337474956066e-06,
+      "loss": 0.7401,
+      "step": 697
+    },
+    {
+      "epoch": 0.314946418499718,
+      "grad_norm": 3.7195496711700793,
+      "learning_rate": 1.6038691481517628e-06,
+      "loss": 0.6073,
+      "step": 698
+    },
+    {
+      "epoch": 0.31539763113367175,
+      "grad_norm": 3.461343986471695,
+      "learning_rate": 1.6027032582732406e-06,
+      "loss": 0.6371,
+      "step": 699
+    },
+    {
+      "epoch": 0.3158488437676255,
+      "grad_norm": 3.0912912625302384,
+      "learning_rate": 1.6015360803516752e-06,
+      "loss": 0.5852,
+      "step": 700
+    },
+    {
+      "epoch": 0.31630005640157927,
+      "grad_norm": 2.735792616102687,
+      "learning_rate": 1.6003676168814536e-06,
+      "loss": 0.4726,
+      "step": 701
+    },
+    {
+      "epoch": 0.31675126903553297,
+      "grad_norm": 3.2460293296612894,
+      "learning_rate": 1.5991978703597112e-06,
+      "loss": 0.6615,
+      "step": 702
+    },
+    {
+      "epoch": 0.3172024816694867,
+      "grad_norm": 3.2723530181109934,
+      "learning_rate": 1.5980268432863239e-06,
+      "loss": 0.5633,
+      "step": 703
+    },
+    {
+      "epoch": 0.3176536943034405,
+      "grad_norm": 3.1604928857576415,
+      "learning_rate": 1.596854538163906e-06,
+      "loss": 0.5592,
+      "step": 704
+    },
+    {
+      "epoch": 0.31810490693739424,
+      "grad_norm": 3.275606024012947,
+      "learning_rate": 1.5956809574978011e-06,
+      "loss": 0.5976,
+      "step": 705
+    },
+    {
+      "epoch": 0.318556119571348,
+      "grad_norm": 3.66434228008565,
+      "learning_rate": 1.594506103796081e-06,
+      "loss": 0.6421,
+      "step": 706
+    },
+    {
+      "epoch": 0.31900733220530175,
+      "grad_norm": 3.1389856099207183,
+      "learning_rate": 1.5933299795695368e-06,
+      "loss": 0.5529,
+      "step": 707
+    },
+    {
+      "epoch": 0.3194585448392555,
+      "grad_norm": 3.040579790837814,
+      "learning_rate": 1.5921525873316753e-06,
+      "loss": 0.4726,
+      "step": 708
+    },
+    {
+      "epoch": 0.31990975747320927,
+      "grad_norm": 3.3660094819872124,
+      "learning_rate": 1.5909739295987122e-06,
+      "loss": 0.5541,
+      "step": 709
+    },
+    {
+      "epoch": 0.320360970107163,
+      "grad_norm": 3.387903668787852,
+      "learning_rate": 1.5897940088895691e-06,
+      "loss": 0.5683,
+      "step": 710
+    },
+    {
+      "epoch": 0.3208121827411167,
+      "grad_norm": 3.2996751014380554,
+      "learning_rate": 1.5886128277258661e-06,
+      "loss": 0.6318,
+      "step": 711
+    },
+    {
+      "epoch": 0.3212633953750705,
+      "grad_norm": 3.4811079536606004,
+      "learning_rate": 1.5874303886319174e-06,
+      "loss": 0.5905,
+      "step": 712
+    },
+    {
+      "epoch": 0.32171460800902424,
+      "grad_norm": 2.974942368471126,
+      "learning_rate": 1.5862466941347243e-06,
+      "loss": 0.5155,
+      "step": 713
+    },
+    {
+      "epoch": 0.322165820642978,
+      "grad_norm": 3.0452805867406942,
+      "learning_rate": 1.5850617467639727e-06,
+      "loss": 0.5682,
+      "step": 714
+    },
+    {
+      "epoch": 0.32261703327693175,
+      "grad_norm": 3.2224139519675403,
+      "learning_rate": 1.5838755490520249e-06,
+      "loss": 0.6019,
+      "step": 715
+    },
+    {
+      "epoch": 0.3230682459108855,
+      "grad_norm": 2.8944069319837817,
+      "learning_rate": 1.5826881035339154e-06,
+      "loss": 0.515,
+      "step": 716
+    },
+    {
+      "epoch": 0.32351945854483927,
+      "grad_norm": 2.441725821700579,
+      "learning_rate": 1.5814994127473465e-06,
+      "loss": 0.3773,
+      "step": 717
+    },
+    {
+      "epoch": 0.323970671178793,
+      "grad_norm": 3.470342462769914,
+      "learning_rate": 1.5803094792326799e-06,
+      "loss": 0.5145,
+      "step": 718
+    },
+    {
+      "epoch": 0.3244218838127468,
+      "grad_norm": 3.56701743592647,
+      "learning_rate": 1.5791183055329352e-06,
+      "loss": 0.5895,
+      "step": 719
+    },
+    {
+      "epoch": 0.3248730964467005,
+      "grad_norm": 3.123117232719042,
+      "learning_rate": 1.5779258941937803e-06,
+      "loss": 0.6079,
+      "step": 720
+    },
+    {
+      "epoch": 0.32532430908065424,
+      "grad_norm": 3.50214397424285,
+      "learning_rate": 1.5767322477635304e-06,
+      "loss": 0.7037,
+      "step": 721
+    },
+    {
+      "epoch": 0.325775521714608,
+      "grad_norm": 3.7189664252368364,
+      "learning_rate": 1.575537368793138e-06,
+      "loss": 0.554,
+      "step": 722
+    },
+    {
+      "epoch": 0.32622673434856175,
+      "grad_norm": 2.9787465398404858,
+      "learning_rate": 1.5743412598361909e-06,
+      "loss": 0.589,
+      "step": 723
+    },
+    {
+      "epoch": 0.3266779469825155,
+      "grad_norm": 3.322004142520188,
+      "learning_rate": 1.5731439234489052e-06,
+      "loss": 0.5615,
+      "step": 724
+    },
+    {
+      "epoch": 0.32712915961646927,
+      "grad_norm": 3.26274181293833,
+      "learning_rate": 1.571945362190121e-06,
+      "loss": 0.5014,
+      "step": 725
+    },
+    {
+      "epoch": 0.327580372250423,
+      "grad_norm": 3.7220892721550234,
+      "learning_rate": 1.5707455786212944e-06,
+      "loss": 0.6807,
+      "step": 726
+    },
+    {
+      "epoch": 0.3280315848843768,
+      "grad_norm": 2.9903955440183148,
+      "learning_rate": 1.569544575306495e-06,
+      "loss": 0.4905,
+      "step": 727
+    },
+    {
+      "epoch": 0.32848279751833054,
+      "grad_norm": 3.4657777627392936,
+      "learning_rate": 1.5683423548123988e-06,
+      "loss": 0.6067,
+      "step": 728
+    },
+    {
+      "epoch": 0.32893401015228424,
+      "grad_norm": 3.4819746921424004,
+      "learning_rate": 1.5671389197082828e-06,
+      "loss": 0.7603,
+      "step": 729
+    },
+    {
+      "epoch": 0.329385222786238,
+      "grad_norm": 3.3128509975650777,
+      "learning_rate": 1.5659342725660204e-06,
+      "loss": 0.5962,
+      "step": 730
+    },
+    {
+      "epoch": 0.32983643542019175,
+      "grad_norm": 3.111013336637488,
+      "learning_rate": 1.5647284159600742e-06,
+      "loss": 0.5375,
+      "step": 731
+    },
+    {
+      "epoch": 0.3302876480541455,
+      "grad_norm": 3.043014983711972,
+      "learning_rate": 1.5635213524674926e-06,
+      "loss": 0.4457,
+      "step": 732
+    },
+    {
+      "epoch": 0.33073886068809927,
+      "grad_norm": 3.176508138409611,
+      "learning_rate": 1.562313084667903e-06,
+      "loss": 0.6087,
+      "step": 733
+    },
+    {
+      "epoch": 0.331190073322053,
+      "grad_norm": 3.020201486936763,
+      "learning_rate": 1.5611036151435057e-06,
+      "loss": 0.4561,
+      "step": 734
+    },
+    {
+      "epoch": 0.3316412859560068,
+      "grad_norm": 3.0231053359967497,
+      "learning_rate": 1.5598929464790703e-06,
+      "loss": 0.5651,
+      "step": 735
+    },
+    {
+      "epoch": 0.33209249858996054,
+      "grad_norm": 3.275662082022039,
+      "learning_rate": 1.5586810812619291e-06,
+      "loss": 0.5691,
+      "step": 736
+    },
+    {
+      "epoch": 0.3325437112239143,
+      "grad_norm": 3.3993781041664337,
+      "learning_rate": 1.55746802208197e-06,
+      "loss": 0.6204,
+      "step": 737
+    },
+    {
+      "epoch": 0.332994923857868,
+      "grad_norm": 2.995463194362283,
+      "learning_rate": 1.5562537715316349e-06,
+      "loss": 0.564,
+      "step": 738
+    },
+    {
+      "epoch": 0.33344613649182175,
+      "grad_norm": 3.134146072367485,
+      "learning_rate": 1.55503833220591e-06,
+      "loss": 0.5233,
+      "step": 739
+    },
+    {
+      "epoch": 0.3338973491257755,
+      "grad_norm": 3.201637769439807,
+      "learning_rate": 1.553821706702322e-06,
+      "loss": 0.6123,
+      "step": 740
+    },
+    {
+      "epoch": 0.33434856175972927,
+      "grad_norm": 3.232574838103331,
+      "learning_rate": 1.5526038976209343e-06,
+      "loss": 0.5694,
+      "step": 741
+    },
+    {
+      "epoch": 0.334799774393683,
+      "grad_norm": 3.3442531523646544,
+      "learning_rate": 1.5513849075643381e-06,
+      "loss": 0.5079,
+      "step": 742
+    },
+    {
+      "epoch": 0.3352509870276368,
+      "grad_norm": 3.0542429928779216,
+      "learning_rate": 1.550164739137649e-06,
+      "loss": 0.5389,
+      "step": 743
+    },
+    {
+      "epoch": 0.33570219966159054,
+      "grad_norm": 3.1568454912469903,
+      "learning_rate": 1.548943394948501e-06,
+      "loss": 0.5308,
+      "step": 744
+    },
+    {
+      "epoch": 0.3361534122955443,
+      "grad_norm": 3.542836071959044,
+      "learning_rate": 1.5477208776070408e-06,
+      "loss": 0.483,
+      "step": 745
+    },
+    {
+      "epoch": 0.33660462492949805,
+      "grad_norm": 3.1222777445652894,
+      "learning_rate": 1.5464971897259219e-06,
+      "loss": 0.5749,
+      "step": 746
+    },
+    {
+      "epoch": 0.33705583756345175,
+      "grad_norm": 3.628624782183597,
+      "learning_rate": 1.5452723339202998e-06,
+      "loss": 0.5553,
+      "step": 747
+    },
+    {
+      "epoch": 0.3375070501974055,
+      "grad_norm": 3.2488842241245988,
+      "learning_rate": 1.5440463128078261e-06,
+      "loss": 0.6232,
+      "step": 748
+    },
+    {
+      "epoch": 0.33795826283135927,
+      "grad_norm": 3.343015867459291,
+      "learning_rate": 1.5428191290086422e-06,
+      "loss": 0.6445,
+      "step": 749
+    },
+    {
+      "epoch": 0.338409475465313,
+      "grad_norm": 3.1055865259237234,
+      "learning_rate": 1.5415907851453747e-06,
+      "loss": 0.5384,
+      "step": 750
+    },
+    {
+      "epoch": 0.3388606880992668,
+      "grad_norm": 3.2377347656439817,
+      "learning_rate": 1.5403612838431298e-06,
+      "loss": 0.5763,
+      "step": 751
+    },
+    {
+      "epoch": 0.33931190073322054,
+      "grad_norm": 3.445908663047367,
+      "learning_rate": 1.539130627729486e-06,
+      "loss": 0.5514,
+      "step": 752
+    },
+    {
+      "epoch": 0.3397631133671743,
+      "grad_norm": 2.911061775987079,
+      "learning_rate": 1.537898819434491e-06,
+      "loss": 0.7235,
+      "step": 753
+    },
+    {
+      "epoch": 0.34021432600112805,
+      "grad_norm": 2.7941569220035625,
+      "learning_rate": 1.5366658615906545e-06,
+      "loss": 0.6081,
+      "step": 754
+    },
+    {
+      "epoch": 0.34066553863508175,
+      "grad_norm": 3.5600515194912847,
+      "learning_rate": 1.5354317568329425e-06,
+      "loss": 0.5073,
+      "step": 755
+    },
+    {
+      "epoch": 0.3411167512690355,
+      "grad_norm": 3.076725754060217,
+      "learning_rate": 1.5341965077987724e-06,
+      "loss": 0.683,
+      "step": 756
+    },
+    {
+      "epoch": 0.34156796390298927,
+      "grad_norm": 3.3055138999806424,
+      "learning_rate": 1.5329601171280073e-06,
+      "loss": 0.4901,
+      "step": 757
+    },
+    {
+      "epoch": 0.342019176536943,
+      "grad_norm": 3.688970910602998,
+      "learning_rate": 1.5317225874629496e-06,
+      "loss": 0.6957,
+      "step": 758
+    },
+    {
+      "epoch": 0.3424703891708968,
+      "grad_norm": 3.01278353896168,
+      "learning_rate": 1.530483921448336e-06,
+      "loss": 0.5273,
+      "step": 759
+    },
+    {
+      "epoch": 0.34292160180485054,
+      "grad_norm": 2.7371700116966697,
+      "learning_rate": 1.5292441217313322e-06,
+      "loss": 0.5151,
+      "step": 760
+    },
+    {
+      "epoch": 0.3433728144388043,
+      "grad_norm": 2.6873243424543682,
+      "learning_rate": 1.5280031909615261e-06,
+      "loss": 0.5559,
+      "step": 761
+    },
+    {
+      "epoch": 0.34382402707275805,
+      "grad_norm": 3.247322078009431,
+      "learning_rate": 1.5267611317909228e-06,
+      "loss": 0.5374,
+      "step": 762
+    },
+    {
+      "epoch": 0.3442752397067118,
+      "grad_norm": 3.2026837973300974,
+      "learning_rate": 1.5255179468739393e-06,
+      "loss": 0.5152,
+      "step": 763
+    },
+    {
+      "epoch": 0.3447264523406655,
+      "grad_norm": 3.0865094325229734,
+      "learning_rate": 1.5242736388673982e-06,
+      "loss": 0.588,
+      "step": 764
+    },
+    {
+      "epoch": 0.34517766497461927,
+      "grad_norm": 3.0866090252108282,
+      "learning_rate": 1.5230282104305226e-06,
+      "loss": 0.5795,
+      "step": 765
+    },
+    {
+      "epoch": 0.345628877608573,
+      "grad_norm": 3.2541619817207814,
+      "learning_rate": 1.5217816642249296e-06,
+      "loss": 0.6493,
+      "step": 766
+    },
+    {
+      "epoch": 0.3460800902425268,
+      "grad_norm": 3.6862448489806967,
+      "learning_rate": 1.5205340029146253e-06,
+      "loss": 0.586,
+      "step": 767
+    },
+    {
+      "epoch": 0.34653130287648054,
+      "grad_norm": 2.6861569549572737,
+      "learning_rate": 1.519285229165999e-06,
+      "loss": 0.4518,
+      "step": 768
+    },
+    {
+      "epoch": 0.3469825155104343,
+      "grad_norm": 3.039215785994512,
+      "learning_rate": 1.5180353456478173e-06,
+      "loss": 0.4648,
+      "step": 769
+    },
+    {
+      "epoch": 0.34743372814438805,
+      "grad_norm": 3.6176540281690914,
+      "learning_rate": 1.5167843550312188e-06,
+      "loss": 0.6346,
+      "step": 770
+    },
+    {
+      "epoch": 0.3478849407783418,
+      "grad_norm": 3.070746867448483,
+      "learning_rate": 1.5155322599897073e-06,
+      "loss": 0.5604,
+      "step": 771
+    },
+    {
+      "epoch": 0.34833615341229557,
+      "grad_norm": 2.73731135380931,
+      "learning_rate": 1.5142790631991478e-06,
+      "loss": 0.5152,
+      "step": 772
+    },
+    {
+      "epoch": 0.34878736604624927,
+      "grad_norm": 3.529456577457423,
+      "learning_rate": 1.5130247673377587e-06,
+      "loss": 0.5609,
+      "step": 773
+    },
+    {
+      "epoch": 0.349238578680203,
+      "grad_norm": 3.4439646240760826,
+      "learning_rate": 1.5117693750861094e-06,
+      "loss": 0.6688,
+      "step": 774
+    },
+    {
+      "epoch": 0.3496897913141568,
+      "grad_norm": 3.3042570111312335,
+      "learning_rate": 1.51051288912711e-06,
+      "loss": 0.4303,
+      "step": 775
+    },
+    {
+      "epoch": 0.35014100394811054,
+      "grad_norm": 2.9401197668624013,
+      "learning_rate": 1.5092553121460088e-06,
+      "loss": 0.4583,
+      "step": 776
+    },
+    {
+      "epoch": 0.3505922165820643,
+      "grad_norm": 3.1073494998401503,
+      "learning_rate": 1.5079966468303863e-06,
+      "loss": 0.5139,
+      "step": 777
+    },
+    {
+      "epoch": 0.35104342921601805,
+      "grad_norm": 3.118708334430417,
+      "learning_rate": 1.5067368958701485e-06,
+      "loss": 0.4641,
+      "step": 778
+    },
+    {
+      "epoch": 0.3514946418499718,
+      "grad_norm": 3.143071377710741,
+      "learning_rate": 1.5054760619575215e-06,
+      "loss": 0.5753,
+      "step": 779
+    },
+    {
+      "epoch": 0.35194585448392557,
+      "grad_norm": 3.1861569232470766,
+      "learning_rate": 1.5042141477870458e-06,
+      "loss": 0.535,
+      "step": 780
+    },
+    {
+      "epoch": 0.3523970671178793,
+      "grad_norm": 2.8866857302075117,
+      "learning_rate": 1.5029511560555706e-06,
+      "loss": 0.46,
+      "step": 781
+    },
+    {
+      "epoch": 0.352848279751833,
+      "grad_norm": 3.3778859647717443,
+      "learning_rate": 1.5016870894622473e-06,
+      "loss": 0.5898,
+      "step": 782
+    },
+    {
+      "epoch": 0.3532994923857868,
+      "grad_norm": 4.172842058459368,
+      "learning_rate": 1.5004219507085262e-06,
+      "loss": 0.5366,
+      "step": 783
+    },
+    {
+      "epoch": 0.35375070501974054,
+      "grad_norm": 3.59549440239923,
+      "learning_rate": 1.499155742498147e-06,
+      "loss": 0.4775,
+      "step": 784
+    },
+    {
+      "epoch": 0.3542019176536943,
+      "grad_norm": 3.0027968674611825,
+      "learning_rate": 1.4978884675371352e-06,
+      "loss": 0.5956,
+      "step": 785
+    },
+    {
+      "epoch": 0.35465313028764806,
+      "grad_norm": 3.385447063632404,
+      "learning_rate": 1.4966201285337976e-06,
+      "loss": 0.6469,
+      "step": 786
+    },
+    {
+      "epoch": 0.3551043429216018,
+      "grad_norm": 2.8706740349542597,
+      "learning_rate": 1.4953507281987134e-06,
+      "loss": 0.4904,
+      "step": 787
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 3.424910056173964,
+      "learning_rate": 1.4940802692447306e-06,
+      "loss": 0.5708,
+      "step": 788
+    },
+    {
+      "epoch": 0.3560067681895093,
+      "grad_norm": 3.0668549250942174,
+      "learning_rate": 1.4928087543869593e-06,
+      "loss": 0.5102,
+      "step": 789
+    },
+    {
+      "epoch": 0.3564579808234631,
+      "grad_norm": 2.9435846287096594,
+      "learning_rate": 1.491536186342766e-06,
+      "loss": 0.4531,
+      "step": 790
+    },
+    {
+      "epoch": 0.3569091934574168,
+      "grad_norm": 3.3647412629748,
+      "learning_rate": 1.4902625678317694e-06,
+      "loss": 0.5754,
+      "step": 791
+    },
+    {
+      "epoch": 0.35736040609137054,
+      "grad_norm": 3.4720397428714147,
+      "learning_rate": 1.4889879015758317e-06,
+      "loss": 0.5361,
+      "step": 792
+    },
+    {
+      "epoch": 0.3578116187253243,
+      "grad_norm": 3.169387365054564,
+      "learning_rate": 1.4877121902990542e-06,
+      "loss": 0.6178,
+      "step": 793
+    },
+    {
+      "epoch": 0.35826283135927806,
+      "grad_norm": 3.1834769003859416,
+      "learning_rate": 1.4864354367277723e-06,
+      "loss": 0.5135,
+      "step": 794
+    },
+    {
+      "epoch": 0.3587140439932318,
+      "grad_norm": 2.72897080565907,
+      "learning_rate": 1.4851576435905486e-06,
+      "loss": 0.5075,
+      "step": 795
+    },
+    {
+      "epoch": 0.35916525662718557,
+      "grad_norm": 3.472851661152734,
+      "learning_rate": 1.4838788136181674e-06,
+      "loss": 0.481,
+      "step": 796
+    },
+    {
+      "epoch": 0.3596164692611393,
+      "grad_norm": 3.3098456095281343,
+      "learning_rate": 1.4825989495436284e-06,
+      "loss": 0.6186,
+      "step": 797
+    },
+    {
+      "epoch": 0.3600676818950931,
+      "grad_norm": 3.5939015234225318,
+      "learning_rate": 1.4813180541021424e-06,
+      "loss": 0.566,
+      "step": 798
+    },
+    {
+      "epoch": 0.36051889452904684,
+      "grad_norm": 3.19607736166632,
+      "learning_rate": 1.4800361300311227e-06,
+      "loss": 0.6465,
+      "step": 799
+    },
+    {
+      "epoch": 0.36097010716300054,
+      "grad_norm": 3.236820425024368,
+      "learning_rate": 1.4787531800701825e-06,
+      "loss": 0.5826,
+      "step": 800
+    },
+    {
+      "epoch": 0.3614213197969543,
+      "grad_norm": 3.1520319660757927,
+      "learning_rate": 1.4774692069611265e-06,
+      "loss": 0.6347,
+      "step": 801
+    },
+    {
+      "epoch": 0.36187253243090806,
+      "grad_norm": 3.4007171711887203,
+      "learning_rate": 1.4761842134479461e-06,
+      "loss": 0.5826,
+      "step": 802
+    },
+    {
+      "epoch": 0.3623237450648618,
+      "grad_norm": 3.322956321362452,
+      "learning_rate": 1.4748982022768136e-06,
+      "loss": 0.5269,
+      "step": 803
+    },
+    {
+      "epoch": 0.36277495769881557,
+      "grad_norm": 3.5976572006549015,
+      "learning_rate": 1.4736111761960763e-06,
+      "loss": 0.6353,
+      "step": 804
+    },
+    {
+      "epoch": 0.3632261703327693,
+      "grad_norm": 3.358228871970233,
+      "learning_rate": 1.4723231379562503e-06,
+      "loss": 0.639,
+      "step": 805
+    },
+    {
+      "epoch": 0.3636773829667231,
+      "grad_norm": 3.066740811078913,
+      "learning_rate": 1.4710340903100142e-06,
+      "loss": 0.6528,
+      "step": 806
+    },
+    {
+      "epoch": 0.36412859560067684,
+      "grad_norm": 3.143447875416717,
+      "learning_rate": 1.4697440360122046e-06,
+      "loss": 0.5159,
+      "step": 807
+    },
+    {
+      "epoch": 0.36457980823463054,
+      "grad_norm": 3.4609033817493087,
+      "learning_rate": 1.4684529778198095e-06,
+      "loss": 0.6577,
+      "step": 808
+    },
+    {
+      "epoch": 0.3650310208685843,
+      "grad_norm": 3.3073777275985243,
+      "learning_rate": 1.467160918491962e-06,
+      "loss": 0.5486,
+      "step": 809
+    },
+    {
+      "epoch": 0.36548223350253806,
+      "grad_norm": 2.955049404596728,
+      "learning_rate": 1.4658678607899346e-06,
+      "loss": 0.4085,
+      "step": 810
+    },
+    {
+      "epoch": 0.3659334461364918,
+      "grad_norm": 3.1794400164176015,
+      "learning_rate": 1.4645738074771331e-06,
+      "loss": 0.5911,
+      "step": 811
+    },
+    {
+      "epoch": 0.36638465877044557,
+      "grad_norm": 3.282347844957748,
+      "learning_rate": 1.4632787613190927e-06,
+      "loss": 0.6132,
+      "step": 812
+    },
+    {
+      "epoch": 0.3668358714043993,
+      "grad_norm": 3.054992196038826,
+      "learning_rate": 1.461982725083468e-06,
+      "loss": 0.5046,
+      "step": 813
+    },
+    {
+      "epoch": 0.3672870840383531,
+      "grad_norm": 2.7569599731911083,
+      "learning_rate": 1.4606857015400315e-06,
+      "loss": 0.5099,
+      "step": 814
+    },
+    {
+      "epoch": 0.36773829667230684,
+      "grad_norm": 3.2540309485705454,
+      "learning_rate": 1.4593876934606649e-06,
+      "loss": 0.5801,
+      "step": 815
+    },
+    {
+      "epoch": 0.3681895093062606,
+      "grad_norm": 3.2943527058039073,
+      "learning_rate": 1.4580887036193537e-06,
+      "loss": 0.6682,
+      "step": 816
+    },
+    {
+      "epoch": 0.3686407219402143,
+      "grad_norm": 3.133778967882415,
+      "learning_rate": 1.4567887347921816e-06,
+      "loss": 0.6227,
+      "step": 817
+    },
+    {
+      "epoch": 0.36909193457416806,
+      "grad_norm": 3.256206658405506,
+      "learning_rate": 1.4554877897573257e-06,
+      "loss": 0.493,
+      "step": 818
+    },
+    {
+      "epoch": 0.3695431472081218,
+      "grad_norm": 3.182329974272021,
+      "learning_rate": 1.4541858712950475e-06,
+      "loss": 0.5302,
+      "step": 819
+    },
+    {
+      "epoch": 0.36999435984207557,
+      "grad_norm": 3.4363468847418077,
+      "learning_rate": 1.4528829821876898e-06,
+      "loss": 0.6303,
+      "step": 820
+    },
+    {
+      "epoch": 0.37044557247602933,
+      "grad_norm": 2.851895955662126,
+      "learning_rate": 1.45157912521967e-06,
+      "loss": 0.5902,
+      "step": 821
+    },
+    {
+      "epoch": 0.3708967851099831,
+      "grad_norm": 2.9192604758917735,
+      "learning_rate": 1.4502743031774735e-06,
+      "loss": 0.4448,
+      "step": 822
+    },
+    {
+      "epoch": 0.37134799774393684,
+      "grad_norm": 3.1234419282519297,
+      "learning_rate": 1.4489685188496485e-06,
+      "loss": 0.6215,
+      "step": 823
+    },
+    {
+      "epoch": 0.3717992103778906,
+      "grad_norm": 3.2175572399766983,
+      "learning_rate": 1.447661775026799e-06,
+      "loss": 0.5736,
+      "step": 824
+    },
+    {
+      "epoch": 0.37225042301184436,
+      "grad_norm": 3.0464327587263593,
+      "learning_rate": 1.4463540745015804e-06,
+      "loss": 0.4518,
+      "step": 825
+    },
+    {
+      "epoch": 0.37270163564579806,
+      "grad_norm": 2.7548027619798114,
+      "learning_rate": 1.4450454200686922e-06,
+      "loss": 0.5115,
+      "step": 826
+    },
+    {
+      "epoch": 0.3731528482797518,
+      "grad_norm": 3.508978871190512,
+      "learning_rate": 1.4437358145248726e-06,
+      "loss": 0.5587,
+      "step": 827
+    },
+    {
+      "epoch": 0.37360406091370557,
+      "grad_norm": 2.8776633436137082,
+      "learning_rate": 1.4424252606688923e-06,
+      "loss": 0.3882,
+      "step": 828
+    },
+    {
+      "epoch": 0.37405527354765933,
+      "grad_norm": 3.0186645553939444,
+      "learning_rate": 1.4411137613015493e-06,
+      "loss": 0.7998,
+      "step": 829
+    },
+    {
+      "epoch": 0.3745064861816131,
+      "grad_norm": 3.1591798420581148,
+      "learning_rate": 1.4398013192256612e-06,
+      "loss": 0.5591,
+      "step": 830
+    },
+    {
+      "epoch": 0.37495769881556684,
+      "grad_norm": 3.465443632908638,
+      "learning_rate": 1.4384879372460614e-06,
+      "loss": 0.5952,
+      "step": 831
+    },
+    {
+      "epoch": 0.3754089114495206,
+      "grad_norm": 3.983463883157992,
+      "learning_rate": 1.4371736181695906e-06,
+      "loss": 0.6167,
+      "step": 832
+    },
+    {
+      "epoch": 0.37586012408347436,
+      "grad_norm": 3.202176645358091,
+      "learning_rate": 1.4358583648050938e-06,
+      "loss": 0.6324,
+      "step": 833
+    },
+    {
+      "epoch": 0.3763113367174281,
+      "grad_norm": 3.729738183408635,
+      "learning_rate": 1.4345421799634117e-06,
+      "loss": 0.5976,
+      "step": 834
+    },
+    {
+      "epoch": 0.3767625493513818,
+      "grad_norm": 3.2963718228707286,
+      "learning_rate": 1.4332250664573753e-06,
+      "loss": 0.575,
+      "step": 835
+    },
+    {
+      "epoch": 0.37721376198533557,
+      "grad_norm": 3.3781633349933236,
+      "learning_rate": 1.4319070271018015e-06,
+      "loss": 0.6036,
+      "step": 836
+    },
+    {
+      "epoch": 0.37766497461928933,
+      "grad_norm": 3.4896071051826,
+      "learning_rate": 1.4305880647134845e-06,
+      "loss": 0.6199,
+      "step": 837
+    },
+    {
+      "epoch": 0.3781161872532431,
+      "grad_norm": 3.4302159345440826,
+      "learning_rate": 1.4292681821111917e-06,
+      "loss": 0.6029,
+      "step": 838
+    },
+    {
+      "epoch": 0.37856739988719684,
+      "grad_norm": 3.2071682961494195,
+      "learning_rate": 1.4279473821156577e-06,
+      "loss": 0.6575,
+      "step": 839
+    },
+    {
+      "epoch": 0.3790186125211506,
+      "grad_norm": 3.417176487620392,
+      "learning_rate": 1.4266256675495775e-06,
+      "loss": 0.5037,
+      "step": 840
+    },
+    {
+      "epoch": 0.37946982515510436,
+      "grad_norm": 3.1535347582785755,
+      "learning_rate": 1.4253030412375992e-06,
+      "loss": 0.6263,
+      "step": 841
+    },
+    {
+      "epoch": 0.3799210377890581,
+      "grad_norm": 3.2890481570320693,
+      "learning_rate": 1.4239795060063208e-06,
+      "loss": 0.6117,
+      "step": 842
+    },
+    {
+      "epoch": 0.38037225042301187,
+      "grad_norm": 3.48926192769284,
+      "learning_rate": 1.422655064684283e-06,
+      "loss": 0.6099,
+      "step": 843
+    },
+    {
+      "epoch": 0.38082346305696557,
+      "grad_norm": 3.9372420818037006,
+      "learning_rate": 1.4213297201019617e-06,
+      "loss": 0.5924,
+      "step": 844
+    },
+    {
+      "epoch": 0.38127467569091933,
+      "grad_norm": 2.7349996691946203,
+      "learning_rate": 1.4200034750917637e-06,
+      "loss": 0.4804,
+      "step": 845
+    },
+    {
+      "epoch": 0.3817258883248731,
+      "grad_norm": 3.3642312266416075,
+      "learning_rate": 1.4186763324880206e-06,
+      "loss": 0.4911,
+      "step": 846
+    },
+    {
+      "epoch": 0.38217710095882684,
+      "grad_norm": 3.2368382839246657,
+      "learning_rate": 1.4173482951269822e-06,
+      "loss": 0.5709,
+      "step": 847
+    },
+    {
+      "epoch": 0.3826283135927806,
+      "grad_norm": 3.5193539688661972,
+      "learning_rate": 1.4160193658468092e-06,
+      "loss": 0.6303,
+      "step": 848
+    },
+    {
+      "epoch": 0.38307952622673436,
+      "grad_norm": 2.9360735003460574,
+      "learning_rate": 1.4146895474875705e-06,
+      "loss": 0.5032,
+      "step": 849
+    },
+    {
+      "epoch": 0.3835307388606881,
+      "grad_norm": 3.4786066803717497,
+      "learning_rate": 1.4133588428912331e-06,
+      "loss": 0.7037,
+      "step": 850
+    },
+    {
+      "epoch": 0.38398195149464187,
+      "grad_norm": 3.1528126266207415,
+      "learning_rate": 1.412027254901659e-06,
+      "loss": 0.5899,
+      "step": 851
+    },
+    {
+      "epoch": 0.38443316412859563,
+      "grad_norm": 3.174247820940736,
+      "learning_rate": 1.4106947863645982e-06,
+      "loss": 0.4902,
+      "step": 852
+    },
+    {
+      "epoch": 0.38488437676254933,
+      "grad_norm": 3.957992235254075,
+      "learning_rate": 1.4093614401276823e-06,
+      "loss": 0.5707,
+      "step": 853
+    },
+    {
+      "epoch": 0.3853355893965031,
+      "grad_norm": 3.218519092249444,
+      "learning_rate": 1.4080272190404184e-06,
+      "loss": 0.6134,
+      "step": 854
+    },
+    {
+      "epoch": 0.38578680203045684,
+      "grad_norm": 3.1459639474006447,
+      "learning_rate": 1.4066921259541836e-06,
+      "loss": 0.6095,
+      "step": 855
+    },
+    {
+      "epoch": 0.3862380146644106,
+      "grad_norm": 3.5759941707522565,
+      "learning_rate": 1.405356163722218e-06,
+      "loss": 0.6358,
+      "step": 856
+    },
+    {
+      "epoch": 0.38668922729836436,
+      "grad_norm": 3.185073838148989,
+      "learning_rate": 1.4040193351996204e-06,
+      "loss": 0.4632,
+      "step": 857
+    },
+    {
+      "epoch": 0.3871404399323181,
+      "grad_norm": 3.109686641994834,
+      "learning_rate": 1.4026816432433398e-06,
+      "loss": 0.5566,
+      "step": 858
+    },
+    {
+      "epoch": 0.38759165256627187,
+      "grad_norm": 3.2959191557613665,
+      "learning_rate": 1.4013430907121703e-06,
+      "loss": 0.538,
+      "step": 859
+    },
+    {
+      "epoch": 0.38804286520022563,
+      "grad_norm": 3.3692958569410263,
+      "learning_rate": 1.4000036804667462e-06,
+      "loss": 0.5392,
+      "step": 860
+    },
+    {
+      "epoch": 0.38849407783417933,
+      "grad_norm": 2.6745649742861133,
+      "learning_rate": 1.3986634153695342e-06,
+      "loss": 0.3742,
+      "step": 861
+    },
+    {
+      "epoch": 0.3889452904681331,
+      "grad_norm": 3.3623036565879945,
+      "learning_rate": 1.3973222982848281e-06,
+      "loss": 0.7448,
+      "step": 862
+    },
+    {
+      "epoch": 0.38939650310208684,
+      "grad_norm": 3.277299965781224,
+      "learning_rate": 1.3959803320787417e-06,
+      "loss": 0.6002,
+      "step": 863
+    },
+    {
+      "epoch": 0.3898477157360406,
+      "grad_norm": 3.0879829136475814,
+      "learning_rate": 1.394637519619205e-06,
+      "loss": 0.4925,
+      "step": 864
+    },
+    {
+      "epoch": 0.39029892836999436,
+      "grad_norm": 3.365807183133468,
+      "learning_rate": 1.3932938637759552e-06,
+      "loss": 0.5599,
+      "step": 865
+    },
+    {
+      "epoch": 0.3907501410039481,
+      "grad_norm": 3.107049993469059,
+      "learning_rate": 1.3919493674205326e-06,
+      "loss": 0.6655,
+      "step": 866
+    },
+    {
+      "epoch": 0.39120135363790187,
+      "grad_norm": 3.053760677478005,
+      "learning_rate": 1.3906040334262731e-06,
+      "loss": 0.5843,
+      "step": 867
+    },
+    {
+      "epoch": 0.39165256627185563,
+      "grad_norm": 3.093046175711346,
+      "learning_rate": 1.3892578646683037e-06,
+      "loss": 0.6043,
+      "step": 868
+    },
+    {
+      "epoch": 0.3921037789058094,
+      "grad_norm": 2.86893503666343,
+      "learning_rate": 1.3879108640235345e-06,
+      "loss": 0.4397,
+      "step": 869
+    },
+    {
+      "epoch": 0.3925549915397631,
+      "grad_norm": 3.118837171623341,
+      "learning_rate": 1.386563034370654e-06,
+      "loss": 0.5386,
+      "step": 870
+    },
+    {
+      "epoch": 0.39300620417371684,
+      "grad_norm": 3.363263594306241,
+      "learning_rate": 1.3852143785901223e-06,
+      "loss": 0.5624,
+      "step": 871
+    },
+    {
+      "epoch": 0.3934574168076706,
+      "grad_norm": 3.186495282439576,
+      "learning_rate": 1.3838648995641644e-06,
+      "loss": 0.4957,
+      "step": 872
+    },
+    {
+      "epoch": 0.39390862944162436,
+      "grad_norm": 3.3353513396229335,
+      "learning_rate": 1.3825146001767653e-06,
+      "loss": 0.5313,
+      "step": 873
+    },
+    {
+      "epoch": 0.3943598420755781,
+      "grad_norm": 3.0521731392139424,
+      "learning_rate": 1.3811634833136637e-06,
+      "loss": 0.4819,
+      "step": 874
+    },
+    {
+      "epoch": 0.3948110547095319,
+      "grad_norm": 3.1624882499122724,
+      "learning_rate": 1.379811551862344e-06,
+      "loss": 0.5611,
+      "step": 875
+    },
+    {
+      "epoch": 0.39526226734348563,
+      "grad_norm": 2.885520973170412,
+      "learning_rate": 1.378458808712032e-06,
+      "loss": 0.5237,
+      "step": 876
+    },
+    {
+      "epoch": 0.3957134799774394,
+      "grad_norm": 3.1923720267502995,
+      "learning_rate": 1.377105256753689e-06,
+      "loss": 0.6018,
+      "step": 877
+    },
+    {
+      "epoch": 0.39616469261139314,
+      "grad_norm": 3.3170122186250968,
+      "learning_rate": 1.375750898880004e-06,
+      "loss": 0.5146,
+      "step": 878
+    },
+    {
+      "epoch": 0.39661590524534684,
+      "grad_norm": 2.9602277801561336,
+      "learning_rate": 1.3743957379853884e-06,
+      "loss": 0.5429,
+      "step": 879
+    },
+    {
+      "epoch": 0.3970671178793006,
+      "grad_norm": 3.1080024064190432,
+      "learning_rate": 1.3730397769659694e-06,
+      "loss": 0.5487,
+      "step": 880
+    },
+    {
+      "epoch": 0.39751833051325436,
+      "grad_norm": 3.439572751157547,
+      "learning_rate": 1.3716830187195854e-06,
+      "loss": 0.5924,
+      "step": 881
+    },
+    {
+      "epoch": 0.3979695431472081,
+      "grad_norm": 3.005894061607398,
+      "learning_rate": 1.3703254661457772e-06,
+      "loss": 0.5435,
+      "step": 882
+    },
+    {
+      "epoch": 0.3984207557811619,
+      "grad_norm": 2.9201664205070172,
+      "learning_rate": 1.3689671221457838e-06,
+      "loss": 0.6007,
+      "step": 883
+    },
+    {
+      "epoch": 0.39887196841511563,
+      "grad_norm": 3.076358828456211,
+      "learning_rate": 1.3676079896225357e-06,
+      "loss": 0.5791,
+      "step": 884
+    },
+    {
+      "epoch": 0.3993231810490694,
+      "grad_norm": 3.0203537087915278,
+      "learning_rate": 1.3662480714806481e-06,
+      "loss": 0.5749,
+      "step": 885
+    },
+    {
+      "epoch": 0.39977439368302314,
+      "grad_norm": 3.0613670662226538,
+      "learning_rate": 1.3648873706264158e-06,
+      "loss": 0.4396,
+      "step": 886
+    },
+    {
+      "epoch": 0.4002256063169769,
+      "grad_norm": 2.6309532474498574,
+      "learning_rate": 1.363525889967805e-06,
+      "loss": 0.4493,
+      "step": 887
+    },
+    {
+      "epoch": 0.4006768189509306,
+      "grad_norm": 3.208084607808158,
+      "learning_rate": 1.3621636324144507e-06,
+      "loss": 0.4828,
+      "step": 888
+    },
+    {
+      "epoch": 0.40112803158488436,
+      "grad_norm": 3.332508518536951,
+      "learning_rate": 1.3608006008776458e-06,
+      "loss": 0.7003,
+      "step": 889
+    },
+    {
+      "epoch": 0.4015792442188381,
+      "grad_norm": 3.3403690099503907,
+      "learning_rate": 1.3594367982703388e-06,
+      "loss": 0.6317,
+      "step": 890
+    },
+    {
+      "epoch": 0.4020304568527919,
+      "grad_norm": 3.4197414818392273,
+      "learning_rate": 1.3580722275071253e-06,
+      "loss": 0.5383,
+      "step": 891
+    },
+    {
+      "epoch": 0.40248166948674563,
+      "grad_norm": 2.913591421357927,
+      "learning_rate": 1.3567068915042433e-06,
+      "loss": 0.4599,
+      "step": 892
+    },
+    {
+      "epoch": 0.4029328821206994,
+      "grad_norm": 3.2044119627020495,
+      "learning_rate": 1.355340793179566e-06,
+      "loss": 0.5566,
+      "step": 893
+    },
+    {
+      "epoch": 0.40338409475465314,
+      "grad_norm": 3.56074826789299,
+      "learning_rate": 1.3539739354525946e-06,
+      "loss": 0.6223,
+      "step": 894
+    },
+    {
+      "epoch": 0.4038353073886069,
+      "grad_norm": 3.0506608223157787,
+      "learning_rate": 1.352606321244455e-06,
+      "loss": 0.7291,
+      "step": 895
+    },
+    {
+      "epoch": 0.40428652002256066,
+      "grad_norm": 3.06038972236859,
+      "learning_rate": 1.3512379534778882e-06,
+      "loss": 0.667,
+      "step": 896
+    },
+    {
+      "epoch": 0.40473773265651436,
+      "grad_norm": 3.094425625713067,
+      "learning_rate": 1.3498688350772472e-06,
+      "loss": 0.5412,
+      "step": 897
+    },
+    {
+      "epoch": 0.4051889452904681,
+      "grad_norm": 3.6102713876984054,
+      "learning_rate": 1.3484989689684879e-06,
+      "loss": 0.6519,
+      "step": 898
+    },
+    {
+      "epoch": 0.4056401579244219,
+      "grad_norm": 3.054024840623251,
+      "learning_rate": 1.347128358079164e-06,
+      "loss": 0.4936,
+      "step": 899
+    },
+    {
+      "epoch": 0.40609137055837563,
+      "grad_norm": 3.4039347396157145,
+      "learning_rate": 1.3457570053384224e-06,
+      "loss": 0.592,
+      "step": 900
+    },
+    {
+      "epoch": 0.4065425831923294,
+      "grad_norm": 3.3228410861857367,
+      "learning_rate": 1.3443849136769945e-06,
+      "loss": 0.4142,
+      "step": 901
+    },
+    {
+      "epoch": 0.40699379582628314,
+      "grad_norm": 3.5304381752960654,
+      "learning_rate": 1.3430120860271905e-06,
+      "loss": 0.4718,
+      "step": 902
+    },
+    {
+      "epoch": 0.4074450084602369,
+      "grad_norm": 3.303441875394312,
+      "learning_rate": 1.3416385253228937e-06,
+      "loss": 0.5454,
+      "step": 903
+    },
+    {
+      "epoch": 0.40789622109419066,
+      "grad_norm": 3.1629313821139706,
+      "learning_rate": 1.3402642344995542e-06,
+      "loss": 0.5869,
+      "step": 904
+    },
+    {
+      "epoch": 0.4083474337281444,
+      "grad_norm": 3.422560274484405,
+      "learning_rate": 1.3388892164941826e-06,
+      "loss": 0.5956,
+      "step": 905
+    },
+    {
+      "epoch": 0.4087986463620981,
+      "grad_norm": 3.09873469521746,
+      "learning_rate": 1.3375134742453434e-06,
+      "loss": 0.545,
+      "step": 906
+    },
+    {
+      "epoch": 0.4092498589960519,
+      "grad_norm": 3.0994234736000164,
+      "learning_rate": 1.3361370106931485e-06,
+      "loss": 0.6644,
+      "step": 907
+    },
+    {
+      "epoch": 0.40970107163000563,
+      "grad_norm": 3.3133585021955976,
+      "learning_rate": 1.3347598287792518e-06,
+      "loss": 0.5737,
+      "step": 908
+    },
+    {
+      "epoch": 0.4101522842639594,
+      "grad_norm": 3.126477001362301,
+      "learning_rate": 1.3333819314468427e-06,
+      "loss": 0.5358,
+      "step": 909
+    },
+    {
+      "epoch": 0.41060349689791314,
+      "grad_norm": 3.843129597851106,
+      "learning_rate": 1.3320033216406385e-06,
+      "loss": 0.5734,
+      "step": 910
+    },
+    {
+      "epoch": 0.4110547095318669,
+      "grad_norm": 3.0621269496803762,
+      "learning_rate": 1.33062400230688e-06,
+      "loss": 0.4941,
+      "step": 911
+    },
+    {
+      "epoch": 0.41150592216582066,
+      "grad_norm": 3.456611325927705,
+      "learning_rate": 1.3292439763933244e-06,
+      "loss": 0.4588,
+      "step": 912
+    },
+    {
+      "epoch": 0.4119571347997744,
+      "grad_norm": 2.847828708168504,
+      "learning_rate": 1.3278632468492377e-06,
+      "loss": 0.5647,
+      "step": 913
+    },
+    {
+      "epoch": 0.4124083474337281,
+      "grad_norm": 3.108813010015726,
+      "learning_rate": 1.3264818166253916e-06,
+      "loss": 0.4924,
+      "step": 914
+    },
+    {
+      "epoch": 0.4128595600676819,
+      "grad_norm": 3.8277286638201558,
+      "learning_rate": 1.325099688674053e-06,
+      "loss": 0.5561,
+      "step": 915
+    },
+    {
+      "epoch": 0.41331077270163563,
+      "grad_norm": 3.147511650054291,
+      "learning_rate": 1.3237168659489825e-06,
+      "loss": 0.5601,
+      "step": 916
+    },
+    {
+      "epoch": 0.4137619853355894,
+      "grad_norm": 3.0966116731178657,
+      "learning_rate": 1.3223333514054232e-06,
+      "loss": 0.5234,
+      "step": 917
+    },
+    {
+      "epoch": 0.41421319796954315,
+      "grad_norm": 2.5545639358135874,
+      "learning_rate": 1.3209491480000977e-06,
+      "loss": 0.4909,
+      "step": 918
+    },
+    {
+      "epoch": 0.4146644106034969,
+      "grad_norm": 3.556454372294347,
+      "learning_rate": 1.319564258691201e-06,
+      "loss": 0.5788,
+      "step": 919
+    },
+    {
+      "epoch": 0.41511562323745066,
+      "grad_norm": 3.1471280862116564,
+      "learning_rate": 1.3181786864383932e-06,
+      "loss": 0.582,
+      "step": 920
+    },
+    {
+      "epoch": 0.4155668358714044,
+      "grad_norm": 3.343836136967912,
+      "learning_rate": 1.3167924342027944e-06,
+      "loss": 0.7171,
+      "step": 921
+    },
+    {
+      "epoch": 0.4160180485053582,
+      "grad_norm": 2.917522532695626,
+      "learning_rate": 1.3154055049469782e-06,
+      "loss": 0.5267,
+      "step": 922
+    },
+    {
+      "epoch": 0.4164692611393119,
+      "grad_norm": 2.758113115417477,
+      "learning_rate": 1.3140179016349646e-06,
+      "loss": 0.5173,
+      "step": 923
+    },
+    {
+      "epoch": 0.41692047377326563,
+      "grad_norm": 3.3089395229486223,
+      "learning_rate": 1.312629627232214e-06,
+      "loss": 0.5977,
+      "step": 924
+    },
+    {
+      "epoch": 0.4173716864072194,
+      "grad_norm": 3.4613943944882237,
+      "learning_rate": 1.3112406847056212e-06,
+      "loss": 0.5058,
+      "step": 925
+    },
+    {
+      "epoch": 0.41782289904117315,
+      "grad_norm": 3.185867032326023,
+      "learning_rate": 1.3098510770235092e-06,
+      "loss": 0.5305,
+      "step": 926
+    },
+    {
+      "epoch": 0.4182741116751269,
+      "grad_norm": 2.9584569775451963,
+      "learning_rate": 1.308460807155622e-06,
+      "loss": 0.4347,
+      "step": 927
+    },
+    {
+      "epoch": 0.41872532430908066,
+      "grad_norm": 3.0943591120917735,
+      "learning_rate": 1.3070698780731192e-06,
+      "loss": 0.6806,
+      "step": 928
+    },
+    {
+      "epoch": 0.4191765369430344,
+      "grad_norm": 3.257203595590827,
+      "learning_rate": 1.3056782927485688e-06,
+      "loss": 0.5482,
+      "step": 929
+    },
+    {
+      "epoch": 0.4196277495769882,
+      "grad_norm": 3.378797664140071,
+      "learning_rate": 1.3042860541559415e-06,
+      "loss": 0.6118,
+      "step": 930
+    },
+    {
+      "epoch": 0.42007896221094193,
+      "grad_norm": 3.782330800744356,
+      "learning_rate": 1.3028931652706039e-06,
+      "loss": 0.5512,
+      "step": 931
+    },
+    {
+      "epoch": 0.42053017484489563,
+      "grad_norm": 3.2959465895996813,
+      "learning_rate": 1.3014996290693127e-06,
+      "loss": 0.557,
+      "step": 932
+    },
+    {
+      "epoch": 0.4209813874788494,
+      "grad_norm": 3.36676862860587,
+      "learning_rate": 1.3001054485302078e-06,
+      "loss": 0.5585,
+      "step": 933
+    },
+    {
+      "epoch": 0.42143260011280315,
+      "grad_norm": 2.9162822229329355,
+      "learning_rate": 1.2987106266328058e-06,
+      "loss": 0.4515,
+      "step": 934
+    },
+    {
+      "epoch": 0.4218838127467569,
+      "grad_norm": 3.489259111346056,
+      "learning_rate": 1.2973151663579947e-06,
+      "loss": 0.5667,
+      "step": 935
+    },
+    {
+      "epoch": 0.42233502538071066,
+      "grad_norm": 3.146522111454242,
+      "learning_rate": 1.295919070688026e-06,
+      "loss": 0.6089,
+      "step": 936
+    },
+    {
+      "epoch": 0.4227862380146644,
+      "grad_norm": 3.061791688694132,
+      "learning_rate": 1.2945223426065095e-06,
+      "loss": 0.5346,
+      "step": 937
+    },
+    {
+      "epoch": 0.4232374506486182,
+      "grad_norm": 3.0568178693062533,
+      "learning_rate": 1.2931249850984064e-06,
+      "loss": 0.5716,
+      "step": 938
+    },
+    {
+      "epoch": 0.42368866328257193,
+      "grad_norm": 3.375298437835067,
+      "learning_rate": 1.2917270011500232e-06,
+      "loss": 0.5784,
+      "step": 939
+    },
+    {
+      "epoch": 0.4241398759165257,
+      "grad_norm": 3.031854170941789,
+      "learning_rate": 1.2903283937490055e-06,
+      "loss": 0.4996,
+      "step": 940
+    },
+    {
+      "epoch": 0.4245910885504794,
+      "grad_norm": 3.3874162086068726,
+      "learning_rate": 1.2889291658843304e-06,
+      "loss": 0.6241,
+      "step": 941
+    },
+    {
+      "epoch": 0.42504230118443315,
+      "grad_norm": 3.1176739539444913,
+      "learning_rate": 1.2875293205463015e-06,
+      "loss": 0.5855,
+      "step": 942
+    },
+    {
+      "epoch": 0.4254935138183869,
+      "grad_norm": 3.302403976138221,
+      "learning_rate": 1.2861288607265424e-06,
+      "loss": 0.5478,
+      "step": 943
+    },
+    {
+      "epoch": 0.42594472645234066,
+      "grad_norm": 3.3731503961513747,
+      "learning_rate": 1.2847277894179888e-06,
+      "loss": 0.6005,
+      "step": 944
+    },
+    {
+      "epoch": 0.4263959390862944,
+      "grad_norm": 2.9075875109978737,
+      "learning_rate": 1.283326109614885e-06,
+      "loss": 0.5674,
+      "step": 945
+    },
+    {
+      "epoch": 0.4268471517202482,
+      "grad_norm": 3.4745646922579554,
+      "learning_rate": 1.2819238243127735e-06,
+      "loss": 0.5554,
+      "step": 946
+    },
+    {
+      "epoch": 0.42729836435420193,
+      "grad_norm": 3.4374933327923203,
+      "learning_rate": 1.2805209365084925e-06,
+      "loss": 0.5387,
+      "step": 947
+    },
+    {
+      "epoch": 0.4277495769881557,
+      "grad_norm": 3.5485274017082404,
+      "learning_rate": 1.2791174492001675e-06,
+      "loss": 0.5332,
+      "step": 948
+    },
+    {
+      "epoch": 0.42820078962210945,
+      "grad_norm": 2.9948929783352463,
+      "learning_rate": 1.2777133653872048e-06,
+      "loss": 0.4852,
+      "step": 949
+    },
+    {
+      "epoch": 0.42865200225606315,
+      "grad_norm": 3.183845074289996,
+      "learning_rate": 1.2763086880702859e-06,
+      "loss": 0.5381,
+      "step": 950
+    },
+    {
+      "epoch": 0.4291032148900169,
+      "grad_norm": 3.3312664393519524,
+      "learning_rate": 1.2749034202513598e-06,
+      "loss": 0.5461,
+      "step": 951
+    },
+    {
+      "epoch": 0.42955442752397066,
+      "grad_norm": 3.248294129712524,
+      "learning_rate": 1.2734975649336383e-06,
+      "loss": 0.6579,
+      "step": 952
+    },
+    {
+      "epoch": 0.4300056401579244,
+      "grad_norm": 3.650970462515364,
+      "learning_rate": 1.2720911251215896e-06,
+      "loss": 0.5921,
+      "step": 953
+    },
+    {
+      "epoch": 0.4304568527918782,
+      "grad_norm": 3.1484500818228627,
+      "learning_rate": 1.270684103820929e-06,
+      "loss": 0.5272,
+      "step": 954
+    },
+    {
+      "epoch": 0.43090806542583193,
+      "grad_norm": 3.0790541876092665,
+      "learning_rate": 1.2692765040386156e-06,
+      "loss": 0.503,
+      "step": 955
+    },
+    {
+      "epoch": 0.4313592780597857,
+      "grad_norm": 3.2768871039849254,
+      "learning_rate": 1.2678683287828449e-06,
+      "loss": 0.5212,
+      "step": 956
+    },
+    {
+      "epoch": 0.43181049069373945,
+      "grad_norm": 3.4878570977168097,
+      "learning_rate": 1.2664595810630422e-06,
+      "loss": 0.6475,
+      "step": 957
+    },
+    {
+      "epoch": 0.4322617033276932,
+      "grad_norm": 3.259311956538079,
+      "learning_rate": 1.2650502638898558e-06,
+      "loss": 0.5158,
+      "step": 958
+    },
+    {
+      "epoch": 0.4327129159616469,
+      "grad_norm": 2.489346150252523,
+      "learning_rate": 1.2636403802751515e-06,
+      "loss": 0.527,
+      "step": 959
+    },
+    {
+      "epoch": 0.43316412859560066,
+      "grad_norm": 3.227790541090488,
+      "learning_rate": 1.2622299332320047e-06,
+      "loss": 0.5149,
+      "step": 960
+    },
+    {
+      "epoch": 0.4336153412295544,
+      "grad_norm": 3.360641383529355,
+      "learning_rate": 1.2608189257746968e-06,
+      "loss": 0.5772,
+      "step": 961
+    },
+    {
+      "epoch": 0.4340665538635082,
+      "grad_norm": 3.3579225913579283,
+      "learning_rate": 1.2594073609187046e-06,
+      "loss": 0.5674,
+      "step": 962
+    },
+    {
+      "epoch": 0.43451776649746193,
+      "grad_norm": 3.6833581489779696,
+      "learning_rate": 1.2579952416806978e-06,
+      "loss": 0.6361,
+      "step": 963
+    },
+    {
+      "epoch": 0.4349689791314157,
+      "grad_norm": 3.68096763034361,
+      "learning_rate": 1.2565825710785303e-06,
+      "loss": 0.5787,
+      "step": 964
+    },
+    {
+      "epoch": 0.43542019176536945,
+      "grad_norm": 3.200182035241666,
+      "learning_rate": 1.2551693521312338e-06,
+      "loss": 0.4692,
+      "step": 965
+    },
+    {
+      "epoch": 0.4358714043993232,
+      "grad_norm": 3.161219776591041,
+      "learning_rate": 1.2537555878590124e-06,
+      "loss": 0.5817,
+      "step": 966
+    },
+    {
+      "epoch": 0.4363226170332769,
+      "grad_norm": 3.3483981216994434,
+      "learning_rate": 1.2523412812832366e-06,
+      "loss": 0.6038,
+      "step": 967
+    },
+    {
+      "epoch": 0.43677382966723066,
+      "grad_norm": 3.664757783306792,
+      "learning_rate": 1.2509264354264337e-06,
+      "loss": 0.5481,
+      "step": 968
+    },
+    {
+      "epoch": 0.4372250423011844,
+      "grad_norm": 3.0682250506770923,
+      "learning_rate": 1.2495110533122848e-06,
+      "loss": 0.5729,
+      "step": 969
+    },
+    {
+      "epoch": 0.4376762549351382,
+      "grad_norm": 3.276987387331576,
+      "learning_rate": 1.2480951379656173e-06,
+      "loss": 0.5415,
+      "step": 970
+    },
+    {
+      "epoch": 0.43812746756909193,
+      "grad_norm": 2.7598150218875084,
+      "learning_rate": 1.2466786924123977e-06,
+      "loss": 0.546,
+      "step": 971
+    },
+    {
+      "epoch": 0.4385786802030457,
+      "grad_norm": 3.232659097993065,
+      "learning_rate": 1.2452617196797258e-06,
+      "loss": 0.4954,
+      "step": 972
+    },
+    {
+      "epoch": 0.43902989283699945,
+      "grad_norm": 3.389259817026544,
+      "learning_rate": 1.2438442227958274e-06,
+      "loss": 0.5887,
+      "step": 973
+    },
+    {
+      "epoch": 0.4394811054709532,
+      "grad_norm": 3.287684474069178,
+      "learning_rate": 1.2424262047900498e-06,
+      "loss": 0.5423,
+      "step": 974
+    },
+    {
+      "epoch": 0.43993231810490696,
+      "grad_norm": 2.9570593801357603,
+      "learning_rate": 1.2410076686928521e-06,
+      "loss": 0.5879,
+      "step": 975
+    },
+    {
+      "epoch": 0.44038353073886066,
+      "grad_norm": 3.0562220060970104,
+      "learning_rate": 1.2395886175358026e-06,
+      "loss": 0.6012,
+      "step": 976
+    },
+    {
+      "epoch": 0.4408347433728144,
+      "grad_norm": 2.742466631605964,
+      "learning_rate": 1.2381690543515691e-06,
+      "loss": 0.5138,
+      "step": 977
+    },
+    {
+      "epoch": 0.4412859560067682,
+      "grad_norm": 3.508015423427429,
+      "learning_rate": 1.236748982173914e-06,
+      "loss": 0.5654,
+      "step": 978
+    },
+    {
+      "epoch": 0.44173716864072193,
+      "grad_norm": 3.1519521341701604,
+      "learning_rate": 1.2353284040376876e-06,
+      "loss": 0.696,
+      "step": 979
+    },
+    {
+      "epoch": 0.4421883812746757,
+      "grad_norm": 3.199203044906545,
+      "learning_rate": 1.2339073229788214e-06,
+      "loss": 0.5913,
+      "step": 980
+    },
+    {
+      "epoch": 0.44263959390862945,
+      "grad_norm": 2.914392046354155,
+      "learning_rate": 1.2324857420343216e-06,
+      "loss": 0.5423,
+      "step": 981
+    },
+    {
+      "epoch": 0.4430908065425832,
+      "grad_norm": 3.208161429147259,
+      "learning_rate": 1.2310636642422623e-06,
+      "loss": 0.6325,
+      "step": 982
+    },
+    {
+      "epoch": 0.44354201917653696,
+      "grad_norm": 3.222015781659111,
+      "learning_rate": 1.2296410926417804e-06,
+      "loss": 0.4924,
+      "step": 983
+    },
+    {
+      "epoch": 0.4439932318104907,
+      "grad_norm": 3.613081662865641,
+      "learning_rate": 1.228218030273068e-06,
+      "loss": 0.7603,
+      "step": 984
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 2.8395745169932316,
+      "learning_rate": 1.2267944801773648e-06,
+      "loss": 0.4732,
+      "step": 985
+    },
+    {
+      "epoch": 0.4448956570783982,
+      "grad_norm": 3.1488834617622157,
+      "learning_rate": 1.2253704453969541e-06,
+      "loss": 0.5935,
+      "step": 986
+    },
+    {
+      "epoch": 0.44534686971235193,
+      "grad_norm": 2.890352804759586,
+      "learning_rate": 1.2239459289751545e-06,
+      "loss": 0.5563,
+      "step": 987
+    },
+    {
+      "epoch": 0.4457980823463057,
+      "grad_norm": 3.344479914937335,
+      "learning_rate": 1.2225209339563143e-06,
+      "loss": 0.6045,
+      "step": 988
+    },
+    {
+      "epoch": 0.44624929498025945,
+      "grad_norm": 2.9945956060125245,
+      "learning_rate": 1.2210954633858042e-06,
+      "loss": 0.6325,
+      "step": 989
+    },
+    {
+      "epoch": 0.4467005076142132,
+      "grad_norm": 3.0243098524774985,
+      "learning_rate": 1.219669520310011e-06,
+      "loss": 0.5677,
+      "step": 990
+    },
+    {
+      "epoch": 0.44715172024816696,
+      "grad_norm": 2.892447973941234,
+      "learning_rate": 1.2182431077763316e-06,
+      "loss": 0.7021,
+      "step": 991
+    },
+    {
+      "epoch": 0.4476029328821207,
+      "grad_norm": 3.700082520722297,
+      "learning_rate": 1.216816228833167e-06,
+      "loss": 0.6754,
+      "step": 992
+    },
+    {
+      "epoch": 0.4480541455160745,
+      "grad_norm": 3.0843377850883766,
+      "learning_rate": 1.2153888865299133e-06,
+      "loss": 0.5611,
+      "step": 993
+    },
+    {
+      "epoch": 0.4485053581500282,
+      "grad_norm": 3.0386431018180806,
+      "learning_rate": 1.213961083916958e-06,
+      "loss": 0.5822,
+      "step": 994
+    },
+    {
+      "epoch": 0.44895657078398193,
+      "grad_norm": 2.8399741907687392,
+      "learning_rate": 1.2125328240456725e-06,
+      "loss": 0.4906,
+      "step": 995
+    },
+    {
+      "epoch": 0.4494077834179357,
+      "grad_norm": 3.027087230934386,
+      "learning_rate": 1.2111041099684042e-06,
+      "loss": 0.6892,
+      "step": 996
+    },
+    {
+      "epoch": 0.44985899605188945,
+      "grad_norm": 2.9568284939288616,
+      "learning_rate": 1.209674944738473e-06,
+      "loss": 0.6299,
+      "step": 997
+    },
+    {
+      "epoch": 0.4503102086858432,
+      "grad_norm": 3.259765282572653,
+      "learning_rate": 1.2082453314101606e-06,
+      "loss": 0.4755,
+      "step": 998
+    },
+    {
+      "epoch": 0.45076142131979696,
+      "grad_norm": 3.447951905086282,
+      "learning_rate": 1.2068152730387088e-06,
+      "loss": 0.6369,
+      "step": 999
+    },
+    {
+      "epoch": 0.4512126339537507,
+      "grad_norm": 3.234955570056143,
+      "learning_rate": 1.2053847726803089e-06,
+      "loss": 0.5156,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4516638465877045,
+      "grad_norm": 3.5463128389089333,
+      "learning_rate": 1.203953833392097e-06,
+      "loss": 0.475,
+      "step": 1001
+    },
+    {
+      "epoch": 0.45211505922165823,
+      "grad_norm": 3.222070894972768,
+      "learning_rate": 1.2025224582321485e-06,
+      "loss": 0.609,
+      "step": 1002
+    },
+    {
+      "epoch": 0.45256627185561193,
+      "grad_norm": 3.0758689339986,
+      "learning_rate": 1.201090650259468e-06,
+      "loss": 0.5973,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4530174844895657,
+      "grad_norm": 3.159127568590662,
+      "learning_rate": 1.1996584125339868e-06,
+      "loss": 0.5238,
+      "step": 1004
+    },
+    {
+      "epoch": 0.45346869712351945,
+      "grad_norm": 3.6135720048902105,
+      "learning_rate": 1.1982257481165545e-06,
+      "loss": 0.675,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4539199097574732,
+      "grad_norm": 3.336465519438658,
+      "learning_rate": 1.1967926600689321e-06,
+      "loss": 0.542,
+      "step": 1006
+    },
+    {
+      "epoch": 0.45437112239142696,
+      "grad_norm": 2.7490928983502854,
+      "learning_rate": 1.1953591514537864e-06,
+      "loss": 0.4707,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4548223350253807,
+      "grad_norm": 3.3908649442402714,
+      "learning_rate": 1.193925225334682e-06,
+      "loss": 0.602,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4552735476593345,
+      "grad_norm": 2.894962512757854,
+      "learning_rate": 1.1924908847760772e-06,
+      "loss": 0.6097,
+      "step": 1009
+    },
+    {
+      "epoch": 0.45572476029328823,
+      "grad_norm": 3.4842483249570155,
+      "learning_rate": 1.191056132843315e-06,
+      "loss": 0.5953,
+      "step": 1010
+    },
+    {
+      "epoch": 0.45617597292724194,
+      "grad_norm": 2.950398511688371,
+      "learning_rate": 1.1896209726026176e-06,
+      "loss": 0.4035,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4566271855611957,
+      "grad_norm": 2.9513437535271185,
+      "learning_rate": 1.1881854071210804e-06,
+      "loss": 0.5398,
+      "step": 1012
+    },
+    {
+      "epoch": 0.45707839819514945,
+      "grad_norm": 2.8613177638799945,
+      "learning_rate": 1.186749439466664e-06,
+      "loss": 0.5525,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4575296108291032,
+      "grad_norm": 3.188791154585419,
+      "learning_rate": 1.1853130727081893e-06,
+      "loss": 0.5192,
+      "step": 1014
+    },
+    {
+      "epoch": 0.45798082346305696,
+      "grad_norm": 3.3146247782552756,
+      "learning_rate": 1.18387630991533e-06,
+      "loss": 0.5131,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4584320360970107,
+      "grad_norm": 3.9250211014882517,
+      "learning_rate": 1.1824391541586055e-06,
+      "loss": 0.5914,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4588832487309645,
+      "grad_norm": 2.9161098414160733,
+      "learning_rate": 1.1810016085093754e-06,
+      "loss": 0.5391,
+      "step": 1017
+    },
+    {
+      "epoch": 0.45933446136491823,
+      "grad_norm": 2.9799917015998987,
+      "learning_rate": 1.179563676039833e-06,
+      "loss": 0.5123,
+      "step": 1018
+    },
+    {
+      "epoch": 0.459785673998872,
+      "grad_norm": 3.530754717880453,
+      "learning_rate": 1.1781253598229982e-06,
+      "loss": 0.5318,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4602368866328257,
+      "grad_norm": 3.375071447655234,
+      "learning_rate": 1.1766866629327097e-06,
+      "loss": 0.5517,
+      "step": 1020
+    },
+    {
+      "epoch": 0.46068809926677945,
+      "grad_norm": 3.388282264796258,
+      "learning_rate": 1.1752475884436213e-06,
+      "loss": 0.5301,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4611393119007332,
+      "grad_norm": 3.4203055686756456,
+      "learning_rate": 1.1738081394311932e-06,
+      "loss": 0.6649,
+      "step": 1022
+    },
+    {
+      "epoch": 0.46159052453468696,
+      "grad_norm": 3.3102903920545232,
+      "learning_rate": 1.172368318971686e-06,
+      "loss": 0.5006,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4620417371686407,
+      "grad_norm": 3.083071988078867,
+      "learning_rate": 1.170928130142154e-06,
+      "loss": 0.4744,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4624929498025945,
+      "grad_norm": 3.348114186262164,
+      "learning_rate": 1.169487576020439e-06,
+      "loss": 0.5852,
+      "step": 1025
+    },
+    {
+      "epoch": 0.46294416243654823,
+      "grad_norm": 3.5395494913322696,
+      "learning_rate": 1.1680466596851635e-06,
+      "loss": 0.5319,
+      "step": 1026
+    },
+    {
+      "epoch": 0.463395375070502,
+      "grad_norm": 3.5359128839904366,
+      "learning_rate": 1.1666053842157232e-06,
+      "loss": 0.5107,
+      "step": 1027
+    },
+    {
+      "epoch": 0.46384658770445575,
+      "grad_norm": 2.6657070438945185,
+      "learning_rate": 1.165163752692283e-06,
+      "loss": 0.5478,
+      "step": 1028
+    },
+    {
+      "epoch": 0.46429780033840945,
+      "grad_norm": 3.0149910558712243,
+      "learning_rate": 1.1637217681957673e-06,
+      "loss": 0.5872,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4647490129723632,
+      "grad_norm": 3.159458756633235,
+      "learning_rate": 1.1622794338078552e-06,
+      "loss": 0.706,
+      "step": 1030
+    },
+    {
+      "epoch": 0.46520022560631696,
+      "grad_norm": 3.0885082043510654,
+      "learning_rate": 1.1608367526109736e-06,
+      "loss": 0.593,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4656514382402707,
+      "grad_norm": 2.9895356676399625,
+      "learning_rate": 1.159393727688291e-06,
+      "loss": 0.5294,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4661026508742245,
+      "grad_norm": 3.2775855911044594,
+      "learning_rate": 1.1579503621237101e-06,
+      "loss": 0.4458,
+      "step": 1033
+    },
+    {
+      "epoch": 0.46655386350817823,
+      "grad_norm": 3.1259902784446894,
+      "learning_rate": 1.1565066590018613e-06,
+      "loss": 0.4544,
+      "step": 1034
+    },
+    {
+      "epoch": 0.467005076142132,
+      "grad_norm": 3.099407676118007,
+      "learning_rate": 1.1550626214080965e-06,
+      "loss": 0.5772,
+      "step": 1035
+    },
+    {
+      "epoch": 0.46745628877608575,
+      "grad_norm": 2.9661310625580333,
+      "learning_rate": 1.1536182524284833e-06,
+      "loss": 0.4604,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4679075014100395,
+      "grad_norm": 3.3426717260171976,
+      "learning_rate": 1.1521735551497966e-06,
+      "loss": 0.6421,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4683587140439932,
+      "grad_norm": 2.976762713490125,
+      "learning_rate": 1.1507285326595126e-06,
+      "loss": 0.597,
+      "step": 1038
+    },
+    {
+      "epoch": 0.46880992667794696,
+      "grad_norm": 3.30725401073655,
+      "learning_rate": 1.1492831880458037e-06,
+      "loss": 0.6313,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4692611393119007,
+      "grad_norm": 2.9639877946890327,
+      "learning_rate": 1.1478375243975295e-06,
+      "loss": 0.4938,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4697123519458545,
+      "grad_norm": 2.786610484320687,
+      "learning_rate": 1.1463915448042326e-06,
+      "loss": 0.4515,
+      "step": 1041
+    },
+    {
+      "epoch": 0.47016356457980824,
+      "grad_norm": 3.135745692122038,
+      "learning_rate": 1.1449452523561294e-06,
+      "loss": 0.5554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.470614777213762,
+      "grad_norm": 2.788435273509232,
+      "learning_rate": 1.143498650144106e-06,
+      "loss": 0.4914,
+      "step": 1043
+    },
+    {
+      "epoch": 0.47106598984771575,
+      "grad_norm": 2.876792135265133,
+      "learning_rate": 1.1420517412597105e-06,
+      "loss": 0.4697,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4715172024816695,
+      "grad_norm": 3.5477556615090093,
+      "learning_rate": 1.1406045287951457e-06,
+      "loss": 0.5816,
+      "step": 1045
+    },
+    {
+      "epoch": 0.47196841511562326,
+      "grad_norm": 2.92573936243044,
+      "learning_rate": 1.1391570158432635e-06,
+      "loss": 0.4241,
+      "step": 1046
+    },
+    {
+      "epoch": 0.47241962774957696,
+      "grad_norm": 3.2379205245375347,
+      "learning_rate": 1.1377092054975583e-06,
+      "loss": 0.5848,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4728708403835307,
+      "grad_norm": 3.2602568198273736,
+      "learning_rate": 1.1362611008521596e-06,
+      "loss": 0.5032,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4733220530174845,
+      "grad_norm": 3.563180746954425,
+      "learning_rate": 1.134812705001826e-06,
+      "loss": 0.6199,
+      "step": 1049
+    },
+    {
+      "epoch": 0.47377326565143824,
+      "grad_norm": 3.258258686215829,
+      "learning_rate": 1.1333640210419386e-06,
+      "loss": 0.6113,
+      "step": 1050
+    },
+    {
+      "epoch": 0.474224478285392,
+      "grad_norm": 3.2628377254999297,
+      "learning_rate": 1.1319150520684944e-06,
+      "loss": 0.5243,
+      "step": 1051
+    },
+    {
+      "epoch": 0.47467569091934575,
+      "grad_norm": 3.027101046392854,
+      "learning_rate": 1.1304658011780984e-06,
+      "loss": 0.5171,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4751269035532995,
+      "grad_norm": 3.649012114499818,
+      "learning_rate": 1.1290162714679594e-06,
+      "loss": 0.6611,
+      "step": 1053
+    },
+    {
+      "epoch": 0.47557811618725326,
+      "grad_norm": 3.1689567292584266,
+      "learning_rate": 1.1275664660358817e-06,
+      "loss": 0.3962,
+      "step": 1054
+    },
+    {
+      "epoch": 0.476029328821207,
+      "grad_norm": 2.92166316898993,
+      "learning_rate": 1.1261163879802587e-06,
+      "loss": 0.4777,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4764805414551607,
+      "grad_norm": 3.137056987986139,
+      "learning_rate": 1.1246660404000658e-06,
+      "loss": 0.5466,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4769317540891145,
+      "grad_norm": 3.382936110474769,
+      "learning_rate": 1.1232154263948556e-06,
+      "loss": 0.5077,
+      "step": 1057
+    },
+    {
+      "epoch": 0.47738296672306824,
+      "grad_norm": 3.517217173416557,
+      "learning_rate": 1.1217645490647494e-06,
+      "loss": 0.5492,
+      "step": 1058
+    },
+    {
+      "epoch": 0.477834179357022,
+      "grad_norm": 3.26075070325087,
+      "learning_rate": 1.1203134115104315e-06,
+      "loss": 0.5016,
+      "step": 1059
+    },
+    {
+      "epoch": 0.47828539199097575,
+      "grad_norm": 3.2324066506117,
+      "learning_rate": 1.1188620168331419e-06,
+      "loss": 0.5466,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4787366046249295,
+      "grad_norm": 2.9767488512886975,
+      "learning_rate": 1.1174103681346708e-06,
+      "loss": 0.5206,
+      "step": 1061
+    },
+    {
+      "epoch": 0.47918781725888326,
+      "grad_norm": 3.1883816288556166,
+      "learning_rate": 1.1159584685173505e-06,
+      "loss": 0.5209,
+      "step": 1062
+    },
+    {
+      "epoch": 0.479639029892837,
+      "grad_norm": 3.2725897850315304,
+      "learning_rate": 1.11450632108405e-06,
+      "loss": 0.6575,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4800902425267907,
+      "grad_norm": 2.90226955653477,
+      "learning_rate": 1.113053928938168e-06,
+      "loss": 0.526,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4805414551607445,
+      "grad_norm": 3.2118769622065724,
+      "learning_rate": 1.1116012951836255e-06,
+      "loss": 0.5494,
+      "step": 1065
+    },
+    {
+      "epoch": 0.48099266779469824,
+      "grad_norm": 2.9794708120012996,
+      "learning_rate": 1.110148422924861e-06,
+      "loss": 0.6493,
+      "step": 1066
+    },
+    {
+      "epoch": 0.481443880428652,
+      "grad_norm": 3.36225003487691,
+      "learning_rate": 1.1086953152668216e-06,
+      "loss": 0.5022,
+      "step": 1067
+    },
+    {
+      "epoch": 0.48189509306260575,
+      "grad_norm": 3.2803450847662754,
+      "learning_rate": 1.1072419753149585e-06,
+      "loss": 0.6551,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4823463056965595,
+      "grad_norm": 3.210329635697201,
+      "learning_rate": 1.1057884061752176e-06,
+      "loss": 0.5731,
+      "step": 1069
+    },
+    {
+      "epoch": 0.48279751833051326,
+      "grad_norm": 3.5448326400292514,
+      "learning_rate": 1.1043346109540369e-06,
+      "loss": 0.5602,
+      "step": 1070
+    },
+    {
+      "epoch": 0.483248730964467,
+      "grad_norm": 3.2599148247025354,
+      "learning_rate": 1.102880592758336e-06,
+      "loss": 0.6967,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4836999435984208,
+      "grad_norm": 3.921908291356212,
+      "learning_rate": 1.1014263546955115e-06,
+      "loss": 0.6903,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4841511562323745,
+      "grad_norm": 3.6723472952642098,
+      "learning_rate": 1.0999718998734298e-06,
+      "loss": 0.7113,
+      "step": 1073
+    },
+    {
+      "epoch": 0.48460236886632824,
+      "grad_norm": 3.1199912636971767,
+      "learning_rate": 1.0985172314004203e-06,
+      "loss": 0.5092,
+      "step": 1074
+    },
+    {
+      "epoch": 0.485053581500282,
+      "grad_norm": 3.084748288084798,
+      "learning_rate": 1.0970623523852698e-06,
+      "loss": 0.5553,
+      "step": 1075
+    },
+    {
+      "epoch": 0.48550479413423575,
+      "grad_norm": 2.8776564795067556,
+      "learning_rate": 1.0956072659372141e-06,
+      "loss": 0.5115,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4859560067681895,
+      "grad_norm": 2.953441401470534,
+      "learning_rate": 1.094151975165933e-06,
+      "loss": 0.6507,
+      "step": 1077
+    },
+    {
+      "epoch": 0.48640721940214326,
+      "grad_norm": 3.690371591944593,
+      "learning_rate": 1.0926964831815424e-06,
+      "loss": 0.6129,
+      "step": 1078
+    },
+    {
+      "epoch": 0.486858432036097,
+      "grad_norm": 2.9774725768337387,
+      "learning_rate": 1.0912407930945887e-06,
+      "loss": 0.6598,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4873096446700508,
+      "grad_norm": 3.304695556663853,
+      "learning_rate": 1.0897849080160411e-06,
+      "loss": 0.596,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48776085730400454,
+      "grad_norm": 3.5963573435316367,
+      "learning_rate": 1.0883288310572862e-06,
+      "loss": 0.6246,
+      "step": 1081
+    },
+    {
+      "epoch": 0.48821206993795824,
+      "grad_norm": 3.155896819738853,
+      "learning_rate": 1.0868725653301205e-06,
+      "loss": 0.5807,
+      "step": 1082
+    },
+    {
+      "epoch": 0.488663282571912,
+      "grad_norm": 3.081080659400599,
+      "learning_rate": 1.0854161139467435e-06,
+      "loss": 0.434,
+      "step": 1083
+    },
+    {
+      "epoch": 0.48911449520586575,
+      "grad_norm": 3.106854377639036,
+      "learning_rate": 1.0839594800197516e-06,
+      "loss": 0.48,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4895657078398195,
+      "grad_norm": 3.7755265948586825,
+      "learning_rate": 1.082502666662132e-06,
+      "loss": 0.6046,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49001692047377327,
+      "grad_norm": 3.1538567782545153,
+      "learning_rate": 1.0810456769872542e-06,
+      "loss": 0.4631,
+      "step": 1086
+    },
+    {
+      "epoch": 0.490468133107727,
+      "grad_norm": 3.493792209571887,
+      "learning_rate": 1.0795885141088652e-06,
+      "loss": 0.4979,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4909193457416808,
+      "grad_norm": 3.2677544727924386,
+      "learning_rate": 1.0781311811410825e-06,
+      "loss": 0.4784,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49137055837563454,
+      "grad_norm": 3.2385254672960335,
+      "learning_rate": 1.0766736811983863e-06,
+      "loss": 0.6579,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4918217710095883,
+      "grad_norm": 3.0410160191967424,
+      "learning_rate": 1.0752160173956144e-06,
+      "loss": 0.4921,
+      "step": 1090
+    },
+    {
+      "epoch": 0.492272983643542,
+      "grad_norm": 3.0814337139592958,
+      "learning_rate": 1.0737581928479538e-06,
+      "loss": 0.4993,
+      "step": 1091
+    },
+    {
+      "epoch": 0.49272419627749575,
+      "grad_norm": 3.1302147016084336,
+      "learning_rate": 1.0723002106709363e-06,
+      "loss": 0.4858,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4931754089114495,
+      "grad_norm": 3.2032365657032846,
+      "learning_rate": 1.0708420739804294e-06,
+      "loss": 0.4662,
+      "step": 1093
+    },
+    {
+      "epoch": 0.49362662154540327,
+      "grad_norm": 2.963453928384545,
+      "learning_rate": 1.0693837858926315e-06,
+      "loss": 0.4786,
+      "step": 1094
+    },
+    {
+      "epoch": 0.494077834179357,
+      "grad_norm": 2.8479265985444835,
+      "learning_rate": 1.0679253495240645e-06,
+      "loss": 0.6417,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4945290468133108,
+      "grad_norm": 3.8316009838064153,
+      "learning_rate": 1.066466767991567e-06,
+      "loss": 0.6002,
+      "step": 1096
+    },
+    {
+      "epoch": 0.49498025944726454,
+      "grad_norm": 3.13289524958788,
+      "learning_rate": 1.0650080444122875e-06,
+      "loss": 0.6372,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4954314720812183,
+      "grad_norm": 3.2639667702613253,
+      "learning_rate": 1.0635491819036792e-06,
+      "loss": 0.6167,
+      "step": 1098
+    },
+    {
+      "epoch": 0.49588268471517205,
+      "grad_norm": 3.049433076188395,
+      "learning_rate": 1.0620901835834912e-06,
+      "loss": 0.5683,
+      "step": 1099
+    },
+    {
+      "epoch": 0.49633389734912575,
+      "grad_norm": 3.387030164756313,
+      "learning_rate": 1.0606310525697627e-06,
+      "loss": 0.6261,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4967851099830795,
+      "grad_norm": 3.402149165888981,
+      "learning_rate": 1.059171791980817e-06,
+      "loss": 0.4562,
+      "step": 1101
+    },
+    {
+      "epoch": 0.49723632261703327,
+      "grad_norm": 3.2602762733367996,
+      "learning_rate": 1.0577124049352548e-06,
+      "loss": 0.626,
+      "step": 1102
+    },
+    {
+      "epoch": 0.497687535250987,
+      "grad_norm": 3.2610142828869777,
+      "learning_rate": 1.0562528945519461e-06,
+      "loss": 0.5353,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4981387478849408,
+      "grad_norm": 2.97682798873228,
+      "learning_rate": 1.0547932639500246e-06,
+      "loss": 0.4905,
+      "step": 1104
+    },
+    {
+      "epoch": 0.49858996051889454,
+      "grad_norm": 2.938963109880475,
+      "learning_rate": 1.0533335162488815e-06,
+      "loss": 0.4562,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4990411731528483,
+      "grad_norm": 3.5017181297984012,
+      "learning_rate": 1.051873654568158e-06,
+      "loss": 0.5793,
+      "step": 1106
+    },
+    {
+      "epoch": 0.49949238578680205,
+      "grad_norm": 2.7407269861856736,
+      "learning_rate": 1.0504136820277384e-06,
+      "loss": 0.5171,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4999435984207558,
+      "grad_norm": 3.2128618819706114,
+      "learning_rate": 1.0489536017477448e-06,
+      "loss": 0.4742,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5003948110547095,
+      "grad_norm": 3.2860218544233737,
+      "learning_rate": 1.0474934168485288e-06,
+      "loss": 0.5809,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5008460236886633,
+      "grad_norm": 3.1537779519919162,
+      "learning_rate": 1.0460331304506655e-06,
+      "loss": 0.5523,
+      "step": 1110
+    },
+    {
+      "epoch": 0.501297236322617,
+      "grad_norm": 2.9629163520820225,
+      "learning_rate": 1.0445727456749483e-06,
+      "loss": 0.5112,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5017484489565708,
+      "grad_norm": 3.197935737383708,
+      "learning_rate": 1.043112265642379e-06,
+      "loss": 0.4732,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5021996615905245,
+      "grad_norm": 2.9033918431675922,
+      "learning_rate": 1.041651693474164e-06,
+      "loss": 0.4971,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5026508742244783,
+      "grad_norm": 3.1684252831494213,
+      "learning_rate": 1.0401910322917064e-06,
+      "loss": 0.5758,
+      "step": 1114
+    },
+    {
+      "epoch": 0.503102086858432,
+      "grad_norm": 2.980345483564053,
+      "learning_rate": 1.0387302852165998e-06,
+      "loss": 0.4851,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5035532994923858,
+      "grad_norm": 3.36987317517813,
+      "learning_rate": 1.037269455370621e-06,
+      "loss": 0.5245,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5040045121263396,
+      "grad_norm": 3.2674734863325945,
+      "learning_rate": 1.035808545875723e-06,
+      "loss": 0.587,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5044557247602933,
+      "grad_norm": 3.089508629837757,
+      "learning_rate": 1.0343475598540307e-06,
+      "loss": 0.4585,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5049069373942471,
+      "grad_norm": 3.058502135032177,
+      "learning_rate": 1.0328865004278315e-06,
+      "loss": 0.62,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5053581500282008,
+      "grad_norm": 3.2132492413810305,
+      "learning_rate": 1.0314253707195703e-06,
+      "loss": 0.5984,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5058093626621546,
+      "grad_norm": 2.9972219397833637,
+      "learning_rate": 1.0299641738518405e-06,
+      "loss": 0.6045,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5062605752961082,
+      "grad_norm": 2.809038897639686,
+      "learning_rate": 1.0285029129473813e-06,
+      "loss": 0.5555,
+      "step": 1122
+    },
+    {
+      "epoch": 0.506711787930062,
+      "grad_norm": 2.659291698186336,
+      "learning_rate": 1.0270415911290671e-06,
+      "loss": 0.5977,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5071630005640158,
+      "grad_norm": 3.0309037627722706,
+      "learning_rate": 1.0255802115199032e-06,
+      "loss": 0.5541,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5076142131979695,
+      "grad_norm": 3.0510921385148646,
+      "learning_rate": 1.024118777243019e-06,
+      "loss": 0.6471,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5080654258319233,
+      "grad_norm": 3.3253019574272176,
+      "learning_rate": 1.022657291421659e-06,
+      "loss": 0.5083,
+      "step": 1126
+    },
+    {
+      "epoch": 0.508516638465877,
+      "grad_norm": 3.1624347051422115,
+      "learning_rate": 1.0211957571791795e-06,
+      "loss": 0.5804,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5089678510998308,
+      "grad_norm": 3.082831702328242,
+      "learning_rate": 1.019734177639039e-06,
+      "loss": 0.5742,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5094190637337845,
+      "grad_norm": 2.855577550248334,
+      "learning_rate": 1.0182725559247945e-06,
+      "loss": 0.5888,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5098702763677383,
+      "grad_norm": 3.5077212889613967,
+      "learning_rate": 1.0168108951600915e-06,
+      "loss": 0.5265,
+      "step": 1130
+    },
+    {
+      "epoch": 0.510321489001692,
+      "grad_norm": 3.196038174031813,
+      "learning_rate": 1.0153491984686593e-06,
+      "loss": 0.5447,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5107727016356458,
+      "grad_norm": 3.0228322304609083,
+      "learning_rate": 1.0138874689743047e-06,
+      "loss": 0.5442,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5112239142695996,
+      "grad_norm": 3.2815180852833095,
+      "learning_rate": 1.0124257098009042e-06,
+      "loss": 0.5194,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5116751269035533,
+      "grad_norm": 2.9460400536573315,
+      "learning_rate": 1.0109639240723973e-06,
+      "loss": 0.5793,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5121263395375071,
+      "grad_norm": 3.2088130319211015,
+      "learning_rate": 1.0095021149127806e-06,
+      "loss": 0.554,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5125775521714608,
+      "grad_norm": 3.3804661649102465,
+      "learning_rate": 1.0080402854461007e-06,
+      "loss": 0.5165,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5130287648054146,
+      "grad_norm": 3.321371296451192,
+      "learning_rate": 1.0065784387964485e-06,
+      "loss": 0.4321,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5134799774393684,
+      "grad_norm": 3.037664620181153,
+      "learning_rate": 1.0051165780879503e-06,
+      "loss": 0.5399,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5139311900733221,
+      "grad_norm": 3.556502110355346,
+      "learning_rate": 1.0036547064447622e-06,
+      "loss": 0.5263,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5143824027072758,
+      "grad_norm": 3.281961668919357,
+      "learning_rate": 1.0021928269910657e-06,
+      "loss": 0.5017,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5148336153412295,
+      "grad_norm": 4.422821086167055,
+      "learning_rate": 1.0007309428510568e-06,
+      "loss": 0.6125,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5152848279751833,
+      "grad_norm": 3.042185816714922,
+      "learning_rate": 9.992690571489431e-07,
+      "loss": 0.4608,
+      "step": 1142
+    },
+    {
+      "epoch": 0.515736040609137,
+      "grad_norm": 3.139974486773792,
+      "learning_rate": 9.978071730089344e-07,
+      "loss": 0.4999,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5161872532430908,
+      "grad_norm": 3.050392508261623,
+      "learning_rate": 9.963452935552377e-07,
+      "loss": 0.5472,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5166384658770445,
+      "grad_norm": 3.1524033877265336,
+      "learning_rate": 9.948834219120498e-07,
+      "loss": 0.5617,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5170896785109983,
+      "grad_norm": 3.045308188934204,
+      "learning_rate": 9.934215612035514e-07,
+      "loss": 0.4082,
+      "step": 1146
+    },
+    {
+      "epoch": 0.517540891144952,
+      "grad_norm": 3.3228413802611225,
+      "learning_rate": 9.91959714553899e-07,
+      "loss": 0.5028,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5179921037789058,
+      "grad_norm": 2.9738895566428285,
+      "learning_rate": 9.904978850872191e-07,
+      "loss": 0.6102,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5184433164128596,
+      "grad_norm": 3.164313244153204,
+      "learning_rate": 9.89036075927603e-07,
+      "loss": 0.541,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5188945290468133,
+      "grad_norm": 3.0993846216386998,
+      "learning_rate": 9.87574290199096e-07,
+      "loss": 0.5887,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5193457416807671,
+      "grad_norm": 3.0457956258147503,
+      "learning_rate": 9.861125310256954e-07,
+      "loss": 0.518,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5197969543147208,
+      "grad_norm": 3.487761052187324,
+      "learning_rate": 9.846508015313406e-07,
+      "loss": 0.5634,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5202481669486746,
+      "grad_norm": 3.0958153646337654,
+      "learning_rate": 9.831891048399084e-07,
+      "loss": 0.486,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5206993795826284,
+      "grad_norm": 2.849378458901456,
+      "learning_rate": 9.817274440752052e-07,
+      "loss": 0.4714,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5211505922165821,
+      "grad_norm": 3.0204334797341867,
+      "learning_rate": 9.802658223609608e-07,
+      "loss": 0.5409,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5216018048505359,
+      "grad_norm": 3.4389615956259414,
+      "learning_rate": 9.78804242820821e-07,
+      "loss": 0.5677,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5220530174844895,
+      "grad_norm": 3.079047759993495,
+      "learning_rate": 9.773427085783413e-07,
+      "loss": 0.5837,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5225042301184433,
+      "grad_norm": 3.2180585801181407,
+      "learning_rate": 9.758812227569812e-07,
+      "loss": 0.6249,
+      "step": 1158
+    },
+    {
+      "epoch": 0.522955442752397,
+      "grad_norm": 3.106738311935934,
+      "learning_rate": 9.744197884800967e-07,
+      "loss": 0.5928,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5234066553863508,
+      "grad_norm": 3.0296372900818676,
+      "learning_rate": 9.72958408870933e-07,
+      "loss": 0.6287,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5238578680203045,
+      "grad_norm": 3.2657856406977164,
+      "learning_rate": 9.714970870526186e-07,
+      "loss": 0.4556,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5243090806542583,
+      "grad_norm": 2.8714718085250497,
+      "learning_rate": 9.700358261481592e-07,
+      "loss": 0.5133,
+      "step": 1162
+    },
+    {
+      "epoch": 0.524760293288212,
+      "grad_norm": 3.1929538694444353,
+      "learning_rate": 9.6857462928043e-07,
+      "loss": 0.5365,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5252115059221658,
+      "grad_norm": 3.1374361855603725,
+      "learning_rate": 9.671134995721684e-07,
+      "loss": 0.5009,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5256627185561196,
+      "grad_norm": 3.0612765295251614,
+      "learning_rate": 9.656524401459692e-07,
+      "loss": 0.4686,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5261139311900733,
+      "grad_norm": 3.0360262730106653,
+      "learning_rate": 9.64191454124277e-07,
+      "loss": 0.5078,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5265651438240271,
+      "grad_norm": 2.7855335523075895,
+      "learning_rate": 9.62730544629379e-07,
+      "loss": 0.4643,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5270163564579808,
+      "grad_norm": 3.2784285165726326,
+      "learning_rate": 9.612697147834003e-07,
+      "loss": 0.5212,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5274675690919346,
+      "grad_norm": 2.9731466081932125,
+      "learning_rate": 9.598089677082933e-07,
+      "loss": 0.5892,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5279187817258884,
+      "grad_norm": 3.496014297140297,
+      "learning_rate": 9.583483065258363e-07,
+      "loss": 0.5361,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5283699943598421,
+      "grad_norm": 3.2554831703742027,
+      "learning_rate": 9.568877343576212e-07,
+      "loss": 0.5437,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5288212069937959,
+      "grad_norm": 3.0774362338181307,
+      "learning_rate": 9.554272543250516e-07,
+      "loss": 0.5781,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5292724196277496,
+      "grad_norm": 3.189104546167665,
+      "learning_rate": 9.539668695493344e-07,
+      "loss": 0.4621,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5297236322617034,
+      "grad_norm": 3.5804034846331607,
+      "learning_rate": 9.525065831514713e-07,
+      "loss": 0.5667,
+      "step": 1174
+    },
+    {
+      "epoch": 0.530174844895657,
+      "grad_norm": 3.0139310674345934,
+      "learning_rate": 9.510463982522551e-07,
+      "loss": 0.6764,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5306260575296108,
+      "grad_norm": 3.6437122981783507,
+      "learning_rate": 9.495863179722615e-07,
+      "loss": 0.5191,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5310772701635645,
+      "grad_norm": 3.780802968968682,
+      "learning_rate": 9.481263454318421e-07,
+      "loss": 0.5795,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5315284827975183,
+      "grad_norm": 3.371236920516789,
+      "learning_rate": 9.466664837511186e-07,
+      "loss": 0.4982,
+      "step": 1178
+    },
+    {
+      "epoch": 0.531979695431472,
+      "grad_norm": 3.174724613560376,
+      "learning_rate": 9.452067360499753e-07,
+      "loss": 0.5753,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5324309080654258,
+      "grad_norm": 2.9536643600933283,
+      "learning_rate": 9.437471054480539e-07,
+      "loss": 0.6468,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5328821206993796,
+      "grad_norm": 3.358069442011158,
+      "learning_rate": 9.422875950647451e-07,
+      "loss": 0.5811,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 3.1595218350137357,
+      "learning_rate": 9.408282080191828e-07,
+      "loss": 0.5539,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5337845459672871,
+      "grad_norm": 3.1013329069241844,
+      "learning_rate": 9.393689474302376e-07,
+      "loss": 0.4659,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5342357586012408,
+      "grad_norm": 3.743900056060665,
+      "learning_rate": 9.379098164165092e-07,
+      "loss": 0.6178,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5346869712351946,
+      "grad_norm": 3.10516339656473,
+      "learning_rate": 9.364508180963207e-07,
+      "loss": 0.5486,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5351381838691484,
+      "grad_norm": 3.7247912121292965,
+      "learning_rate": 9.349919555877124e-07,
+      "loss": 0.6168,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5355893965031021,
+      "grad_norm": 3.523392877840991,
+      "learning_rate": 9.335332320084331e-07,
+      "loss": 0.6578,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5360406091370559,
+      "grad_norm": 3.029040493416117,
+      "learning_rate": 9.320746504759354e-07,
+      "loss": 0.6227,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5364918217710096,
+      "grad_norm": 3.494834853466634,
+      "learning_rate": 9.306162141073685e-07,
+      "loss": 0.6807,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5369430344049634,
+      "grad_norm": 3.149740148496557,
+      "learning_rate": 9.291579260195708e-07,
+      "loss": 0.5903,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5373942470389171,
+      "grad_norm": 3.267822705572205,
+      "learning_rate": 9.27699789329064e-07,
+      "loss": 0.4847,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5378454596728709,
+      "grad_norm": 3.714908029228556,
+      "learning_rate": 9.262418071520463e-07,
+      "loss": 0.6019,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5382966723068245,
+      "grad_norm": 2.824698681778885,
+      "learning_rate": 9.247839826043857e-07,
+      "loss": 0.4482,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5387478849407783,
+      "grad_norm": 3.151682454847607,
+      "learning_rate": 9.233263188016138e-07,
+      "loss": 0.4968,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5391990975747321,
+      "grad_norm": 2.9790615764183355,
+      "learning_rate": 9.218688188589174e-07,
+      "loss": 0.5278,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5396503102086858,
+      "grad_norm": 2.9627521296692474,
+      "learning_rate": 9.204114858911346e-07,
+      "loss": 0.5431,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5401015228426396,
+      "grad_norm": 2.992708202976521,
+      "learning_rate": 9.189543230127461e-07,
+      "loss": 0.5062,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5405527354765933,
+      "grad_norm": 2.8085674068135016,
+      "learning_rate": 9.174973333378683e-07,
+      "loss": 0.5778,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5410039481105471,
+      "grad_norm": 3.324978073123737,
+      "learning_rate": 9.160405199802486e-07,
+      "loss": 0.5476,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5414551607445008,
+      "grad_norm": 2.6559214068968937,
+      "learning_rate": 9.145838860532567e-07,
+      "loss": 0.5372,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5419063733784546,
+      "grad_norm": 2.9815500519398763,
+      "learning_rate": 9.131274346698795e-07,
+      "loss": 0.5114,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5423575860124084,
+      "grad_norm": 3.4127009993684196,
+      "learning_rate": 9.116711689427136e-07,
+      "loss": 0.6166,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5428087986463621,
+      "grad_norm": 3.226502001346105,
+      "learning_rate": 9.102150919839589e-07,
+      "loss": 0.5777,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5432600112803159,
+      "grad_norm": 3.2761488720411585,
+      "learning_rate": 9.087592069054118e-07,
+      "loss": 0.4045,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5437112239142696,
+      "grad_norm": 3.2522550379255803,
+      "learning_rate": 9.073035168184579e-07,
+      "loss": 0.5614,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5441624365482234,
+      "grad_norm": 3.158338840928273,
+      "learning_rate": 9.058480248340671e-07,
+      "loss": 0.5347,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5446136491821771,
+      "grad_norm": 3.128846961573522,
+      "learning_rate": 9.043927340627857e-07,
+      "loss": 0.4551,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5450648618161309,
+      "grad_norm": 3.2010639237879333,
+      "learning_rate": 9.029376476147301e-07,
+      "loss": 0.6002,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5455160744500847,
+      "grad_norm": 3.0281082746251857,
+      "learning_rate": 9.014827685995794e-07,
+      "loss": 0.7246,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5459672870840383,
+      "grad_norm": 3.011888432511697,
+      "learning_rate": 9.0002810012657e-07,
+      "loss": 0.5268,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5464184997179921,
+      "grad_norm": 3.2243121571383386,
+      "learning_rate": 8.985736453044886e-07,
+      "loss": 0.5456,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5468697123519458,
+      "grad_norm": 3.0825653650623126,
+      "learning_rate": 8.97119407241664e-07,
+      "loss": 0.5749,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5473209249858996,
+      "grad_norm": 2.661130517981536,
+      "learning_rate": 8.956653890459632e-07,
+      "loss": 0.5596,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5477721376198533,
+      "grad_norm": 3.724775101530356,
+      "learning_rate": 8.942115938247823e-07,
+      "loss": 0.5477,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5482233502538071,
+      "grad_norm": 3.6650788579826585,
+      "learning_rate": 8.927580246850416e-07,
+      "loss": 0.4938,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5486745628877608,
+      "grad_norm": 3.139755611427215,
+      "learning_rate": 8.913046847331784e-07,
+      "loss": 0.6102,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5491257755217146,
+      "grad_norm": 2.9266705632636856,
+      "learning_rate": 8.89851577075139e-07,
+      "loss": 0.5387,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5495769881556684,
+      "grad_norm": 3.033908297559131,
+      "learning_rate": 8.883987048163746e-07,
+      "loss": 0.4825,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5500282007896221,
+      "grad_norm": 2.961499235356465,
+      "learning_rate": 8.869460710618323e-07,
+      "loss": 0.5537,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5504794134235759,
+      "grad_norm": 2.58750526796041,
+      "learning_rate": 8.8549367891595e-07,
+      "loss": 0.414,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5509306260575296,
+      "grad_norm": 3.2580895273085684,
+      "learning_rate": 8.840415314826496e-07,
+      "loss": 0.5591,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5513818386914834,
+      "grad_norm": 3.416770692667306,
+      "learning_rate": 8.825896318653292e-07,
+      "loss": 0.5914,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5518330513254371,
+      "grad_norm": 3.3868054346511385,
+      "learning_rate": 8.811379831668578e-07,
+      "loss": 0.5391,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5522842639593909,
+      "grad_norm": 3.513167625744527,
+      "learning_rate": 8.796865884895685e-07,
+      "loss": 0.6352,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5527354765933447,
+      "grad_norm": 3.2759348364567495,
+      "learning_rate": 8.782354509352505e-07,
+      "loss": 0.4788,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5531866892272984,
+      "grad_norm": 3.16036405291083,
+      "learning_rate": 8.767845736051445e-07,
+      "loss": 0.5454,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5536379018612522,
+      "grad_norm": 3.0291347992929563,
+      "learning_rate": 8.753339595999343e-07,
+      "loss": 0.4814,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5540891144952058,
+      "grad_norm": 3.419410089774215,
+      "learning_rate": 8.738836120197414e-07,
+      "loss": 0.6572,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5545403271291596,
+      "grad_norm": 3.4196948382062895,
+      "learning_rate": 8.724335339641183e-07,
+      "loss": 0.5833,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5549915397631133,
+      "grad_norm": 2.95345812029554,
+      "learning_rate": 8.709837285320405e-07,
+      "loss": 0.5173,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5554427523970671,
+      "grad_norm": 3.277178241020724,
+      "learning_rate": 8.695341988219013e-07,
+      "loss": 0.4623,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5558939650310208,
+      "grad_norm": 3.4090915194634466,
+      "learning_rate": 8.68084947931506e-07,
+      "loss": 0.54,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5563451776649746,
+      "grad_norm": 2.9424477792511072,
+      "learning_rate": 8.666359789580612e-07,
+      "loss": 0.4951,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5567963902989284,
+      "grad_norm": 2.8093499614133166,
+      "learning_rate": 8.651872949981741e-07,
+      "loss": 0.5294,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5572476029328821,
+      "grad_norm": 3.403220558325019,
+      "learning_rate": 8.637388991478404e-07,
+      "loss": 0.5359,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5576988155668359,
+      "grad_norm": 2.7915279885504294,
+      "learning_rate": 8.622907945024417e-07,
+      "loss": 0.4772,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5581500282007896,
+      "grad_norm": 3.402216398468153,
+      "learning_rate": 8.608429841567364e-07,
+      "loss": 0.6865,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5586012408347434,
+      "grad_norm": 3.4495430024602993,
+      "learning_rate": 8.593954712048544e-07,
+      "loss": 0.5511,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5590524534686971,
+      "grad_norm": 3.141295268567425,
+      "learning_rate": 8.579482587402899e-07,
+      "loss": 0.6404,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5595036661026509,
+      "grad_norm": 3.5009079206917257,
+      "learning_rate": 8.565013498558941e-07,
+      "loss": 0.5031,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5599548787366047,
+      "grad_norm": 3.4205180711589063,
+      "learning_rate": 8.550547476438708e-07,
+      "loss": 0.7429,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5604060913705584,
+      "grad_norm": 2.773602345160653,
+      "learning_rate": 8.536084551957676e-07,
+      "loss": 0.5418,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5608573040045122,
+      "grad_norm": 3.1366076676001704,
+      "learning_rate": 8.521624756024704e-07,
+      "loss": 0.554,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5613085166384659,
+      "grad_norm": 3.1430067736085676,
+      "learning_rate": 8.507168119541963e-07,
+      "loss": 0.5359,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5617597292724196,
+      "grad_norm": 3.3366605489268717,
+      "learning_rate": 8.492714673404871e-07,
+      "loss": 0.6201,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5622109419063733,
+      "grad_norm": 2.9317529934404396,
+      "learning_rate": 8.478264448502036e-07,
+      "loss": 0.5387,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5626621545403271,
+      "grad_norm": 2.8606734322020233,
+      "learning_rate": 8.463817475715168e-07,
+      "loss": 0.5679,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5631133671742808,
+      "grad_norm": 2.921942814808396,
+      "learning_rate": 8.449373785919034e-07,
+      "loss": 0.6017,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5635645798082346,
+      "grad_norm": 3.108017504971404,
+      "learning_rate": 8.434933409981389e-07,
+      "loss": 0.5762,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5640157924421884,
+      "grad_norm": 3.2256170994042828,
+      "learning_rate": 8.420496378762899e-07,
+      "loss": 0.5252,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5644670050761421,
+      "grad_norm": 2.733940948644607,
+      "learning_rate": 8.406062723117089e-07,
+      "loss": 0.4403,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5649182177100959,
+      "grad_norm": 3.130827807611649,
+      "learning_rate": 8.391632473890262e-07,
+      "loss": 0.5186,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5653694303440496,
+      "grad_norm": 2.8551274180587187,
+      "learning_rate": 8.377205661921452e-07,
+      "loss": 0.5046,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5658206429780034,
+      "grad_norm": 3.0734293806965263,
+      "learning_rate": 8.36278231804233e-07,
+      "loss": 0.5113,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5662718556119571,
+      "grad_norm": 3.0293954304204593,
+      "learning_rate": 8.348362473077169e-07,
+      "loss": 0.4135,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5667230682459109,
+      "grad_norm": 3.2711429622012087,
+      "learning_rate": 8.333946157842767e-07,
+      "loss": 0.5911,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5671742808798647,
+      "grad_norm": 3.21348983820472,
+      "learning_rate": 8.319533403148366e-07,
+      "loss": 0.5845,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5676254935138184,
+      "grad_norm": 2.977221190242596,
+      "learning_rate": 8.305124239795608e-07,
+      "loss": 0.4975,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5680767061477722,
+      "grad_norm": 3.317575862791347,
+      "learning_rate": 8.29071869857846e-07,
+      "loss": 0.5145,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5685279187817259,
+      "grad_norm": 3.0078612380810217,
+      "learning_rate": 8.27631681028314e-07,
+      "loss": 0.4761,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5689791314156797,
+      "grad_norm": 2.9469364940545684,
+      "learning_rate": 8.26191860568807e-07,
+      "loss": 0.4903,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5694303440496334,
+      "grad_norm": 2.8641257568656804,
+      "learning_rate": 8.247524115563788e-07,
+      "loss": 0.4808,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5698815566835871,
+      "grad_norm": 3.3655191743332313,
+      "learning_rate": 8.233133370672903e-07,
+      "loss": 0.684,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5703327693175408,
+      "grad_norm": 3.2105020379563602,
+      "learning_rate": 8.218746401770021e-07,
+      "loss": 0.5674,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5707839819514946,
+      "grad_norm": 3.5642628952355975,
+      "learning_rate": 8.204363239601668e-07,
+      "loss": 0.501,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5712351945854484,
+      "grad_norm": 2.838113061969968,
+      "learning_rate": 8.189983914906248e-07,
+      "loss": 0.616,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5716864072194021,
+      "grad_norm": 2.9737366718198626,
+      "learning_rate": 8.175608458413947e-07,
+      "loss": 0.576,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5721376198533559,
+      "grad_norm": 2.7885730115959597,
+      "learning_rate": 8.161236900846702e-07,
+      "loss": 0.4278,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5725888324873096,
+      "grad_norm": 3.4834364046330792,
+      "learning_rate": 8.146869272918108e-07,
+      "loss": 0.5024,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5730400451212634,
+      "grad_norm": 3.0305631394109054,
+      "learning_rate": 8.132505605333361e-07,
+      "loss": 0.542,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5734912577552171,
+      "grad_norm": 2.7776093782076705,
+      "learning_rate": 8.118145928789197e-07,
+      "loss": 0.382,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5739424703891709,
+      "grad_norm": 2.982019078285388,
+      "learning_rate": 8.103790273973823e-07,
+      "loss": 0.5661,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5743936830231247,
+      "grad_norm": 2.9581883979360155,
+      "learning_rate": 8.089438671566852e-07,
+      "loss": 0.5193,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5748448956570784,
+      "grad_norm": 2.973062841681966,
+      "learning_rate": 8.075091152239229e-07,
+      "loss": 0.5088,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5752961082910322,
+      "grad_norm": 3.4183675101644595,
+      "learning_rate": 8.060747746653179e-07,
+      "loss": 0.6613,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5757473209249859,
+      "grad_norm": 3.5528615876910012,
+      "learning_rate": 8.046408485462136e-07,
+      "loss": 0.5776,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5761985335589397,
+      "grad_norm": 3.332215928387168,
+      "learning_rate": 8.032073399310677e-07,
+      "loss": 0.4554,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5766497461928934,
+      "grad_norm": 3.1567473951867178,
+      "learning_rate": 8.017742518834453e-07,
+      "loss": 0.5779,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5771009588268472,
+      "grad_norm": 3.1547565169857723,
+      "learning_rate": 8.003415874660129e-07,
+      "loss": 0.6327,
+      "step": 1279
+    },
+    {
+      "epoch": 0.577552171460801,
+      "grad_norm": 2.9713115658035547,
+      "learning_rate": 7.989093497405322e-07,
+      "loss": 0.5887,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5780033840947546,
+      "grad_norm": 3.2508596246355834,
+      "learning_rate": 7.974775417678517e-07,
+      "loss": 0.5491,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5784545967287084,
+      "grad_norm": 3.2239593887990288,
+      "learning_rate": 7.960461666079029e-07,
+      "loss": 0.5808,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5789058093626621,
+      "grad_norm": 3.3138402872343966,
+      "learning_rate": 7.946152273196911e-07,
+      "loss": 0.5569,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5793570219966159,
+      "grad_norm": 3.434837325852875,
+      "learning_rate": 7.931847269612911e-07,
+      "loss": 0.5351,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5798082346305696,
+      "grad_norm": 3.234624298819192,
+      "learning_rate": 7.917546685898391e-07,
+      "loss": 0.5941,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5802594472645234,
+      "grad_norm": 3.441772930399985,
+      "learning_rate": 7.903250552615272e-07,
+      "loss": 0.6435,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5807106598984771,
+      "grad_norm": 3.2690887674264895,
+      "learning_rate": 7.888958900315959e-07,
+      "loss": 0.5054,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5811618725324309,
+      "grad_norm": 3.9833861316481256,
+      "learning_rate": 7.874671759543278e-07,
+      "loss": 0.599,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5816130851663847,
+      "grad_norm": 2.868845567268956,
+      "learning_rate": 7.860389160830419e-07,
+      "loss": 0.4934,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5820642978003384,
+      "grad_norm": 3.047832823905869,
+      "learning_rate": 7.846111134700867e-07,
+      "loss": 0.5081,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5825155104342922,
+      "grad_norm": 3.275881282842554,
+      "learning_rate": 7.831837711668332e-07,
+      "loss": 0.5563,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5829667230682459,
+      "grad_norm": 3.14320311170746,
+      "learning_rate": 7.817568922236681e-07,
+      "loss": 0.5681,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5834179357021997,
+      "grad_norm": 3.524498053095352,
+      "learning_rate": 7.80330479689989e-07,
+      "loss": 0.6049,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5838691483361534,
+      "grad_norm": 3.118060779938019,
+      "learning_rate": 7.789045366141961e-07,
+      "loss": 0.6507,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5843203609701072,
+      "grad_norm": 3.3177816993811065,
+      "learning_rate": 7.774790660436857e-07,
+      "loss": 0.4724,
+      "step": 1295
+    },
+    {
+      "epoch": 0.584771573604061,
+      "grad_norm": 2.837955287125353,
+      "learning_rate": 7.760540710248454e-07,
+      "loss": 0.51,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5852227862380147,
+      "grad_norm": 3.3668111960232117,
+      "learning_rate": 7.746295546030458e-07,
+      "loss": 0.637,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5856739988719684,
+      "grad_norm": 3.103539045578858,
+      "learning_rate": 7.732055198226351e-07,
+      "loss": 0.4814,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5861252115059221,
+      "grad_norm": 3.0421653752914413,
+      "learning_rate": 7.717819697269321e-07,
+      "loss": 0.525,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5865764241398759,
+      "grad_norm": 3.644103387205874,
+      "learning_rate": 7.703589073582193e-07,
+      "loss": 0.6064,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5870276367738296,
+      "grad_norm": 2.8055962606604092,
+      "learning_rate": 7.689363357577378e-07,
+      "loss": 0.5143,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5874788494077834,
+      "grad_norm": 3.277739442754899,
+      "learning_rate": 7.675142579656788e-07,
+      "loss": 0.5523,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5879300620417371,
+      "grad_norm": 3.460236702752055,
+      "learning_rate": 7.660926770211787e-07,
+      "loss": 0.5705,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5883812746756909,
+      "grad_norm": 3.260180098933021,
+      "learning_rate": 7.646715959623125e-07,
+      "loss": 0.6103,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5888324873096447,
+      "grad_norm": 3.4686183563952824,
+      "learning_rate": 7.632510178260859e-07,
+      "loss": 0.6584,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5892836999435984,
+      "grad_norm": 3.2006762499940353,
+      "learning_rate": 7.618309456484308e-07,
+      "loss": 0.4899,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5897349125775522,
+      "grad_norm": 2.9971855862973267,
+      "learning_rate": 7.604113824641973e-07,
+      "loss": 0.6366,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5901861252115059,
+      "grad_norm": 3.1873994214172807,
+      "learning_rate": 7.589923313071479e-07,
+      "loss": 0.6843,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5906373378454597,
+      "grad_norm": 3.2537367572253824,
+      "learning_rate": 7.575737952099505e-07,
+      "loss": 0.5228,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5910885504794134,
+      "grad_norm": 3.079550130831521,
+      "learning_rate": 7.561557772041725e-07,
+      "loss": 0.4874,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5915397631133672,
+      "grad_norm": 3.2716458004854543,
+      "learning_rate": 7.547382803202742e-07,
+      "loss": 0.5857,
+      "step": 1311
+    },
+    {
+      "epoch": 0.591990975747321,
+      "grad_norm": 3.1743175723654735,
+      "learning_rate": 7.533213075876022e-07,
+      "loss": 0.529,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5924421883812747,
+      "grad_norm": 3.8188908521422458,
+      "learning_rate": 7.519048620343825e-07,
+      "loss": 0.6333,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5928934010152285,
+      "grad_norm": 2.8402686230355156,
+      "learning_rate": 7.504889466877149e-07,
+      "loss": 0.5294,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5933446136491822,
+      "grad_norm": 3.207555735354162,
+      "learning_rate": 7.490735645735666e-07,
+      "loss": 0.4968,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5937958262831359,
+      "grad_norm": 3.0695972720324622,
+      "learning_rate": 7.476587187167635e-07,
+      "loss": 0.5465,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5942470389170896,
+      "grad_norm": 2.9505661618967283,
+      "learning_rate": 7.462444121409875e-07,
+      "loss": 0.6166,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5946982515510434,
+      "grad_norm": 3.2187802075234737,
+      "learning_rate": 7.448306478687663e-07,
+      "loss": 0.4585,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5951494641849971,
+      "grad_norm": 3.1253329687327818,
+      "learning_rate": 7.434174289214696e-07,
+      "loss": 0.5872,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5956006768189509,
+      "grad_norm": 2.945507566127143,
+      "learning_rate": 7.420047583193018e-07,
+      "loss": 0.5749,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5960518894529047,
+      "grad_norm": 2.8046185112075506,
+      "learning_rate": 7.405926390812952e-07,
+      "loss": 0.6378,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5965031020868584,
+      "grad_norm": 3.3367837174035246,
+      "learning_rate": 7.391810742253035e-07,
+      "loss": 0.5555,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5969543147208122,
+      "grad_norm": 3.361270385049288,
+      "learning_rate": 7.377700667679952e-07,
+      "loss": 0.4942,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5974055273547659,
+      "grad_norm": 3.198915242112322,
+      "learning_rate": 7.363596197248488e-07,
+      "loss": 0.4949,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5978567399887197,
+      "grad_norm": 3.329606223960123,
+      "learning_rate": 7.349497361101442e-07,
+      "loss": 0.5877,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5983079526226734,
+      "grad_norm": 3.519739136879892,
+      "learning_rate": 7.335404189369578e-07,
+      "loss": 0.645,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5987591652566272,
+      "grad_norm": 3.0247206849836976,
+      "learning_rate": 7.321316712171551e-07,
+      "loss": 0.4796,
+      "step": 1327
+    },
+    {
+      "epoch": 0.599210377890581,
+      "grad_norm": 3.186177089807284,
+      "learning_rate": 7.307234959613842e-07,
+      "loss": 0.5669,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5996615905245347,
+      "grad_norm": 3.130095831838903,
+      "learning_rate": 7.293158961790714e-07,
+      "loss": 0.5828,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6001128031584885,
+      "grad_norm": 3.384495690869306,
+      "learning_rate": 7.279088748784105e-07,
+      "loss": 0.6257,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6005640157924422,
+      "grad_norm": 3.3442845371887135,
+      "learning_rate": 7.265024350663615e-07,
+      "loss": 0.5929,
+      "step": 1331
+    },
+    {
+      "epoch": 0.601015228426396,
+      "grad_norm": 3.329300257806507,
+      "learning_rate": 7.250965797486404e-07,
+      "loss": 0.5622,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6014664410603497,
+      "grad_norm": 3.137210816893963,
+      "learning_rate": 7.236913119297144e-07,
+      "loss": 0.4791,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6019176536943034,
+      "grad_norm": 3.2475644991308297,
+      "learning_rate": 7.222866346127952e-07,
+      "loss": 0.5562,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6023688663282571,
+      "grad_norm": 2.964087669113336,
+      "learning_rate": 7.208825507998325e-07,
+      "loss": 0.5684,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6028200789622109,
+      "grad_norm": 3.41721759180177,
+      "learning_rate": 7.194790634915075e-07,
+      "loss": 0.5463,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6032712915961647,
+      "grad_norm": 2.8453268851912923,
+      "learning_rate": 7.180761756872267e-07,
+      "loss": 0.5151,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6037225042301184,
+      "grad_norm": 3.2646935195604483,
+      "learning_rate": 7.166738903851153e-07,
+      "loss": 0.4982,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6041737168640722,
+      "grad_norm": 2.970220890446487,
+      "learning_rate": 7.152722105820112e-07,
+      "loss": 0.5154,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6046249294980259,
+      "grad_norm": 3.4187797390981114,
+      "learning_rate": 7.138711392734578e-07,
+      "loss": 0.4689,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6050761421319797,
+      "grad_norm": 2.985186967050454,
+      "learning_rate": 7.124706794536983e-07,
+      "loss": 0.5244,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6055273547659334,
+      "grad_norm": 3.455866470082041,
+      "learning_rate": 7.110708341156698e-07,
+      "loss": 0.5728,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6059785673998872,
+      "grad_norm": 3.6106718461697977,
+      "learning_rate": 7.096716062509947e-07,
+      "loss": 0.5279,
+      "step": 1343
+    },
+    {
+      "epoch": 0.606429780033841,
+      "grad_norm": 4.105211530606187,
+      "learning_rate": 7.082729988499768e-07,
+      "loss": 0.6094,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6068809926677947,
+      "grad_norm": 2.693788412848332,
+      "learning_rate": 7.068750149015936e-07,
+      "loss": 0.4555,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6073322053017485,
+      "grad_norm": 3.0065637977545676,
+      "learning_rate": 7.054776573934905e-07,
+      "loss": 0.565,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6077834179357022,
+      "grad_norm": 3.289187175236741,
+      "learning_rate": 7.04080929311974e-07,
+      "loss": 0.5429,
+      "step": 1347
+    },
+    {
+      "epoch": 0.608234630569656,
+      "grad_norm": 3.7046080997067903,
+      "learning_rate": 7.026848336420052e-07,
+      "loss": 0.5946,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6086858432036097,
+      "grad_norm": 2.9252591310820586,
+      "learning_rate": 7.012893733671943e-07,
+      "loss": 0.4993,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6091370558375635,
+      "grad_norm": 2.8882168795713894,
+      "learning_rate": 6.998945514697923e-07,
+      "loss": 0.4905,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6095882684715171,
+      "grad_norm": 3.583128007061669,
+      "learning_rate": 6.985003709306871e-07,
+      "loss": 0.4726,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6100394811054709,
+      "grad_norm": 3.130137883692283,
+      "learning_rate": 6.97106834729396e-07,
+      "loss": 0.4106,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6104906937394247,
+      "grad_norm": 3.0703439274804682,
+      "learning_rate": 6.957139458440584e-07,
+      "loss": 0.5424,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6109419063733784,
+      "grad_norm": 3.0314143588988496,
+      "learning_rate": 6.943217072514311e-07,
+      "loss": 0.6235,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6113931190073322,
+      "grad_norm": 3.495368048533246,
+      "learning_rate": 6.929301219268805e-07,
+      "loss": 0.664,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6118443316412859,
+      "grad_norm": 3.2783156409995318,
+      "learning_rate": 6.915391928443779e-07,
+      "loss": 0.5994,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6122955442752397,
+      "grad_norm": 3.1900987577022075,
+      "learning_rate": 6.90148922976491e-07,
+      "loss": 0.473,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6127467569091934,
+      "grad_norm": 3.542481502196449,
+      "learning_rate": 6.887593152943789e-07,
+      "loss": 0.6318,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6131979695431472,
+      "grad_norm": 3.0512134064626077,
+      "learning_rate": 6.873703727677862e-07,
+      "loss": 0.4577,
+      "step": 1359
+    },
+    {
+      "epoch": 0.613649182177101,
+      "grad_norm": 3.0267553714028814,
+      "learning_rate": 6.859820983650355e-07,
+      "loss": 0.5358,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6141003948110547,
+      "grad_norm": 3.320872889591731,
+      "learning_rate": 6.845944950530218e-07,
+      "loss": 0.5442,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6145516074450085,
+      "grad_norm": 3.2701182397565347,
+      "learning_rate": 6.832075657972054e-07,
+      "loss": 0.5837,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6150028200789622,
+      "grad_norm": 3.3678828828961533,
+      "learning_rate": 6.818213135616071e-07,
+      "loss": 0.663,
+      "step": 1363
+    },
+    {
+      "epoch": 0.615454032712916,
+      "grad_norm": 3.1223684191824854,
+      "learning_rate": 6.804357413087992e-07,
+      "loss": 0.529,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6159052453468697,
+      "grad_norm": 3.686360531224555,
+      "learning_rate": 6.790508519999023e-07,
+      "loss": 0.6493,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6163564579808235,
+      "grad_norm": 2.9271029540244005,
+      "learning_rate": 6.776666485945769e-07,
+      "loss": 0.5776,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6168076706147773,
+      "grad_norm": 3.4729445853273737,
+      "learning_rate": 6.762831340510174e-07,
+      "loss": 0.512,
+      "step": 1367
+    },
+    {
+      "epoch": 0.617258883248731,
+      "grad_norm": 3.14390720715517,
+      "learning_rate": 6.749003113259466e-07,
+      "loss": 0.5518,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6177100958826847,
+      "grad_norm": 3.138843726117474,
+      "learning_rate": 6.735181833746087e-07,
+      "loss": 0.6031,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6181613085166384,
+      "grad_norm": 3.272624979852151,
+      "learning_rate": 6.721367531507626e-07,
+      "loss": 0.5369,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6186125211505922,
+      "grad_norm": 3.562938902127868,
+      "learning_rate": 6.70756023606676e-07,
+      "loss": 0.6634,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6190637337845459,
+      "grad_norm": 2.937710815940084,
+      "learning_rate": 6.6937599769312e-07,
+      "loss": 0.6368,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6195149464184997,
+      "grad_norm": 3.2281736301884596,
+      "learning_rate": 6.679966783593615e-07,
+      "loss": 0.5482,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6199661590524534,
+      "grad_norm": 3.46989691765386,
+      "learning_rate": 6.666180685531575e-07,
+      "loss": 0.5121,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6204173716864072,
+      "grad_norm": 2.9512285549656525,
+      "learning_rate": 6.65240171220748e-07,
+      "loss": 0.6428,
+      "step": 1375
+    },
+    {
+      "epoch": 0.620868584320361,
+      "grad_norm": 3.5246959007009044,
+      "learning_rate": 6.638629893068515e-07,
+      "loss": 0.5731,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6213197969543147,
+      "grad_norm": 3.0891012405500304,
+      "learning_rate": 6.62486525754657e-07,
+      "loss": 0.4931,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6217710095882685,
+      "grad_norm": 2.9732550708765233,
+      "learning_rate": 6.611107835058174e-07,
+      "loss": 0.4172,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 2.933874068194838,
+      "learning_rate": 6.59735765500446e-07,
+      "loss": 0.5248,
+      "step": 1379
+    },
+    {
+      "epoch": 0.622673434856176,
+      "grad_norm": 3.1023005159339165,
+      "learning_rate": 6.583614746771064e-07,
+      "loss": 0.5489,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6231246474901297,
+      "grad_norm": 2.9712725239339477,
+      "learning_rate": 6.569879139728096e-07,
+      "loss": 0.5301,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6235758601240835,
+      "grad_norm": 3.1648033323683045,
+      "learning_rate": 6.556150863230054e-07,
+      "loss": 0.5682,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6240270727580373,
+      "grad_norm": 2.7399149382162564,
+      "learning_rate": 6.542429946615773e-07,
+      "loss": 0.4796,
+      "step": 1383
+    },
+    {
+      "epoch": 0.624478285391991,
+      "grad_norm": 2.8260892433984073,
+      "learning_rate": 6.528716419208361e-07,
+      "loss": 0.5453,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6249294980259448,
+      "grad_norm": 3.145384660363434,
+      "learning_rate": 6.515010310315125e-07,
+      "loss": 0.638,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6253807106598985,
+      "grad_norm": 3.0447519086357846,
+      "learning_rate": 6.50131164922753e-07,
+      "loss": 0.5322,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6258319232938522,
+      "grad_norm": 3.4448461436407594,
+      "learning_rate": 6.487620465221117e-07,
+      "loss": 0.5624,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6262831359278059,
+      "grad_norm": 3.0855788626135023,
+      "learning_rate": 6.473936787555452e-07,
+      "loss": 0.4002,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6267343485617597,
+      "grad_norm": 2.943659166164125,
+      "learning_rate": 6.460260645474052e-07,
+      "loss": 0.4525,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6271855611957134,
+      "grad_norm": 2.9945684467434135,
+      "learning_rate": 6.44659206820434e-07,
+      "loss": 0.4683,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6276367738296672,
+      "grad_norm": 2.929306663688666,
+      "learning_rate": 6.432931084957566e-07,
+      "loss": 0.5541,
+      "step": 1391
+    },
+    {
+      "epoch": 0.628087986463621,
+      "grad_norm": 3.0908644677993817,
+      "learning_rate": 6.419277724928747e-07,
+      "loss": 0.5163,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6285391990975747,
+      "grad_norm": 3.001491521797685,
+      "learning_rate": 6.405632017296614e-07,
+      "loss": 0.6203,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6289904117315285,
+      "grad_norm": 3.1474986627885895,
+      "learning_rate": 6.391993991223543e-07,
+      "loss": 0.544,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6294416243654822,
+      "grad_norm": 3.2290177386909966,
+      "learning_rate": 6.378363675855494e-07,
+      "loss": 0.5752,
+      "step": 1395
+    },
+    {
+      "epoch": 0.629892836999436,
+      "grad_norm": 3.222879746589294,
+      "learning_rate": 6.364741100321947e-07,
+      "loss": 0.4536,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6303440496333897,
+      "grad_norm": 3.0349129642474755,
+      "learning_rate": 6.351126293735842e-07,
+      "loss": 0.4456,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6307952622673435,
+      "grad_norm": 3.218838258944328,
+      "learning_rate": 6.33751928519352e-07,
+      "loss": 0.5432,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6312464749012973,
+      "grad_norm": 3.2926190596476284,
+      "learning_rate": 6.323920103774644e-07,
+      "loss": 0.5973,
+      "step": 1399
+    },
+    {
+      "epoch": 0.631697687535251,
+      "grad_norm": 3.3559310121476775,
+      "learning_rate": 6.310328778542162e-07,
+      "loss": 0.5248,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6321489001692048,
+      "grad_norm": 3.6153857414340793,
+      "learning_rate": 6.296745338542229e-07,
+      "loss": 0.6443,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6326001128031585,
+      "grad_norm": 3.1304394163996623,
+      "learning_rate": 6.283169812804146e-07,
+      "loss": 0.5722,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6330513254371123,
+      "grad_norm": 3.0269630181138396,
+      "learning_rate": 6.269602230340304e-07,
+      "loss": 0.5002,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6335025380710659,
+      "grad_norm": 2.901434555273011,
+      "learning_rate": 6.256042620146118e-07,
+      "loss": 0.4313,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6339537507050197,
+      "grad_norm": 2.904472954933817,
+      "learning_rate": 6.242491011199963e-07,
+      "loss": 0.5316,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6344049633389734,
+      "grad_norm": 3.569318854020443,
+      "learning_rate": 6.228947432463111e-07,
+      "loss": 0.6989,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6348561759729272,
+      "grad_norm": 3.70336661181829,
+      "learning_rate": 6.21541191287968e-07,
+      "loss": 0.4956,
+      "step": 1407
+    },
+    {
+      "epoch": 0.635307388606881,
+      "grad_norm": 2.835895432940531,
+      "learning_rate": 6.201884481376562e-07,
+      "loss": 0.5343,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6357586012408347,
+      "grad_norm": 3.154332164144894,
+      "learning_rate": 6.188365166863365e-07,
+      "loss": 0.4946,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6362098138747885,
+      "grad_norm": 2.999966045677623,
+      "learning_rate": 6.174853998232346e-07,
+      "loss": 0.4792,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6366610265087422,
+      "grad_norm": 2.823054257582786,
+      "learning_rate": 6.161351004358359e-07,
+      "loss": 0.4774,
+      "step": 1411
+    },
+    {
+      "epoch": 0.637112239142696,
+      "grad_norm": 3.1087901532322473,
+      "learning_rate": 6.14785621409878e-07,
+      "loss": 0.5926,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6375634517766497,
+      "grad_norm": 3.19373907556398,
+      "learning_rate": 6.13436965629346e-07,
+      "loss": 0.4821,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6380146644106035,
+      "grad_norm": 3.7849347226642136,
+      "learning_rate": 6.120891359764655e-07,
+      "loss": 0.5625,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6384658770445573,
+      "grad_norm": 3.214785840415643,
+      "learning_rate": 6.107421353316964e-07,
+      "loss": 0.5056,
+      "step": 1415
+    },
+    {
+      "epoch": 0.638917089678511,
+      "grad_norm": 3.2716930771356387,
+      "learning_rate": 6.093959665737267e-07,
+      "loss": 0.5974,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6393683023124648,
+      "grad_norm": 2.657202103063792,
+      "learning_rate": 6.080506325794674e-07,
+      "loss": 0.458,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6398195149464185,
+      "grad_norm": 3.4613963613287284,
+      "learning_rate": 6.067061362240448e-07,
+      "loss": 0.5536,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6402707275803723,
+      "grad_norm": 3.1871447082433213,
+      "learning_rate": 6.053624803807951e-07,
+      "loss": 0.5803,
+      "step": 1419
+    },
+    {
+      "epoch": 0.640721940214326,
+      "grad_norm": 3.1180639042969274,
+      "learning_rate": 6.040196679212582e-07,
+      "loss": 0.4619,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6411731528482798,
+      "grad_norm": 3.179821687875411,
+      "learning_rate": 6.026777017151718e-07,
+      "loss": 0.584,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6416243654822334,
+      "grad_norm": 3.917598055933022,
+      "learning_rate": 6.013365846304657e-07,
+      "loss": 0.4427,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6420755781161872,
+      "grad_norm": 3.259154452067742,
+      "learning_rate": 5.999963195332536e-07,
+      "loss": 0.5559,
+      "step": 1423
+    },
+    {
+      "epoch": 0.642526790750141,
+      "grad_norm": 3.4742010090894806,
+      "learning_rate": 5.986569092878296e-07,
+      "loss": 0.5994,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6429780033840947,
+      "grad_norm": 3.753798551895116,
+      "learning_rate": 5.973183567566604e-07,
+      "loss": 0.6971,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6434292160180485,
+      "grad_norm": 2.941155690931197,
+      "learning_rate": 5.959806648003796e-07,
+      "loss": 0.516,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6438804286520022,
+      "grad_norm": 2.9533012729371326,
+      "learning_rate": 5.946438362777819e-07,
+      "loss": 0.6212,
+      "step": 1427
+    },
+    {
+      "epoch": 0.644331641285956,
+      "grad_norm": 3.3329411483235964,
+      "learning_rate": 5.933078740458166e-07,
+      "loss": 0.6503,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6447828539199097,
+      "grad_norm": 3.498152521685046,
+      "learning_rate": 5.919727809595815e-07,
+      "loss": 0.5699,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6452340665538635,
+      "grad_norm": 3.638822037194711,
+      "learning_rate": 5.906385598723178e-07,
+      "loss": 0.5854,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6456852791878173,
+      "grad_norm": 3.381793813968072,
+      "learning_rate": 5.893052136354017e-07,
+      "loss": 0.4964,
+      "step": 1431
+    },
+    {
+      "epoch": 0.646136491821771,
+      "grad_norm": 3.3009296858703268,
+      "learning_rate": 5.879727450983412e-07,
+      "loss": 0.593,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6465877044557248,
+      "grad_norm": 3.550794325305599,
+      "learning_rate": 5.866411571087671e-07,
+      "loss": 0.5447,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6470389170896785,
+      "grad_norm": 3.3261654579811992,
+      "learning_rate": 5.853104525124297e-07,
+      "loss": 0.5494,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6474901297236323,
+      "grad_norm": 3.428245584517914,
+      "learning_rate": 5.839806341531908e-07,
+      "loss": 0.6307,
+      "step": 1435
+    },
+    {
+      "epoch": 0.647941342357586,
+      "grad_norm": 3.0629393848960365,
+      "learning_rate": 5.82651704873018e-07,
+      "loss": 0.5884,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6483925549915398,
+      "grad_norm": 2.948994135334962,
+      "learning_rate": 5.813236675119793e-07,
+      "loss": 0.4318,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6488437676254936,
+      "grad_norm": 3.760438297720544,
+      "learning_rate": 5.79996524908236e-07,
+      "loss": 0.5324,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6492949802594473,
+      "grad_norm": 3.49087174910647,
+      "learning_rate": 5.786702798980388e-07,
+      "loss": 0.5048,
+      "step": 1439
+    },
+    {
+      "epoch": 0.649746192893401,
+      "grad_norm": 2.9974874928675614,
+      "learning_rate": 5.773449353157171e-07,
+      "loss": 0.5413,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6501974055273547,
+      "grad_norm": 3.777448022885416,
+      "learning_rate": 5.76020493993679e-07,
+      "loss": 0.5136,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6506486181613085,
+      "grad_norm": 3.210445403802698,
+      "learning_rate": 5.74696958762401e-07,
+      "loss": 0.5167,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6510998307952622,
+      "grad_norm": 3.336796933709652,
+      "learning_rate": 5.733743324504224e-07,
+      "loss": 0.5119,
+      "step": 1443
+    },
+    {
+      "epoch": 0.651551043429216,
+      "grad_norm": 3.3282911026676176,
+      "learning_rate": 5.720526178843418e-07,
+      "loss": 0.5884,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6520022560631697,
+      "grad_norm": 3.21378303870658,
+      "learning_rate": 5.707318178888082e-07,
+      "loss": 0.538,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6524534686971235,
+      "grad_norm": 2.969064914210708,
+      "learning_rate": 5.694119352865159e-07,
+      "loss": 0.5845,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6529046813310773,
+      "grad_norm": 3.30695390232686,
+      "learning_rate": 5.680929728981991e-07,
+      "loss": 0.5142,
+      "step": 1447
+    },
+    {
+      "epoch": 0.653355893965031,
+      "grad_norm": 3.0732966058972604,
+      "learning_rate": 5.667749335426246e-07,
+      "loss": 0.4201,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6538071065989848,
+      "grad_norm": 3.254645914821211,
+      "learning_rate": 5.654578200365885e-07,
+      "loss": 0.6627,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6542583192329385,
+      "grad_norm": 2.9760909272929683,
+      "learning_rate": 5.641416351949062e-07,
+      "loss": 0.4217,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6547095318668923,
+      "grad_norm": 2.997343851542973,
+      "learning_rate": 5.628263818304091e-07,
+      "loss": 0.5031,
+      "step": 1451
+    },
+    {
+      "epoch": 0.655160744500846,
+      "grad_norm": 3.0077707371721862,
+      "learning_rate": 5.615120627539387e-07,
+      "loss": 0.4886,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6556119571347998,
+      "grad_norm": 3.5304208905303915,
+      "learning_rate": 5.601986807743387e-07,
+      "loss": 0.5286,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6560631697687536,
+      "grad_norm": 3.6248543822772463,
+      "learning_rate": 5.588862386984509e-07,
+      "loss": 0.6958,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6565143824027073,
+      "grad_norm": 3.153359939961065,
+      "learning_rate": 5.575747393311078e-07,
+      "loss": 0.5553,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6569655950366611,
+      "grad_norm": 3.0726315593606093,
+      "learning_rate": 5.562641854751274e-07,
+      "loss": 0.4999,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6574168076706147,
+      "grad_norm": 2.6907674515493873,
+      "learning_rate": 5.54954579931308e-07,
+      "loss": 0.4409,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6578680203045685,
+      "grad_norm": 2.976942150158183,
+      "learning_rate": 5.536459254984194e-07,
+      "loss": 0.5027,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6583192329385222,
+      "grad_norm": 4.099210213112422,
+      "learning_rate": 5.523382249732009e-07,
+      "loss": 0.5853,
+      "step": 1459
+    },
+    {
+      "epoch": 0.658770445572476,
+      "grad_norm": 3.5743013453464427,
+      "learning_rate": 5.510314811503519e-07,
+      "loss": 0.6072,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6592216582064298,
+      "grad_norm": 3.0791361228178205,
+      "learning_rate": 5.497256968225263e-07,
+      "loss": 0.4773,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6596728708403835,
+      "grad_norm": 3.127066793240839,
+      "learning_rate": 5.484208747803301e-07,
+      "loss": 0.6281,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6601240834743373,
+      "grad_norm": 3.1864970517695275,
+      "learning_rate": 5.4711701781231e-07,
+      "loss": 0.5335,
+      "step": 1463
+    },
+    {
+      "epoch": 0.660575296108291,
+      "grad_norm": 2.7910338972873676,
+      "learning_rate": 5.458141287049525e-07,
+      "loss": 0.4367,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6610265087422448,
+      "grad_norm": 3.112978170764596,
+      "learning_rate": 5.445122102426745e-07,
+      "loss": 0.4787,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6614777213761985,
+      "grad_norm": 3.2428968725230436,
+      "learning_rate": 5.432112652078179e-07,
+      "loss": 0.4979,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6619289340101523,
+      "grad_norm": 3.64761923117337,
+      "learning_rate": 5.419112963806467e-07,
+      "loss": 0.6766,
+      "step": 1467
+    },
+    {
+      "epoch": 0.662380146644106,
+      "grad_norm": 3.1374440807407966,
+      "learning_rate": 5.406123065393351e-07,
+      "loss": 0.585,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6628313592780598,
+      "grad_norm": 3.0743306845398872,
+      "learning_rate": 5.393142984599684e-07,
+      "loss": 0.5432,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6632825719120136,
+      "grad_norm": 3.0220167684458,
+      "learning_rate": 5.380172749165321e-07,
+      "loss": 0.6321,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6637337845459673,
+      "grad_norm": 3.2506207588342115,
+      "learning_rate": 5.367212386809073e-07,
+      "loss": 0.4709,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6641849971799211,
+      "grad_norm": 2.7879147797631463,
+      "learning_rate": 5.354261925228666e-07,
+      "loss": 0.4185,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6646362098138748,
+      "grad_norm": 3.372496137176609,
+      "learning_rate": 5.341321392100655e-07,
+      "loss": 0.5043,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6650874224478286,
+      "grad_norm": 3.3089075981861913,
+      "learning_rate": 5.32839081508038e-07,
+      "loss": 0.5101,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6655386350817822,
+      "grad_norm": 3.141227268377863,
+      "learning_rate": 5.315470221801905e-07,
+      "loss": 0.5693,
+      "step": 1475
+    },
+    {
+      "epoch": 0.665989847715736,
+      "grad_norm": 3.200943952541811,
+      "learning_rate": 5.302559639877952e-07,
+      "loss": 0.5679,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6664410603496898,
+      "grad_norm": 3.264422897311328,
+      "learning_rate": 5.289659096899859e-07,
+      "loss": 0.5888,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6668922729836435,
+      "grad_norm": 3.2617925134864456,
+      "learning_rate": 5.2767686204375e-07,
+      "loss": 0.618,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6673434856175973,
+      "grad_norm": 3.103687607569448,
+      "learning_rate": 5.263888238039234e-07,
+      "loss": 0.4523,
+      "step": 1479
+    },
+    {
+      "epoch": 0.667794698251551,
+      "grad_norm": 3.288349444384094,
+      "learning_rate": 5.251017977231862e-07,
+      "loss": 0.5756,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6682459108855048,
+      "grad_norm": 3.3422954628890325,
+      "learning_rate": 5.238157865520538e-07,
+      "loss": 0.5607,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6686971235194585,
+      "grad_norm": 2.9514424726562414,
+      "learning_rate": 5.225307930388736e-07,
+      "loss": 0.5046,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6691483361534123,
+      "grad_norm": 3.381792256243614,
+      "learning_rate": 5.212468199298177e-07,
+      "loss": 0.5893,
+      "step": 1483
+    },
+    {
+      "epoch": 0.669599548787366,
+      "grad_norm": 2.974690993406971,
+      "learning_rate": 5.199638699688771e-07,
+      "loss": 0.6419,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6700507614213198,
+      "grad_norm": 3.323658905812843,
+      "learning_rate": 5.186819458978577e-07,
+      "loss": 0.575,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6705019740552736,
+      "grad_norm": 3.1513857698424426,
+      "learning_rate": 5.174010504563715e-07,
+      "loss": 0.4734,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6709531866892273,
+      "grad_norm": 2.8289877159247427,
+      "learning_rate": 5.161211863818328e-07,
+      "loss": 0.5453,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6714043993231811,
+      "grad_norm": 3.288966427725167,
+      "learning_rate": 5.148423564094516e-07,
+      "loss": 0.5392,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6718556119571348,
+      "grad_norm": 3.224946055142921,
+      "learning_rate": 5.135645632722276e-07,
+      "loss": 0.5281,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6723068245910886,
+      "grad_norm": 3.5118397978024625,
+      "learning_rate": 5.122878097009459e-07,
+      "loss": 0.4893,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6727580372250423,
+      "grad_norm": 3.288392437565718,
+      "learning_rate": 5.110120984241686e-07,
+      "loss": 0.528,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6732092498589961,
+      "grad_norm": 3.3121054039299254,
+      "learning_rate": 5.097374321682303e-07,
+      "loss": 0.5505,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6736604624929498,
+      "grad_norm": 3.1922673429233317,
+      "learning_rate": 5.084638136572337e-07,
+      "loss": 0.5267,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6741116751269035,
+      "grad_norm": 3.1998010986229546,
+      "learning_rate": 5.071912456130409e-07,
+      "loss": 0.6137,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6745628877608573,
+      "grad_norm": 2.9001454970418448,
+      "learning_rate": 5.059197307552697e-07,
+      "loss": 0.5806,
+      "step": 1495
+    },
+    {
+      "epoch": 0.675014100394811,
+      "grad_norm": 3.381015585223304,
+      "learning_rate": 5.046492718012869e-07,
+      "loss": 0.5509,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6754653130287648,
+      "grad_norm": 3.518395239900884,
+      "learning_rate": 5.033798714662023e-07,
+      "loss": 0.5119,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6759165256627185,
+      "grad_norm": 2.9919505064871705,
+      "learning_rate": 5.021115324628647e-07,
+      "loss": 0.5477,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6763677382966723,
+      "grad_norm": 3.1660202084920095,
+      "learning_rate": 5.008442575018533e-07,
+      "loss": 0.4414,
+      "step": 1499
+    },
+    {
+      "epoch": 0.676818950930626,
+      "grad_norm": 3.3320708896626527,
+      "learning_rate": 4.995780492914736e-07,
+      "loss": 0.6299,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6772701635645798,
+      "grad_norm": 2.872291861282731,
+      "learning_rate": 4.983129105377527e-07,
+      "loss": 0.5124,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6777213761985336,
+      "grad_norm": 3.1930837083504997,
+      "learning_rate": 4.970488439444296e-07,
+      "loss": 0.576,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6781725888324873,
+      "grad_norm": 3.172575535409837,
+      "learning_rate": 4.957858522129544e-07,
+      "loss": 0.5872,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6786238014664411,
+      "grad_norm": 3.0896523132819573,
+      "learning_rate": 4.945239380424786e-07,
+      "loss": 0.5427,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6790750141003948,
+      "grad_norm": 2.8752434836894243,
+      "learning_rate": 4.932631041298513e-07,
+      "loss": 0.505,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6795262267343486,
+      "grad_norm": 2.973393299716313,
+      "learning_rate": 4.920033531696136e-07,
+      "loss": 0.5431,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6799774393683023,
+      "grad_norm": 2.9302261733817225,
+      "learning_rate": 4.907446878539912e-07,
+      "loss": 0.5481,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6804286520022561,
+      "grad_norm": 2.6511702770482266,
+      "learning_rate": 4.894871108728903e-07,
+      "loss": 0.4476,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6808798646362099,
+      "grad_norm": 3.741059117355058,
+      "learning_rate": 4.882306249138909e-07,
+      "loss": 0.5295,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6813310772701635,
+      "grad_norm": 3.1716709800903184,
+      "learning_rate": 4.86975232662241e-07,
+      "loss": 0.6054,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6817822899041173,
+      "grad_norm": 2.6341030249717026,
+      "learning_rate": 4.857209368008524e-07,
+      "loss": 0.5144,
+      "step": 1511
+    },
+    {
+      "epoch": 0.682233502538071,
+      "grad_norm": 3.422325855185031,
+      "learning_rate": 4.844677400102929e-07,
+      "loss": 0.538,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6826847151720248,
+      "grad_norm": 3.1391602045967444,
+      "learning_rate": 4.832156449687811e-07,
+      "loss": 0.5469,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6831359278059785,
+      "grad_norm": 3.7792088207002283,
+      "learning_rate": 4.819646543521824e-07,
+      "loss": 0.579,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6835871404399323,
+      "grad_norm": 3.639662253926238,
+      "learning_rate": 4.807147708340009e-07,
+      "loss": 0.5355,
+      "step": 1515
+    },
+    {
+      "epoch": 0.684038353073886,
+      "grad_norm": 3.0964113537901916,
+      "learning_rate": 4.794659970853748e-07,
+      "loss": 0.5282,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6844895657078398,
+      "grad_norm": 3.3232471100038894,
+      "learning_rate": 4.782183357750707e-07,
+      "loss": 0.6801,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6849407783417936,
+      "grad_norm": 3.3328995149094305,
+      "learning_rate": 4.769717895694774e-07,
+      "loss": 0.5651,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6853919909757473,
+      "grad_norm": 2.9076854833073575,
+      "learning_rate": 4.7572636113260176e-07,
+      "loss": 0.6316,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6858432036097011,
+      "grad_norm": 3.5121904313302874,
+      "learning_rate": 4.744820531260608e-07,
+      "loss": 0.4922,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6862944162436548,
+      "grad_norm": 2.946462024735822,
+      "learning_rate": 4.732388682090771e-07,
+      "loss": 0.4305,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6867456288776086,
+      "grad_norm": 3.3183567408446493,
+      "learning_rate": 4.7199680903847426e-07,
+      "loss": 0.6194,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6871968415115624,
+      "grad_norm": 2.970061784609735,
+      "learning_rate": 4.7075587826866767e-07,
+      "loss": 0.5048,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6876480541455161,
+      "grad_norm": 3.3231922435824757,
+      "learning_rate": 4.695160785516639e-07,
+      "loss": 0.6005,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6880992667794699,
+      "grad_norm": 3.248780126563559,
+      "learning_rate": 4.6827741253705054e-07,
+      "loss": 0.5268,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6885504794134236,
+      "grad_norm": 2.9816778422414076,
+      "learning_rate": 4.670398828719926e-07,
+      "loss": 0.5717,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6890016920473774,
+      "grad_norm": 3.2874556360576883,
+      "learning_rate": 4.658034922012275e-07,
+      "loss": 0.5634,
+      "step": 1527
+    },
+    {
+      "epoch": 0.689452904681331,
+      "grad_norm": 3.2116212084250373,
+      "learning_rate": 4.6456824316705725e-07,
+      "loss": 0.5602,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6899041173152848,
+      "grad_norm": 3.4186841183082364,
+      "learning_rate": 4.6333413840934575e-07,
+      "loss": 0.6244,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6903553299492385,
+      "grad_norm": 3.1280202580241614,
+      "learning_rate": 4.621011805655093e-07,
+      "loss": 0.4692,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6908065425831923,
+      "grad_norm": 2.973990284802939,
+      "learning_rate": 4.60869372270514e-07,
+      "loss": 0.5565,
+      "step": 1531
+    },
+    {
+      "epoch": 0.691257755217146,
+      "grad_norm": 3.7131828396505706,
+      "learning_rate": 4.5963871615687045e-07,
+      "loss": 0.6157,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6917089678510998,
+      "grad_norm": 2.8869266707436823,
+      "learning_rate": 4.584092148546254e-07,
+      "loss": 0.6354,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6921601804850536,
+      "grad_norm": 3.3918689462338265,
+      "learning_rate": 4.5718087099135773e-07,
+      "loss": 0.5702,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6926113931190073,
+      "grad_norm": 3.2001428288161344,
+      "learning_rate": 4.5595368719217397e-07,
+      "loss": 0.5083,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6930626057529611,
+      "grad_norm": 3.084523211589884,
+      "learning_rate": 4.5472766607970024e-07,
+      "loss": 0.5727,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6935138183869148,
+      "grad_norm": 3.157807512575682,
+      "learning_rate": 4.5350281027407843e-07,
+      "loss": 0.5678,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6939650310208686,
+      "grad_norm": 3.2071372806652905,
+      "learning_rate": 4.522791223929596e-07,
+      "loss": 0.5597,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6944162436548224,
+      "grad_norm": 3.225339649209973,
+      "learning_rate": 4.51056605051499e-07,
+      "loss": 0.5693,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6948674562887761,
+      "grad_norm": 3.0766839452947856,
+      "learning_rate": 4.4983526086235103e-07,
+      "loss": 0.5093,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6953186689227299,
+      "grad_norm": 3.3661435682809535,
+      "learning_rate": 4.4861509243566164e-07,
+      "loss": 0.6246,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6957698815566836,
+      "grad_norm": 2.782237105741097,
+      "learning_rate": 4.4739610237906545e-07,
+      "loss": 0.5195,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6962210941906374,
+      "grad_norm": 3.1390391034941563,
+      "learning_rate": 4.461782932976782e-07,
+      "loss": 0.6254,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6966723068245911,
+      "grad_norm": 2.837001264646572,
+      "learning_rate": 4.4496166779409026e-07,
+      "loss": 0.5682,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6971235194585449,
+      "grad_norm": 3.151679334520682,
+      "learning_rate": 4.437462284683653e-07,
+      "loss": 0.5967,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6975747320924985,
+      "grad_norm": 3.2614811697248096,
+      "learning_rate": 4.4253197791802967e-07,
+      "loss": 0.4696,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6980259447264523,
+      "grad_norm": 3.1064249116523706,
+      "learning_rate": 4.4131891873807103e-07,
+      "loss": 0.5832,
+      "step": 1547
+    },
+    {
+      "epoch": 0.698477157360406,
+      "grad_norm": 3.7129814910314884,
+      "learning_rate": 4.401070535209296e-07,
+      "loss": 0.6575,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6989283699943598,
+      "grad_norm": 3.5598656251240293,
+      "learning_rate": 4.3889638485649405e-07,
+      "loss": 0.5255,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6993795826283136,
+      "grad_norm": 3.0419102505278355,
+      "learning_rate": 4.3768691533209735e-07,
+      "loss": 0.5589,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6998307952622673,
+      "grad_norm": 3.250677316394861,
+      "learning_rate": 4.364786475325072e-07,
+      "loss": 0.4775,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7002820078962211,
+      "grad_norm": 3.012263276867671,
+      "learning_rate": 4.3527158403992567e-07,
+      "loss": 0.503,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7007332205301748,
+      "grad_norm": 3.120176925995655,
+      "learning_rate": 4.3406572743397975e-07,
+      "loss": 0.6046,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7011844331641286,
+      "grad_norm": 3.134496372403675,
+      "learning_rate": 4.3286108029171685e-07,
+      "loss": 0.5061,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7016356457980824,
+      "grad_norm": 3.5153313421474373,
+      "learning_rate": 4.3165764518760104e-07,
+      "loss": 0.556,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7020868584320361,
+      "grad_norm": 3.361928587880027,
+      "learning_rate": 4.304554246935049e-07,
+      "loss": 0.5555,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7025380710659899,
+      "grad_norm": 3.113389444703251,
+      "learning_rate": 4.292544213787056e-07,
+      "loss": 0.508,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7029892836999436,
+      "grad_norm": 2.951119900702608,
+      "learning_rate": 4.280546378098792e-07,
+      "loss": 0.4627,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7034404963338974,
+      "grad_norm": 3.3068113437135747,
+      "learning_rate": 4.2685607655109446e-07,
+      "loss": 0.4641,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7038917089678511,
+      "grad_norm": 3.443473545652759,
+      "learning_rate": 4.256587401638091e-07,
+      "loss": 0.5881,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7043429216018049,
+      "grad_norm": 3.672614498272048,
+      "learning_rate": 4.244626312068622e-07,
+      "loss": 0.5562,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7047941342357587,
+      "grad_norm": 3.1807964103718316,
+      "learning_rate": 4.232677522364696e-07,
+      "loss": 0.514,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7052453468697123,
+      "grad_norm": 2.8379876257690118,
+      "learning_rate": 4.220741058062194e-07,
+      "loss": 0.6209,
+      "step": 1563
+    },
+    {
+      "epoch": 0.705696559503666,
+      "grad_norm": 3.115298174903329,
+      "learning_rate": 4.2088169446706487e-07,
+      "loss": 0.5598,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7061477721376198,
+      "grad_norm": 3.616655218693416,
+      "learning_rate": 4.1969052076732005e-07,
+      "loss": 0.5609,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7065989847715736,
+      "grad_norm": 3.209766337154335,
+      "learning_rate": 4.185005872526538e-07,
+      "loss": 0.5969,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7070501974055273,
+      "grad_norm": 3.469400141091681,
+      "learning_rate": 4.1731189646608434e-07,
+      "loss": 0.6814,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7075014100394811,
+      "grad_norm": 3.1397943853367374,
+      "learning_rate": 4.161244509479751e-07,
+      "loss": 0.5017,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7079526226734348,
+      "grad_norm": 3.5877431815311436,
+      "learning_rate": 4.1493825323602737e-07,
+      "loss": 0.6132,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7084038353073886,
+      "grad_norm": 2.681923696233627,
+      "learning_rate": 4.137533058652758e-07,
+      "loss": 0.516,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7088550479413424,
+      "grad_norm": 3.389585300189951,
+      "learning_rate": 4.12569611368083e-07,
+      "loss": 0.5234,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7093062605752961,
+      "grad_norm": 3.180876954986152,
+      "learning_rate": 4.113871722741337e-07,
+      "loss": 0.4276,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7097574732092499,
+      "grad_norm": 3.932081857493054,
+      "learning_rate": 4.1020599111043084e-07,
+      "loss": 0.5718,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7102086858432036,
+      "grad_norm": 3.0660045119592416,
+      "learning_rate": 4.09026070401288e-07,
+      "loss": 0.4444,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7106598984771574,
+      "grad_norm": 2.9150397257470377,
+      "learning_rate": 4.078474126683249e-07,
+      "loss": 0.5647,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 3.260434948896054,
+      "learning_rate": 4.0667002043046304e-07,
+      "loss": 0.5476,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7115623237450649,
+      "grad_norm": 3.2788448633519867,
+      "learning_rate": 4.0549389620391896e-07,
+      "loss": 0.5775,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7120135363790187,
+      "grad_norm": 3.0034190041080064,
+      "learning_rate": 4.0431904250219893e-07,
+      "loss": 0.5515,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7124647490129724,
+      "grad_norm": 3.2382651735959223,
+      "learning_rate": 4.031454618360944e-07,
+      "loss": 0.4369,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7129159616469262,
+      "grad_norm": 3.0531108303993135,
+      "learning_rate": 4.0197315671367593e-07,
+      "loss": 0.5064,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7133671742808798,
+      "grad_norm": 2.9453398873817025,
+      "learning_rate": 4.008021296402888e-07,
+      "loss": 0.427,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7138183869148336,
+      "grad_norm": 3.265675895615181,
+      "learning_rate": 3.996323831185462e-07,
+      "loss": 0.6767,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7142695995487873,
+      "grad_norm": 2.9358827987112535,
+      "learning_rate": 3.984639196483245e-07,
+      "loss": 0.5785,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7147208121827411,
+      "grad_norm": 3.3867250177331782,
+      "learning_rate": 3.9729674172675954e-07,
+      "loss": 0.5236,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7151720248166948,
+      "grad_norm": 3.31448568517522,
+      "learning_rate": 3.961308518482372e-07,
+      "loss": 0.609,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7156232374506486,
+      "grad_norm": 3.71248092590625,
+      "learning_rate": 3.949662525043934e-07,
+      "loss": 0.6091,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7160744500846024,
+      "grad_norm": 3.8185840047248125,
+      "learning_rate": 3.9380294618410434e-07,
+      "loss": 0.6641,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7165256627185561,
+      "grad_norm": 3.3840710545866632,
+      "learning_rate": 3.92640935373483e-07,
+      "loss": 0.5043,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7169768753525099,
+      "grad_norm": 3.109852062826725,
+      "learning_rate": 3.9148022255587487e-07,
+      "loss": 0.5068,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7174280879864636,
+      "grad_norm": 3.0466909947937184,
+      "learning_rate": 3.9032081021185024e-07,
+      "loss": 0.8023,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7178793006204174,
+      "grad_norm": 3.494445337066531,
+      "learning_rate": 3.891627008192007e-07,
+      "loss": 0.6285,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7183305132543711,
+      "grad_norm": 3.0639765960841387,
+      "learning_rate": 3.8800589685293294e-07,
+      "loss": 0.4803,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7187817258883249,
+      "grad_norm": 3.370717647642269,
+      "learning_rate": 3.868504007852641e-07,
+      "loss": 0.5468,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7192329385222787,
+      "grad_norm": 3.252865929125753,
+      "learning_rate": 3.8569621508561666e-07,
+      "loss": 0.5753,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7196841511562324,
+      "grad_norm": 3.1394814071549484,
+      "learning_rate": 3.845433422206118e-07,
+      "loss": 0.4729,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7201353637901862,
+      "grad_norm": 3.1631245452667973,
+      "learning_rate": 3.833917846540651e-07,
+      "loss": 0.6474,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7205865764241399,
+      "grad_norm": 3.162870418400208,
+      "learning_rate": 3.8224154484698234e-07,
+      "loss": 0.4195,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7210377890580937,
+      "grad_norm": 3.161335436131402,
+      "learning_rate": 3.8109262525755183e-07,
+      "loss": 0.5058,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7214890016920473,
+      "grad_norm": 3.3185670575255433,
+      "learning_rate": 3.7994502834114085e-07,
+      "loss": 0.4979,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7219402143260011,
+      "grad_norm": 3.176929378675548,
+      "learning_rate": 3.7879875655029017e-07,
+      "loss": 0.5728,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7223914269599548,
+      "grad_norm": 3.1684081635191066,
+      "learning_rate": 3.7765381233470794e-07,
+      "loss": 0.4601,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7228426395939086,
+      "grad_norm": 3.127695957202654,
+      "learning_rate": 3.765101981412665e-07,
+      "loss": 0.5002,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7232938522278624,
+      "grad_norm": 3.219550164344783,
+      "learning_rate": 3.753679164139947e-07,
+      "loss": 0.5385,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7237450648618161,
+      "grad_norm": 3.601540512399754,
+      "learning_rate": 3.742269695940734e-07,
+      "loss": 0.664,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7241962774957699,
+      "grad_norm": 3.375042559176746,
+      "learning_rate": 3.730873601198326e-07,
+      "loss": 0.6187,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7246474901297236,
+      "grad_norm": 3.0597581559151794,
+      "learning_rate": 3.7194909042674115e-07,
+      "loss": 0.4807,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7250987027636774,
+      "grad_norm": 2.9753785627396105,
+      "learning_rate": 3.708121629474077e-07,
+      "loss": 0.5375,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7255499153976311,
+      "grad_norm": 2.9762038394272765,
+      "learning_rate": 3.6967658011157054e-07,
+      "loss": 0.4504,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7260011280315849,
+      "grad_norm": 3.1965956621894063,
+      "learning_rate": 3.6854234434609477e-07,
+      "loss": 0.5014,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7264523406655387,
+      "grad_norm": 3.499485873703523,
+      "learning_rate": 3.6740945807496736e-07,
+      "loss": 0.4979,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7269035532994924,
+      "grad_norm": 3.2588338513217274,
+      "learning_rate": 3.662779237192899e-07,
+      "loss": 0.5223,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7273547659334462,
+      "grad_norm": 3.0490054858602584,
+      "learning_rate": 3.6514774369727676e-07,
+      "loss": 0.5,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7278059785673999,
+      "grad_norm": 3.12273878068711,
+      "learning_rate": 3.6401892042424556e-07,
+      "loss": 0.5755,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7282571912013537,
+      "grad_norm": 3.073986060170658,
+      "learning_rate": 3.6289145631261554e-07,
+      "loss": 0.5491,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7287084038353074,
+      "grad_norm": 3.290781958913942,
+      "learning_rate": 3.617653537719022e-07,
+      "loss": 0.5943,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7291596164692611,
+      "grad_norm": 3.400579286261728,
+      "learning_rate": 3.606406152087095e-07,
+      "loss": 0.5581,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7296108291032148,
+      "grad_norm": 3.450947773384265,
+      "learning_rate": 3.595172430267279e-07,
+      "loss": 0.5873,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7300620417371686,
+      "grad_norm": 2.8791493307158014,
+      "learning_rate": 3.583952396267269e-07,
+      "loss": 0.5159,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7305132543711224,
+      "grad_norm": 3.3597310964990994,
+      "learning_rate": 3.572746074065509e-07,
+      "loss": 0.5292,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7309644670050761,
+      "grad_norm": 3.2911307641618457,
+      "learning_rate": 3.56155348761114e-07,
+      "loss": 0.6023,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7314156796390299,
+      "grad_norm": 3.208895269155749,
+      "learning_rate": 3.550374660823948e-07,
+      "loss": 0.6399,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7318668922729836,
+      "grad_norm": 3.411723131804797,
+      "learning_rate": 3.539209617594311e-07,
+      "loss": 0.5386,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7323181049069374,
+      "grad_norm": 3.017726904850449,
+      "learning_rate": 3.5280583817831577e-07,
+      "loss": 0.5199,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7327693175408911,
+      "grad_norm": 3.448141993818582,
+      "learning_rate": 3.516920977221898e-07,
+      "loss": 0.6594,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7332205301748449,
+      "grad_norm": 3.1018446805384268,
+      "learning_rate": 3.5057974277123935e-07,
+      "loss": 0.432,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7336717428087987,
+      "grad_norm": 3.292832340665522,
+      "learning_rate": 3.494687757026894e-07,
+      "loss": 0.6043,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7341229554427524,
+      "grad_norm": 3.37654871788207,
+      "learning_rate": 3.483591988907972e-07,
+      "loss": 0.6076,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7345741680767062,
+      "grad_norm": 2.9139739574958794,
+      "learning_rate": 3.472510147068515e-07,
+      "loss": 0.5498,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7350253807106599,
+      "grad_norm": 3.0177269585869415,
+      "learning_rate": 3.461442255191628e-07,
+      "loss": 0.4855,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7354765933446137,
+      "grad_norm": 3.5169507921427083,
+      "learning_rate": 3.4503883369306176e-07,
+      "loss": 0.5087,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7359278059785674,
+      "grad_norm": 3.028359718330194,
+      "learning_rate": 3.439348415908918e-07,
+      "loss": 0.5116,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7363790186125212,
+      "grad_norm": 3.3434023450295416,
+      "learning_rate": 3.42832251572005e-07,
+      "loss": 0.623,
+      "step": 1632
+    },
+    {
+      "epoch": 0.736830231246475,
+      "grad_norm": 3.0861385669753822,
+      "learning_rate": 3.4173106599275827e-07,
+      "loss": 0.5863,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7372814438804286,
+      "grad_norm": 3.4871736340602344,
+      "learning_rate": 3.406312872065047e-07,
+      "loss": 0.6979,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7377326565143824,
+      "grad_norm": 3.3007203094285322,
+      "learning_rate": 3.395329175635935e-07,
+      "loss": 0.5642,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7381838691483361,
+      "grad_norm": 3.042983266092788,
+      "learning_rate": 3.384359594113606e-07,
+      "loss": 0.5794,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7386350817822899,
+      "grad_norm": 3.1322148623384174,
+      "learning_rate": 3.373404150941258e-07,
+      "loss": 0.5112,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7390862944162436,
+      "grad_norm": 3.3451076966064037,
+      "learning_rate": 3.3624628695318845e-07,
+      "loss": 0.5563,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7395375070501974,
+      "grad_norm": 2.898388910225105,
+      "learning_rate": 3.3515357732682e-07,
+      "loss": 0.5745,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7399887196841511,
+      "grad_norm": 3.3344993868159536,
+      "learning_rate": 3.34062288550261e-07,
+      "loss": 0.5693,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7404399323181049,
+      "grad_norm": 3.399041172987303,
+      "learning_rate": 3.3297242295571526e-07,
+      "loss": 0.4952,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7408911449520587,
+      "grad_norm": 3.552119321092643,
+      "learning_rate": 3.3188398287234496e-07,
+      "loss": 0.5461,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7413423575860124,
+      "grad_norm": 3.1787290156115406,
+      "learning_rate": 3.307969706262669e-07,
+      "loss": 0.6037,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7417935702199662,
+      "grad_norm": 3.5000027663014492,
+      "learning_rate": 3.2971138854054505e-07,
+      "loss": 0.5315,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7422447828539199,
+      "grad_norm": 3.1720666447080808,
+      "learning_rate": 3.286272389351874e-07,
+      "loss": 0.5329,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7426959954878737,
+      "grad_norm": 3.09152659928442,
+      "learning_rate": 3.2754452412714153e-07,
+      "loss": 0.4382,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7431472081218274,
+      "grad_norm": 2.798294812436822,
+      "learning_rate": 3.2646324643028664e-07,
+      "loss": 0.4975,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7435984207557812,
+      "grad_norm": 3.1293398353599393,
+      "learning_rate": 3.2538340815543287e-07,
+      "loss": 0.463,
+      "step": 1648
+    },
+    {
+      "epoch": 0.744049633389735,
+      "grad_norm": 3.291575473368299,
+      "learning_rate": 3.243050116103128e-07,
+      "loss": 0.4967,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7445008460236887,
+      "grad_norm": 3.1465605176299825,
+      "learning_rate": 3.2322805909957795e-07,
+      "loss": 0.5417,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7449520586576425,
+      "grad_norm": 2.9534255260466624,
+      "learning_rate": 3.221525529247949e-07,
+      "loss": 0.484,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7454032712915961,
+      "grad_norm": 3.2799558091998215,
+      "learning_rate": 3.21078495384438e-07,
+      "loss": 0.5042,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7458544839255499,
+      "grad_norm": 3.3413912958342102,
+      "learning_rate": 3.20005888773886e-07,
+      "loss": 0.5555,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7463056965595036,
+      "grad_norm": 3.5148486290813548,
+      "learning_rate": 3.189347353854173e-07,
+      "loss": 0.5452,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7467569091934574,
+      "grad_norm": 3.235645631726398,
+      "learning_rate": 3.178650375082038e-07,
+      "loss": 0.5124,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7472081218274111,
+      "grad_norm": 3.335777914025934,
+      "learning_rate": 3.16796797428308e-07,
+      "loss": 0.551,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7476593344613649,
+      "grad_norm": 3.3336251599770117,
+      "learning_rate": 3.157300174286759e-07,
+      "loss": 0.4606,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7481105470953187,
+      "grad_norm": 3.3954076174897683,
+      "learning_rate": 3.1466469978913325e-07,
+      "loss": 0.5403,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7485617597292724,
+      "grad_norm": 3.1227703274844614,
+      "learning_rate": 3.136008467863814e-07,
+      "loss": 0.4239,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7490129723632262,
+      "grad_norm": 3.144214684776084,
+      "learning_rate": 3.125384606939908e-07,
+      "loss": 0.4472,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7494641849971799,
+      "grad_norm": 3.075384566057599,
+      "learning_rate": 3.114775437823971e-07,
+      "loss": 0.4441,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7499153976311337,
+      "grad_norm": 3.407644282112315,
+      "learning_rate": 3.104180983188963e-07,
+      "loss": 0.6542,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7503666102650874,
+      "grad_norm": 3.0829123476918703,
+      "learning_rate": 3.0936012656763933e-07,
+      "loss": 0.5066,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7508178228990412,
+      "grad_norm": 3.197121809224084,
+      "learning_rate": 3.0830363078962853e-07,
+      "loss": 0.5604,
+      "step": 1664
+    },
+    {
+      "epoch": 0.751269035532995,
+      "grad_norm": 3.4708630812887162,
+      "learning_rate": 3.0724861324271136e-07,
+      "loss": 0.4919,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7517202481669487,
+      "grad_norm": 3.2521891509707173,
+      "learning_rate": 3.061950761815755e-07,
+      "loss": 0.6188,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7521714608009025,
+      "grad_norm": 3.047231852620941,
+      "learning_rate": 3.0514302185774653e-07,
+      "loss": 0.5921,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7526226734348562,
+      "grad_norm": 3.2765361212999204,
+      "learning_rate": 3.040924525195786e-07,
+      "loss": 0.5818,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7530738860688099,
+      "grad_norm": 3.299009096476747,
+      "learning_rate": 3.030433704122549e-07,
+      "loss": 0.5344,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7535250987027636,
+      "grad_norm": 2.905729407389529,
+      "learning_rate": 3.0199577777777875e-07,
+      "loss": 0.5905,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7539763113367174,
+      "grad_norm": 2.886347624748166,
+      "learning_rate": 3.0094967685497017e-07,
+      "loss": 0.5058,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7544275239706711,
+      "grad_norm": 3.3966367759928184,
+      "learning_rate": 2.999050698794624e-07,
+      "loss": 0.6179,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7548787366046249,
+      "grad_norm": 3.6135542251703874,
+      "learning_rate": 2.9886195908369504e-07,
+      "loss": 0.5298,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7553299492385787,
+      "grad_norm": 3.3688596527662864,
+      "learning_rate": 2.9782034669691027e-07,
+      "loss": 0.5498,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7557811618725324,
+      "grad_norm": 3.2566793348820537,
+      "learning_rate": 2.9678023494514815e-07,
+      "loss": 0.5749,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7562323745064862,
+      "grad_norm": 3.0116943004039576,
+      "learning_rate": 2.957416260512414e-07,
+      "loss": 0.4711,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7566835871404399,
+      "grad_norm": 3.52833333229487,
+      "learning_rate": 2.94704522234812e-07,
+      "loss": 0.5615,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7571347997743937,
+      "grad_norm": 3.075366987645075,
+      "learning_rate": 2.936689257122642e-07,
+      "loss": 0.485,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7575860124083474,
+      "grad_norm": 3.3966268101373216,
+      "learning_rate": 2.926348386967813e-07,
+      "loss": 0.6323,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7580372250423012,
+      "grad_norm": 2.8446858198235483,
+      "learning_rate": 2.9160226339832137e-07,
+      "loss": 0.4601,
+      "step": 1680
+    },
+    {
+      "epoch": 0.758488437676255,
+      "grad_norm": 3.3602462761752934,
+      "learning_rate": 2.9057120202361094e-07,
+      "loss": 0.4662,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7589396503102087,
+      "grad_norm": 3.3791987672981794,
+      "learning_rate": 2.895416567761414e-07,
+      "loss": 0.6081,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7593908629441625,
+      "grad_norm": 3.4250371237897217,
+      "learning_rate": 2.8851362985616387e-07,
+      "loss": 0.5844,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7598420755781162,
+      "grad_norm": 3.46770568979179,
+      "learning_rate": 2.874871234606846e-07,
+      "loss": 0.5083,
+      "step": 1684
+    },
+    {
+      "epoch": 0.76029328821207,
+      "grad_norm": 3.4131489026616357,
+      "learning_rate": 2.86462139783461e-07,
+      "loss": 0.5537,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7607445008460237,
+      "grad_norm": 3.270548918149311,
+      "learning_rate": 2.854386810149955e-07,
+      "loss": 0.4998,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7611957134799774,
+      "grad_norm": 3.530660029135758,
+      "learning_rate": 2.8441674934253135e-07,
+      "loss": 0.5188,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7616469261139311,
+      "grad_norm": 3.386488218146081,
+      "learning_rate": 2.8339634695005e-07,
+      "loss": 0.529,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7620981387478849,
+      "grad_norm": 3.1388677063133716,
+      "learning_rate": 2.823774760182619e-07,
+      "loss": 0.6524,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7625493513818387,
+      "grad_norm": 3.4194129454744906,
+      "learning_rate": 2.813601387246073e-07,
+      "loss": 0.6083,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7630005640157924,
+      "grad_norm": 3.253968578183408,
+      "learning_rate": 2.8034433724324715e-07,
+      "loss": 0.5288,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7634517766497462,
+      "grad_norm": 2.8786430697629752,
+      "learning_rate": 2.7933007374506045e-07,
+      "loss": 0.4924,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7639029892836999,
+      "grad_norm": 3.7085567077881403,
+      "learning_rate": 2.783173503976405e-07,
+      "loss": 0.6246,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7643542019176537,
+      "grad_norm": 2.977804758799455,
+      "learning_rate": 2.7730616936528763e-07,
+      "loss": 0.5176,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7648054145516074,
+      "grad_norm": 3.327404515557504,
+      "learning_rate": 2.7629653280900744e-07,
+      "loss": 0.5808,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7652566271855612,
+      "grad_norm": 3.069935757487436,
+      "learning_rate": 2.7528844288650345e-07,
+      "loss": 0.4892,
+      "step": 1696
+    },
+    {
+      "epoch": 0.765707839819515,
+      "grad_norm": 3.2131875481565553,
+      "learning_rate": 2.7428190175217437e-07,
+      "loss": 0.5039,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7661590524534687,
+      "grad_norm": 2.9817468613273737,
+      "learning_rate": 2.7327691155710976e-07,
+      "loss": 0.5042,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7666102650874225,
+      "grad_norm": 3.3064977092399097,
+      "learning_rate": 2.7227347444908347e-07,
+      "loss": 0.4653,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7670614777213762,
+      "grad_norm": 3.0238137206901183,
+      "learning_rate": 2.7127159257255136e-07,
+      "loss": 0.464,
+      "step": 1700
+    },
+    {
+      "epoch": 0.76751269035533,
+      "grad_norm": 3.169051965683476,
+      "learning_rate": 2.7027126806864465e-07,
+      "loss": 0.708,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7679639029892837,
+      "grad_norm": 3.133987039383184,
+      "learning_rate": 2.692725030751668e-07,
+      "loss": 0.5191,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7684151156232375,
+      "grad_norm": 3.1774428494885063,
+      "learning_rate": 2.6827529972658816e-07,
+      "loss": 0.5386,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7688663282571913,
+      "grad_norm": 3.1659341840204474,
+      "learning_rate": 2.6727966015404144e-07,
+      "loss": 0.5514,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7693175408911449,
+      "grad_norm": 3.567018735324809,
+      "learning_rate": 2.662855864853184e-07,
+      "loss": 0.5817,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7697687535250987,
+      "grad_norm": 3.1177868647304514,
+      "learning_rate": 2.6529308084486334e-07,
+      "loss": 0.5047,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7702199661590524,
+      "grad_norm": 3.6947986746375534,
+      "learning_rate": 2.643021453537695e-07,
+      "loss": 0.5996,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7706711787930062,
+      "grad_norm": 3.026216223845999,
+      "learning_rate": 2.633127821297754e-07,
+      "loss": 0.5287,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7711223914269599,
+      "grad_norm": 3.4446775186050633,
+      "learning_rate": 2.623249932872589e-07,
+      "loss": 0.5906,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7715736040609137,
+      "grad_norm": 3.5855936274951534,
+      "learning_rate": 2.6133878093723227e-07,
+      "loss": 0.5337,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7720248166948674,
+      "grad_norm": 3.28611792990953,
+      "learning_rate": 2.603541471873405e-07,
+      "loss": 0.552,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7724760293288212,
+      "grad_norm": 3.4992707402338823,
+      "learning_rate": 2.5937109414185364e-07,
+      "loss": 0.4644,
+      "step": 1712
+    },
+    {
+      "epoch": 0.772927241962775,
+      "grad_norm": 2.9170590421526104,
+      "learning_rate": 2.583896239016643e-07,
+      "loss": 0.4364,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7733784545967287,
+      "grad_norm": 2.7757056581823782,
+      "learning_rate": 2.5740973856428205e-07,
+      "loss": 0.55,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7738296672306825,
+      "grad_norm": 3.4368687127239195,
+      "learning_rate": 2.56431440223829e-07,
+      "loss": 0.6541,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7742808798646362,
+      "grad_norm": 3.655602500119933,
+      "learning_rate": 2.5545473097103723e-07,
+      "loss": 0.5804,
+      "step": 1716
+    },
+    {
+      "epoch": 0.77473209249859,
+      "grad_norm": 2.9071464976070995,
+      "learning_rate": 2.5447961289324024e-07,
+      "loss": 0.4734,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7751833051325437,
+      "grad_norm": 3.0403001990908383,
+      "learning_rate": 2.5350608807437356e-07,
+      "loss": 0.4796,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7756345177664975,
+      "grad_norm": 3.401670150665487,
+      "learning_rate": 2.525341585949662e-07,
+      "loss": 0.6389,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7760857304004513,
+      "grad_norm": 3.2058804560472125,
+      "learning_rate": 2.5156382653213783e-07,
+      "loss": 0.5616,
+      "step": 1720
+    },
+    {
+      "epoch": 0.776536943034405,
+      "grad_norm": 2.9830405398098097,
+      "learning_rate": 2.5059509395959523e-07,
+      "loss": 0.4926,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7769881556683587,
+      "grad_norm": 3.120574964296058,
+      "learning_rate": 2.496279629476261e-07,
+      "loss": 0.5861,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7774393683023124,
+      "grad_norm": 3.390755640754869,
+      "learning_rate": 2.4866243556309554e-07,
+      "loss": 0.5654,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7778905809362662,
+      "grad_norm": 2.892098028372284,
+      "learning_rate": 2.476985138694415e-07,
+      "loss": 0.4677,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7783417935702199,
+      "grad_norm": 3.250949839615332,
+      "learning_rate": 2.467361999266704e-07,
+      "loss": 0.4674,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7787930062041737,
+      "grad_norm": 3.149351653501838,
+      "learning_rate": 2.4577549579135317e-07,
+      "loss": 0.5966,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7792442188381274,
+      "grad_norm": 2.983139524124556,
+      "learning_rate": 2.448164035166199e-07,
+      "loss": 0.4793,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7796954314720812,
+      "grad_norm": 3.5014980256358172,
+      "learning_rate": 2.438589251521558e-07,
+      "loss": 0.5634,
+      "step": 1728
+    },
+    {
+      "epoch": 0.780146644106035,
+      "grad_norm": 2.9881299554568757,
+      "learning_rate": 2.4290306274419793e-07,
+      "loss": 0.5164,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7805978567399887,
+      "grad_norm": 3.116274993285628,
+      "learning_rate": 2.419488183355284e-07,
+      "loss": 0.5957,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7810490693739425,
+      "grad_norm": 3.148408583672423,
+      "learning_rate": 2.409961939654729e-07,
+      "loss": 0.4581,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7815002820078962,
+      "grad_norm": 2.9310980700141784,
+      "learning_rate": 2.40045191669894e-07,
+      "loss": 0.5326,
+      "step": 1732
+    },
+    {
+      "epoch": 0.78195149464185,
+      "grad_norm": 2.841995600846533,
+      "learning_rate": 2.39095813481188e-07,
+      "loss": 0.3975,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7824027072758037,
+      "grad_norm": 3.360507399857603,
+      "learning_rate": 2.381480614282807e-07,
+      "loss": 0.5217,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7828539199097575,
+      "grad_norm": 3.3345269121171786,
+      "learning_rate": 2.37201937536622e-07,
+      "loss": 0.6034,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7833051325437113,
+      "grad_norm": 3.2457173184176327,
+      "learning_rate": 2.362574438281827e-07,
+      "loss": 0.5572,
+      "step": 1736
+    },
+    {
+      "epoch": 0.783756345177665,
+      "grad_norm": 2.5998859002425654,
+      "learning_rate": 2.353145823214495e-07,
+      "loss": 0.4872,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7842075578116188,
+      "grad_norm": 3.3464133340861433,
+      "learning_rate": 2.3437335503142063e-07,
+      "loss": 0.5293,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7846587704455725,
+      "grad_norm": 3.4826017860389027,
+      "learning_rate": 2.3343376396960278e-07,
+      "loss": 0.487,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7851099830795262,
+      "grad_norm": 3.359966006460181,
+      "learning_rate": 2.3249581114400507e-07,
+      "loss": 0.6411,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7855611957134799,
+      "grad_norm": 3.108189759258103,
+      "learning_rate": 2.3155949855913515e-07,
+      "loss": 0.5771,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7860124083474337,
+      "grad_norm": 3.206925034569629,
+      "learning_rate": 2.306248282159965e-07,
+      "loss": 0.5679,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7864636209813874,
+      "grad_norm": 3.174330922672863,
+      "learning_rate": 2.2969180211208195e-07,
+      "loss": 0.5643,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7869148336153412,
+      "grad_norm": 3.1224925255844,
+      "learning_rate": 2.2876042224137081e-07,
+      "loss": 0.5932,
+      "step": 1744
+    },
+    {
+      "epoch": 0.787366046249295,
+      "grad_norm": 3.1025659930391107,
+      "learning_rate": 2.2783069059432415e-07,
+      "loss": 0.5617,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7878172588832487,
+      "grad_norm": 3.4166793932429416,
+      "learning_rate": 2.2690260915788029e-07,
+      "loss": 0.6249,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7882684715172025,
+      "grad_norm": 3.62085438351349,
+      "learning_rate": 2.2597617991545158e-07,
+      "loss": 0.5649,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7887196841511562,
+      "grad_norm": 3.5339972671901885,
+      "learning_rate": 2.2505140484691897e-07,
+      "loss": 0.6335,
+      "step": 1748
+    },
+    {
+      "epoch": 0.78917089678511,
+      "grad_norm": 3.4985614392697415,
+      "learning_rate": 2.2412828592862798e-07,
+      "loss": 0.6082,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7896221094190637,
+      "grad_norm": 3.267096741544989,
+      "learning_rate": 2.2320682513338595e-07,
+      "loss": 0.5948,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7900733220530175,
+      "grad_norm": 3.0173344659741486,
+      "learning_rate": 2.2228702443045454e-07,
+      "loss": 0.5234,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7905245346869713,
+      "grad_norm": 2.7526133536838358,
+      "learning_rate": 2.213688857855499e-07,
+      "loss": 0.4142,
+      "step": 1752
+    },
+    {
+      "epoch": 0.790975747320925,
+      "grad_norm": 3.509571340813883,
+      "learning_rate": 2.2045241116083467e-07,
+      "loss": 0.602,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7914269599548788,
+      "grad_norm": 3.3780395439717323,
+      "learning_rate": 2.195376025149156e-07,
+      "loss": 0.5572,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7918781725888325,
+      "grad_norm": 3.3496281451911325,
+      "learning_rate": 2.1862446180283966e-07,
+      "loss": 0.5613,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7923293852227863,
+      "grad_norm": 3.0508967113065713,
+      "learning_rate": 2.1771299097608864e-07,
+      "loss": 0.47,
+      "step": 1756
+    },
+    {
+      "epoch": 0.79278059785674,
+      "grad_norm": 3.1585837200192697,
+      "learning_rate": 2.1680319198257568e-07,
+      "loss": 0.4856,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7932318104906937,
+      "grad_norm": 2.8592335827975366,
+      "learning_rate": 2.1589506676664125e-07,
+      "loss": 0.5273,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7936830231246474,
+      "grad_norm": 3.5664002528716128,
+      "learning_rate": 2.1498861726904838e-07,
+      "loss": 0.4738,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7941342357586012,
+      "grad_norm": 3.3257110040445705,
+      "learning_rate": 2.1408384542697953e-07,
+      "loss": 0.5896,
+      "step": 1760
+    },
+    {
+      "epoch": 0.794585448392555,
+      "grad_norm": 3.2893705188785627,
+      "learning_rate": 2.131807531740315e-07,
+      "loss": 0.5908,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7950366610265087,
+      "grad_norm": 3.2439198495340724,
+      "learning_rate": 2.1227934244021106e-07,
+      "loss": 0.5418,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7954878736604625,
+      "grad_norm": 2.890014525204039,
+      "learning_rate": 2.113796151519327e-07,
+      "loss": 0.5701,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7959390862944162,
+      "grad_norm": 2.9555433016653243,
+      "learning_rate": 2.10481573232012e-07,
+      "loss": 0.5724,
+      "step": 1764
+    },
+    {
+      "epoch": 0.79639029892837,
+      "grad_norm": 3.1595499191577985,
+      "learning_rate": 2.0958521859966317e-07,
+      "loss": 0.58,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7968415115623237,
+      "grad_norm": 3.2406481995005683,
+      "learning_rate": 2.0869055317049454e-07,
+      "loss": 0.5815,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7972927241962775,
+      "grad_norm": 3.013567529689706,
+      "learning_rate": 2.0779757885650407e-07,
+      "loss": 0.4969,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7977439368302313,
+      "grad_norm": 3.285613814118948,
+      "learning_rate": 2.0690629756607647e-07,
+      "loss": 0.5407,
+      "step": 1768
+    },
+    {
+      "epoch": 0.798195149464185,
+      "grad_norm": 3.05321531009751,
+      "learning_rate": 2.0601671120397747e-07,
+      "loss": 0.5402,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7986463620981388,
+      "grad_norm": 3.3714424558205,
+      "learning_rate": 2.0512882167135047e-07,
+      "loss": 0.5022,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7990975747320925,
+      "grad_norm": 3.3008668639596284,
+      "learning_rate": 2.042426308657138e-07,
+      "loss": 0.5349,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7995487873660463,
+      "grad_norm": 3.5138662487604444,
+      "learning_rate": 2.0335814068095336e-07,
+      "loss": 0.5402,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9913939825555014,
+      "learning_rate": 2.0247535300732265e-07,
+      "loss": 0.5572,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8004512126339538,
+      "grad_norm": 3.6435559524273406,
+      "learning_rate": 2.0159426973143568e-07,
+      "loss": 0.4992,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8009024252679074,
+      "grad_norm": 3.2412230999711817,
+      "learning_rate": 2.0071489273626374e-07,
+      "loss": 0.6071,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8013536379018612,
+      "grad_norm": 2.969248009278257,
+      "learning_rate": 1.9983722390113255e-07,
+      "loss": 0.5,
+      "step": 1776
+    },
+    {
+      "epoch": 0.801804850535815,
+      "grad_norm": 2.768580600024177,
+      "learning_rate": 1.9896126510171641e-07,
+      "loss": 0.4454,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8022560631697687,
+      "grad_norm": 3.2476101536079787,
+      "learning_rate": 1.9808701821003614e-07,
+      "loss": 0.5911,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8027072758037225,
+      "grad_norm": 2.9616657048741244,
+      "learning_rate": 1.972144850944526e-07,
+      "loss": 0.5681,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8031584884376762,
+      "grad_norm": 3.105204828849402,
+      "learning_rate": 1.963436676196649e-07,
+      "loss": 0.5379,
+      "step": 1780
+    },
+    {
+      "epoch": 0.80360970107163,
+      "grad_norm": 3.292817844044641,
+      "learning_rate": 1.95474567646706e-07,
+      "loss": 0.5259,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8040609137055837,
+      "grad_norm": 3.264350635402717,
+      "learning_rate": 1.9460718703293765e-07,
+      "loss": 0.4252,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8045121263395375,
+      "grad_norm": 2.829404764906008,
+      "learning_rate": 1.9374152763204777e-07,
+      "loss": 0.5153,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8049633389734913,
+      "grad_norm": 3.2480177263767063,
+      "learning_rate": 1.9287759129404536e-07,
+      "loss": 0.4709,
+      "step": 1784
+    },
+    {
+      "epoch": 0.805414551607445,
+      "grad_norm": 2.899713242999487,
+      "learning_rate": 1.920153798652574e-07,
+      "loss": 0.4611,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8058657642413988,
+      "grad_norm": 3.1635502073692345,
+      "learning_rate": 1.9115489518832418e-07,
+      "loss": 0.4709,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8063169768753525,
+      "grad_norm": 3.1508557769862438,
+      "learning_rate": 1.9029613910219577e-07,
+      "loss": 0.5433,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8067681895093063,
+      "grad_norm": 3.5210820961966425,
+      "learning_rate": 1.8943911344212872e-07,
+      "loss": 0.6146,
+      "step": 1788
+    },
+    {
+      "epoch": 0.80721940214326,
+      "grad_norm": 3.3366881648206403,
+      "learning_rate": 1.8858382003968077e-07,
+      "loss": 0.5368,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8076706147772138,
+      "grad_norm": 3.3247727610257005,
+      "learning_rate": 1.8773026072270759e-07,
+      "loss": 0.8037,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8081218274111676,
+      "grad_norm": 3.4817252725861763,
+      "learning_rate": 1.8687843731535956e-07,
+      "loss": 0.5695,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8085730400451213,
+      "grad_norm": 3.336918219285363,
+      "learning_rate": 1.8602835163807662e-07,
+      "loss": 0.6866,
+      "step": 1792
+    },
+    {
+      "epoch": 0.809024252679075,
+      "grad_norm": 3.614272310931608,
+      "learning_rate": 1.8518000550758527e-07,
+      "loss": 0.4701,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8094754653130287,
+      "grad_norm": 3.219275535851942,
+      "learning_rate": 1.843334007368943e-07,
+      "loss": 0.6077,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8099266779469825,
+      "grad_norm": 2.89680741186188,
+      "learning_rate": 1.8348853913529083e-07,
+      "loss": 0.6083,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8103778905809362,
+      "grad_norm": 3.6574243799798385,
+      "learning_rate": 1.8264542250833748e-07,
+      "loss": 0.6222,
+      "step": 1796
+    },
+    {
+      "epoch": 0.81082910321489,
+      "grad_norm": 3.0840696058845793,
+      "learning_rate": 1.8180405265786657e-07,
+      "loss": 0.5185,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8112803158488437,
+      "grad_norm": 3.027696852321835,
+      "learning_rate": 1.8096443138197804e-07,
+      "loss": 0.5319,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8117315284827975,
+      "grad_norm": 3.159795651179558,
+      "learning_rate": 1.801265604750347e-07,
+      "loss": 0.5356,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8121827411167513,
+      "grad_norm": 3.3263427018468628,
+      "learning_rate": 1.792904417276584e-07,
+      "loss": 0.6881,
+      "step": 1800
+    },
+    {
+      "epoch": 0.812633953750705,
+      "grad_norm": 3.491430348543849,
+      "learning_rate": 1.7845607692672726e-07,
+      "loss": 0.6757,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8130851663846588,
+      "grad_norm": 3.452496098068734,
+      "learning_rate": 1.776234678553702e-07,
+      "loss": 0.634,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8135363790186125,
+      "grad_norm": 3.046194568097652,
+      "learning_rate": 1.7679261629296405e-07,
+      "loss": 0.5394,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8139875916525663,
+      "grad_norm": 3.41189405392855,
+      "learning_rate": 1.7596352401513025e-07,
+      "loss": 0.6579,
+      "step": 1804
+    },
+    {
+      "epoch": 0.81443880428652,
+      "grad_norm": 3.3370355365005433,
+      "learning_rate": 1.7513619279372982e-07,
+      "loss": 0.584,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8148900169204738,
+      "grad_norm": 3.1564896201684296,
+      "learning_rate": 1.743106243968605e-07,
+      "loss": 0.454,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8153412295544276,
+      "grad_norm": 3.4096665409080305,
+      "learning_rate": 1.7348682058885244e-07,
+      "loss": 0.5079,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8157924421883813,
+      "grad_norm": 3.422237532293,
+      "learning_rate": 1.7266478313026467e-07,
+      "loss": 0.49,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8162436548223351,
+      "grad_norm": 2.9182124049243043,
+      "learning_rate": 1.71844513777882e-07,
+      "loss": 0.5393,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8166948674562888,
+      "grad_norm": 3.298575760582701,
+      "learning_rate": 1.7102601428470986e-07,
+      "loss": 0.5132,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8171460800902425,
+      "grad_norm": 3.1076000971626323,
+      "learning_rate": 1.7020928639997133e-07,
+      "loss": 0.5096,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8175972927241962,
+      "grad_norm": 2.9576537154131914,
+      "learning_rate": 1.6939433186910435e-07,
+      "loss": 0.4751,
+      "step": 1812
+    },
+    {
+      "epoch": 0.81804850535815,
+      "grad_norm": 3.1438542828461884,
+      "learning_rate": 1.6858115243375516e-07,
+      "loss": 0.635,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8184997179921037,
+      "grad_norm": 3.3925384992387566,
+      "learning_rate": 1.6776974983177827e-07,
+      "loss": 0.5006,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8189509306260575,
+      "grad_norm": 3.155615848697049,
+      "learning_rate": 1.6696012579722986e-07,
+      "loss": 0.5519,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8194021432600113,
+      "grad_norm": 3.2055385260502516,
+      "learning_rate": 1.6615228206036524e-07,
+      "loss": 0.5571,
+      "step": 1816
+    },
+    {
+      "epoch": 0.819853355893965,
+      "grad_norm": 3.3036737895825485,
+      "learning_rate": 1.6534622034763556e-07,
+      "loss": 0.5955,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8203045685279188,
+      "grad_norm": 3.246191316789231,
+      "learning_rate": 1.6454194238168318e-07,
+      "loss": 0.4955,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8207557811618725,
+      "grad_norm": 3.7297330377785847,
+      "learning_rate": 1.6373944988133815e-07,
+      "loss": 0.6329,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8212069937958263,
+      "grad_norm": 3.3710927359486176,
+      "learning_rate": 1.6293874456161516e-07,
+      "loss": 0.6353,
+      "step": 1820
+    },
+    {
+      "epoch": 0.82165820642978,
+      "grad_norm": 3.075825293080748,
+      "learning_rate": 1.621398281337093e-07,
+      "loss": 0.5383,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8221094190637338,
+      "grad_norm": 3.0510999526952673,
+      "learning_rate": 1.6134270230499292e-07,
+      "loss": 0.4723,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8225606316976876,
+      "grad_norm": 2.990800015781615,
+      "learning_rate": 1.6054736877901154e-07,
+      "loss": 0.5556,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8230118443316413,
+      "grad_norm": 3.2691319322666086,
+      "learning_rate": 1.5975382925547965e-07,
+      "loss": 0.6269,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8234630569655951,
+      "grad_norm": 3.36974210150047,
+      "learning_rate": 1.5896208543027911e-07,
+      "loss": 0.6239,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8239142695995488,
+      "grad_norm": 3.2846656450026863,
+      "learning_rate": 1.5817213899545289e-07,
+      "loss": 0.5369,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8243654822335026,
+      "grad_norm": 3.6310256851443357,
+      "learning_rate": 1.5738399163920356e-07,
+      "loss": 0.5704,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8248166948674562,
+      "grad_norm": 2.8015522594684414,
+      "learning_rate": 1.5659764504588845e-07,
+      "loss": 0.5314,
+      "step": 1828
+    },
+    {
+      "epoch": 0.82526790750141,
+      "grad_norm": 2.9898520678955145,
+      "learning_rate": 1.558131008960163e-07,
+      "loss": 0.4779,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8257191201353637,
+      "grad_norm": 3.1623967382624345,
+      "learning_rate": 1.5503036086624454e-07,
+      "loss": 0.5098,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8261703327693175,
+      "grad_norm": 3.4072485811873543,
+      "learning_rate": 1.5424942662937434e-07,
+      "loss": 0.563,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8266215454032713,
+      "grad_norm": 2.855993326323611,
+      "learning_rate": 1.5347029985434777e-07,
+      "loss": 0.4537,
+      "step": 1832
+    },
+    {
+      "epoch": 0.827072758037225,
+      "grad_norm": 3.4851239971474195,
+      "learning_rate": 1.5269298220624505e-07,
+      "loss": 0.6242,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8275239706711788,
+      "grad_norm": 3.388860727365143,
+      "learning_rate": 1.5191747534627819e-07,
+      "loss": 0.5445,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8279751833051325,
+      "grad_norm": 3.554115255182286,
+      "learning_rate": 1.5114378093179147e-07,
+      "loss": 0.5375,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8284263959390863,
+      "grad_norm": 2.655545323566283,
+      "learning_rate": 1.5037190061625427e-07,
+      "loss": 0.3131,
+      "step": 1836
+    },
+    {
+      "epoch": 0.82887760857304,
+      "grad_norm": 3.3338895146881002,
+      "learning_rate": 1.4960183604925968e-07,
+      "loss": 0.5707,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8293288212069938,
+      "grad_norm": 3.218344870415116,
+      "learning_rate": 1.4883358887652042e-07,
+      "loss": 0.5006,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8297800338409476,
+      "grad_norm": 3.1468119802066203,
+      "learning_rate": 1.4806716073986504e-07,
+      "loss": 0.6662,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8302312464749013,
+      "grad_norm": 3.5160680404028493,
+      "learning_rate": 1.4730255327723452e-07,
+      "loss": 0.5002,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8306824591088551,
+      "grad_norm": 3.260575279604774,
+      "learning_rate": 1.4653976812267898e-07,
+      "loss": 0.5138,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8311336717428088,
+      "grad_norm": 2.9649200815505456,
+      "learning_rate": 1.457788069063538e-07,
+      "loss": 0.5762,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8315848843767626,
+      "grad_norm": 2.942782774050298,
+      "learning_rate": 1.4501967125451718e-07,
+      "loss": 0.5477,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8320360970107163,
+      "grad_norm": 3.0094337078574127,
+      "learning_rate": 1.442623627895251e-07,
+      "loss": 0.4498,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8324873096446701,
+      "grad_norm": 3.225704846247271,
+      "learning_rate": 1.4350688312982862e-07,
+      "loss": 0.6199,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8329385222786237,
+      "grad_norm": 3.1240951847264076,
+      "learning_rate": 1.4275323388997117e-07,
+      "loss": 0.4944,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8333897349125775,
+      "grad_norm": 3.5826594571402963,
+      "learning_rate": 1.4200141668058396e-07,
+      "loss": 0.6276,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8338409475465313,
+      "grad_norm": 3.3094708533679102,
+      "learning_rate": 1.412514331083826e-07,
+      "loss": 0.6394,
+      "step": 1848
+    },
+    {
+      "epoch": 0.834292160180485,
+      "grad_norm": 3.5135736501243446,
+      "learning_rate": 1.4050328477616458e-07,
+      "loss": 0.5399,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8347433728144388,
+      "grad_norm": 3.482173704832845,
+      "learning_rate": 1.3975697328280456e-07,
+      "loss": 0.4879,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8351945854483925,
+      "grad_norm": 3.0662306539692543,
+      "learning_rate": 1.3901250022325283e-07,
+      "loss": 0.4534,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8356457980823463,
+      "grad_norm": 3.0598921330176214,
+      "learning_rate": 1.382698671885295e-07,
+      "loss": 0.4628,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8360970107163,
+      "grad_norm": 3.1520019214759456,
+      "learning_rate": 1.3752907576572347e-07,
+      "loss": 0.6246,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8365482233502538,
+      "grad_norm": 2.6571093677090363,
+      "learning_rate": 1.3679012753798724e-07,
+      "loss": 0.4899,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8369994359842076,
+      "grad_norm": 2.9836867986668523,
+      "learning_rate": 1.3605302408453356e-07,
+      "loss": 0.4217,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8374506486181613,
+      "grad_norm": 2.898222008405317,
+      "learning_rate": 1.3531776698063436e-07,
+      "loss": 0.5292,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8379018612521151,
+      "grad_norm": 2.950876249789325,
+      "learning_rate": 1.3458435779761425e-07,
+      "loss": 0.4231,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8383530738860688,
+      "grad_norm": 3.174963854424885,
+      "learning_rate": 1.3385279810284956e-07,
+      "loss": 0.4779,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8388042865200226,
+      "grad_norm": 3.135298443318726,
+      "learning_rate": 1.3312308945976347e-07,
+      "loss": 0.5428,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8392554991539763,
+      "grad_norm": 3.5971103192719505,
+      "learning_rate": 1.3239523342782344e-07,
+      "loss": 0.5815,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8397067117879301,
+      "grad_norm": 3.413779126654346,
+      "learning_rate": 1.3166923156253817e-07,
+      "loss": 0.5965,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8401579244218839,
+      "grad_norm": 3.397541524879773,
+      "learning_rate": 1.309450854154528e-07,
+      "loss": 0.6834,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8406091370558376,
+      "grad_norm": 3.328000762597347,
+      "learning_rate": 1.3022279653414725e-07,
+      "loss": 0.64,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8410603496897913,
+      "grad_norm": 3.0084605090395353,
+      "learning_rate": 1.2950236646223244e-07,
+      "loss": 0.5704,
+      "step": 1864
+    },
+    {
+      "epoch": 0.841511562323745,
+      "grad_norm": 3.2459425568437053,
+      "learning_rate": 1.2878379673934615e-07,
+      "loss": 0.5634,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8419627749576988,
+      "grad_norm": 2.974415189859473,
+      "learning_rate": 1.2806708890115137e-07,
+      "loss": 0.4767,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8424139875916525,
+      "grad_norm": 3.3391852048885813,
+      "learning_rate": 1.2735224447933102e-07,
+      "loss": 0.528,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8428652002256063,
+      "grad_norm": 3.263226862766161,
+      "learning_rate": 1.2663926500158618e-07,
+      "loss": 0.6521,
+      "step": 1868
+    },
+    {
+      "epoch": 0.84331641285956,
+      "grad_norm": 3.208837226211113,
+      "learning_rate": 1.2592815199163244e-07,
+      "loss": 0.4281,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8437676254935138,
+      "grad_norm": 3.704953009121333,
+      "learning_rate": 1.2521890696919602e-07,
+      "loss": 0.6876,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8442188381274676,
+      "grad_norm": 3.1149574368441075,
+      "learning_rate": 1.245115314500118e-07,
+      "loss": 0.5546,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8446700507614213,
+      "grad_norm": 3.2673584564454083,
+      "learning_rate": 1.2380602694581888e-07,
+      "loss": 0.5626,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8451212633953751,
+      "grad_norm": 3.6047315203131025,
+      "learning_rate": 1.2310239496435748e-07,
+      "loss": 0.6006,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8455724760293288,
+      "grad_norm": 3.3754199818311004,
+      "learning_rate": 1.224006370093672e-07,
+      "loss": 0.5484,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8460236886632826,
+      "grad_norm": 3.299566444490645,
+      "learning_rate": 1.2170075458058083e-07,
+      "loss": 0.5683,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8464749012972363,
+      "grad_norm": 3.1617484748413154,
+      "learning_rate": 1.2100274917372477e-07,
+      "loss": 0.5741,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8469261139311901,
+      "grad_norm": 3.2176898145338533,
+      "learning_rate": 1.203066222805129e-07,
+      "loss": 0.5392,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8473773265651439,
+      "grad_norm": 2.6695851978983662,
+      "learning_rate": 1.1961237538864467e-07,
+      "loss": 0.586,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8478285391990976,
+      "grad_norm": 3.2205550742995057,
+      "learning_rate": 1.189200099818024e-07,
+      "loss": 0.5225,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8482797518330514,
+      "grad_norm": 3.1361790423595144,
+      "learning_rate": 1.1822952753964666e-07,
+      "loss": 0.4771,
+      "step": 1880
+    },
+    {
+      "epoch": 0.848730964467005,
+      "grad_norm": 2.9693045075564726,
+      "learning_rate": 1.1754092953781425e-07,
+      "loss": 0.582,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8491821771009588,
+      "grad_norm": 3.010501184967778,
+      "learning_rate": 1.168542174479148e-07,
+      "loss": 0.4967,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8496333897349125,
+      "grad_norm": 2.9444396444036416,
+      "learning_rate": 1.1616939273752713e-07,
+      "loss": 0.597,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8500846023688663,
+      "grad_norm": 3.1244893204773887,
+      "learning_rate": 1.1548645687019742e-07,
+      "loss": 0.476,
+      "step": 1884
+    },
+    {
+      "epoch": 0.85053581500282,
+      "grad_norm": 2.895489653963696,
+      "learning_rate": 1.1480541130543431e-07,
+      "loss": 0.4374,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8509870276367738,
+      "grad_norm": 3.6679522394726902,
+      "learning_rate": 1.1412625749870675e-07,
+      "loss": 0.6452,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8514382402707276,
+      "grad_norm": 3.4012095936653255,
+      "learning_rate": 1.1344899690144138e-07,
+      "loss": 0.6069,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8518894529046813,
+      "grad_norm": 3.4427078769517427,
+      "learning_rate": 1.1277363096101833e-07,
+      "loss": 0.4773,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8523406655386351,
+      "grad_norm": 2.8891427428108485,
+      "learning_rate": 1.1210016112076869e-07,
+      "loss": 0.5172,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8527918781725888,
+      "grad_norm": 3.1882213169736326,
+      "learning_rate": 1.1142858881997153e-07,
+      "loss": 0.6214,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8532430908065426,
+      "grad_norm": 3.2631974785916578,
+      "learning_rate": 1.107589154938503e-07,
+      "loss": 0.3816,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8536943034404963,
+      "grad_norm": 3.646300547484839,
+      "learning_rate": 1.10091142573571e-07,
+      "loss": 0.5078,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8541455160744501,
+      "grad_norm": 3.0800229323313446,
+      "learning_rate": 1.0942527148623736e-07,
+      "loss": 0.5741,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8545967287084039,
+      "grad_norm": 3.4437111992470695,
+      "learning_rate": 1.0876130365488878e-07,
+      "loss": 0.4977,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8550479413423576,
+      "grad_norm": 3.415687476852208,
+      "learning_rate": 1.0809924049849816e-07,
+      "loss": 0.5176,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8554991539763114,
+      "grad_norm": 3.423266820929565,
+      "learning_rate": 1.0743908343196629e-07,
+      "loss": 0.5477,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8559503666102651,
+      "grad_norm": 3.319978860721829,
+      "learning_rate": 1.067808338661219e-07,
+      "loss": 0.6,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8564015792442189,
+      "grad_norm": 3.2937834521614766,
+      "learning_rate": 1.0612449320771644e-07,
+      "loss": 0.6547,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8568527918781725,
+      "grad_norm": 2.7650015075242136,
+      "learning_rate": 1.0547006285942162e-07,
+      "loss": 0.4555,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8573040045121263,
+      "grad_norm": 3.2864394879320495,
+      "learning_rate": 1.0481754421982758e-07,
+      "loss": 0.5315,
+      "step": 1900
+    },
+    {
+      "epoch": 0.85775521714608,
+      "grad_norm": 3.359413365564091,
+      "learning_rate": 1.0416693868343795e-07,
+      "loss": 0.4423,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8582064297800338,
+      "grad_norm": 2.9927056449757714,
+      "learning_rate": 1.0351824764066819e-07,
+      "loss": 0.5731,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8586576424139876,
+      "grad_norm": 3.61532226087724,
+      "learning_rate": 1.0287147247784244e-07,
+      "loss": 0.6,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8591088550479413,
+      "grad_norm": 2.9636046590677902,
+      "learning_rate": 1.0222661457718985e-07,
+      "loss": 0.683,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8595600676818951,
+      "grad_norm": 3.2418007105981337,
+      "learning_rate": 1.015836753168431e-07,
+      "loss": 0.5704,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8600112803158488,
+      "grad_norm": 3.270787762505259,
+      "learning_rate": 1.0094265607083374e-07,
+      "loss": 0.5531,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8604624929498026,
+      "grad_norm": 3.4241238059817753,
+      "learning_rate": 1.0030355820908997e-07,
+      "loss": 0.4724,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8609137055837564,
+      "grad_norm": 3.120442346322047,
+      "learning_rate": 9.966638309743481e-08,
+      "loss": 0.5969,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8613649182177101,
+      "grad_norm": 2.8700401112362983,
+      "learning_rate": 9.903113209758096e-08,
+      "loss": 0.639,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8618161308516639,
+      "grad_norm": 2.90952553707339,
+      "learning_rate": 9.839780656712959e-08,
+      "loss": 0.4335,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8622673434856176,
+      "grad_norm": 3.15352399159264,
+      "learning_rate": 9.776640785956702e-08,
+      "loss": 0.602,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8627185561195714,
+      "grad_norm": 3.558953242347675,
+      "learning_rate": 9.713693732426131e-08,
+      "loss": 0.5247,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8631697687535251,
+      "grad_norm": 3.058394803478645,
+      "learning_rate": 9.65093963064606e-08,
+      "loss": 0.5442,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8636209813874789,
+      "grad_norm": 3.0343736318962646,
+      "learning_rate": 9.588378614728864e-08,
+      "loss": 0.4304,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8640721940214326,
+      "grad_norm": 3.103162455870594,
+      "learning_rate": 9.526010818374309e-08,
+      "loss": 0.6602,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8645234066553864,
+      "grad_norm": 3.345412268134743,
+      "learning_rate": 9.46383637486925e-08,
+      "loss": 0.5215,
+      "step": 1916
+    },
+    {
+      "epoch": 0.86497461928934,
+      "grad_norm": 2.950588770881482,
+      "learning_rate": 9.401855417087234e-08,
+      "loss": 0.432,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8654258319232938,
+      "grad_norm": 2.9028721420815438,
+      "learning_rate": 9.34006807748845e-08,
+      "loss": 0.5536,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8658770445572476,
+      "grad_norm": 3.422382660689821,
+      "learning_rate": 9.278474488119182e-08,
+      "loss": 0.5321,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8663282571912013,
+      "grad_norm": 3.052528705427497,
+      "learning_rate": 9.217074780611688e-08,
+      "loss": 0.5474,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8667794698251551,
+      "grad_norm": 3.1194719596998484,
+      "learning_rate": 9.155869086183921e-08,
+      "loss": 0.5148,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8672306824591088,
+      "grad_norm": 2.9070890387145494,
+      "learning_rate": 9.094857535639156e-08,
+      "loss": 0.4189,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8676818950930626,
+      "grad_norm": 3.177762649776062,
+      "learning_rate": 9.03404025936576e-08,
+      "loss": 0.4111,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8681331077270164,
+      "grad_norm": 3.4014028718324885,
+      "learning_rate": 8.973417387336946e-08,
+      "loss": 0.5615,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8685843203609701,
+      "grad_norm": 3.314370098733984,
+      "learning_rate": 8.91298904911043e-08,
+      "loss": 0.5898,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8690355329949239,
+      "grad_norm": 3.2243090136724772,
+      "learning_rate": 8.852755373828235e-08,
+      "loss": 0.5159,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8694867456288776,
+      "grad_norm": 3.1990156923608506,
+      "learning_rate": 8.792716490216335e-08,
+      "loss": 0.5222,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8699379582628314,
+      "grad_norm": 3.0312256180294996,
+      "learning_rate": 8.732872526584379e-08,
+      "loss": 0.7078,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8703891708967851,
+      "grad_norm": 3.0987202755918983,
+      "learning_rate": 8.67322361082553e-08,
+      "loss": 0.4882,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8708403835307389,
+      "grad_norm": 2.7196240122489446,
+      "learning_rate": 8.613769870416066e-08,
+      "loss": 0.5035,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8712915961646926,
+      "grad_norm": 2.9325956476707224,
+      "learning_rate": 8.554511432415145e-08,
+      "loss": 0.4407,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8717428087986464,
+      "grad_norm": 3.563913043226757,
+      "learning_rate": 8.495448423464568e-08,
+      "loss": 0.6164,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8721940214326002,
+      "grad_norm": 3.0026341334651323,
+      "learning_rate": 8.436580969788431e-08,
+      "loss": 0.4445,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8726452340665538,
+      "grad_norm": 3.2398204365530656,
+      "learning_rate": 8.377909197193011e-08,
+      "loss": 0.5261,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8730964467005076,
+      "grad_norm": 3.552891767506158,
+      "learning_rate": 8.319433231066264e-08,
+      "loss": 0.683,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8735476593344613,
+      "grad_norm": 2.857576906676008,
+      "learning_rate": 8.261153196377813e-08,
+      "loss": 0.4816,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8739988719684151,
+      "grad_norm": 3.172569622874956,
+      "learning_rate": 8.20306921767847e-08,
+      "loss": 0.4672,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8744500846023688,
+      "grad_norm": 3.515851107453122,
+      "learning_rate": 8.145181419100034e-08,
+      "loss": 0.6234,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8749012972363226,
+      "grad_norm": 3.439159617356599,
+      "learning_rate": 8.08748992435514e-08,
+      "loss": 0.5984,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8753525098702764,
+      "grad_norm": 3.3171348098631857,
+      "learning_rate": 8.02999485673681e-08,
+      "loss": 0.5059,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8758037225042301,
+      "grad_norm": 3.008859919440221,
+      "learning_rate": 7.972696339118346e-08,
+      "loss": 0.5124,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8762549351381839,
+      "grad_norm": 3.3407440096371626,
+      "learning_rate": 7.91559449395296e-08,
+      "loss": 0.5261,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8767061477721376,
+      "grad_norm": 3.2680192388068985,
+      "learning_rate": 7.858689443273547e-08,
+      "loss": 0.4824,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8771573604060914,
+      "grad_norm": 3.2302885233475247,
+      "learning_rate": 7.801981308692507e-08,
+      "loss": 0.5301,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8776085730400451,
+      "grad_norm": 3.3144955548693846,
+      "learning_rate": 7.745470211401273e-08,
+      "loss": 0.5292,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8780597856739989,
+      "grad_norm": 3.6338069876593972,
+      "learning_rate": 7.689156272170316e-08,
+      "loss": 0.6645,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8785109983079527,
+      "grad_norm": 3.320705078381082,
+      "learning_rate": 7.633039611348701e-08,
+      "loss": 0.5731,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8789622109419064,
+      "grad_norm": 3.400427088861012,
+      "learning_rate": 7.577120348863864e-08,
+      "loss": 0.5424,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8794134235758602,
+      "grad_norm": 2.952134565078521,
+      "learning_rate": 7.521398604221451e-08,
+      "loss": 0.4211,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8798646362098139,
+      "grad_norm": 3.1223077524770804,
+      "learning_rate": 7.465874496504943e-08,
+      "loss": 0.4525,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8803158488437677,
+      "grad_norm": 3.3651190746871213,
+      "learning_rate": 7.410548144375417e-08,
+      "loss": 0.5718,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8807670614777213,
+      "grad_norm": 2.8278868757405995,
+      "learning_rate": 7.355419666071406e-08,
+      "loss": 0.5105,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8812182741116751,
+      "grad_norm": 3.3691970534959346,
+      "learning_rate": 7.300489179408476e-08,
+      "loss": 0.4373,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8816694867456288,
+      "grad_norm": 3.1396491136987374,
+      "learning_rate": 7.245756801779158e-08,
+      "loss": 0.5241,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8821206993795826,
+      "grad_norm": 2.9990018272726338,
+      "learning_rate": 7.191222650152528e-08,
+      "loss": 0.5162,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8825719120135364,
+      "grad_norm": 2.872850179093246,
+      "learning_rate": 7.136886841074052e-08,
+      "loss": 0.4716,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8830231246474901,
+      "grad_norm": 3.7713414016675255,
+      "learning_rate": 7.082749490665351e-08,
+      "loss": 0.6061,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8834743372814439,
+      "grad_norm": 3.476662997099093,
+      "learning_rate": 7.028810714623846e-08,
+      "loss": 0.4683,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8839255499153976,
+      "grad_norm": 3.2389720481182764,
+      "learning_rate": 6.975070628222646e-08,
+      "loss": 0.5448,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8843767625493514,
+      "grad_norm": 3.1694385457983127,
+      "learning_rate": 6.921529346310218e-08,
+      "loss": 0.5136,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8848279751833051,
+      "grad_norm": 3.6314284620989787,
+      "learning_rate": 6.868186983310131e-08,
+      "loss": 0.4929,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8852791878172589,
+      "grad_norm": 3.123575388660666,
+      "learning_rate": 6.81504365322092e-08,
+      "loss": 0.5635,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8857304004512127,
+      "grad_norm": 2.6321261997764807,
+      "learning_rate": 6.76209946961569e-08,
+      "loss": 0.4772,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8861816130851664,
+      "grad_norm": 3.210961705882933,
+      "learning_rate": 6.709354545641987e-08,
+      "loss": 0.5765,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8866328257191202,
+      "grad_norm": 3.102505340660919,
+      "learning_rate": 6.65680899402149e-08,
+      "loss": 0.6185,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8870840383530739,
+      "grad_norm": 3.318813163094756,
+      "learning_rate": 6.604462927049804e-08,
+      "loss": 0.5496,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8875352509870277,
+      "grad_norm": 3.693495485179525,
+      "learning_rate": 6.552316456596252e-08,
+      "loss": 0.6225,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8879864636209814,
+      "grad_norm": 2.7909920602545957,
+      "learning_rate": 6.500369694103558e-08,
+      "loss": 0.4657,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8884376762549352,
+      "grad_norm": 3.8268919463461315,
+      "learning_rate": 6.44862275058763e-08,
+      "loss": 0.5475,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 2.8955466658153273,
+      "learning_rate": 6.397075736637403e-08,
+      "loss": 0.4761,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8893401015228426,
+      "grad_norm": 3.0393342411424054,
+      "learning_rate": 6.345728762414503e-08,
+      "loss": 0.5003,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8897913141567964,
+      "grad_norm": 3.0480382524304677,
+      "learning_rate": 6.294581937653042e-08,
+      "loss": 0.5091,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8902425267907501,
+      "grad_norm": 3.2472044245989027,
+      "learning_rate": 6.243635371659395e-08,
+      "loss": 0.5822,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8906937394247039,
+      "grad_norm": 2.9373896773076926,
+      "learning_rate": 6.192889173311966e-08,
+      "loss": 0.5317,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8911449520586576,
+      "grad_norm": 3.4088834880295034,
+      "learning_rate": 6.142343451060972e-08,
+      "loss": 0.5696,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8915961646926114,
+      "grad_norm": 3.2061069267520645,
+      "learning_rate": 6.091998312928171e-08,
+      "loss": 0.5439,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8920473773265651,
+      "grad_norm": 2.9538556019221733,
+      "learning_rate": 6.04185386650662e-08,
+      "loss": 0.4199,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8924985899605189,
+      "grad_norm": 3.3347490781585125,
+      "learning_rate": 5.99191021896055e-08,
+      "loss": 0.585,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8929498025944727,
+      "grad_norm": 3.252069083231302,
+      "learning_rate": 5.9421674770249844e-08,
+      "loss": 0.5721,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8934010152284264,
+      "grad_norm": 3.373090912932044,
+      "learning_rate": 5.8926257470056415e-08,
+      "loss": 0.6677,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8938522278623802,
+      "grad_norm": 3.470419272509584,
+      "learning_rate": 5.8432851347786414e-08,
+      "loss": 0.6336,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8943034404963339,
+      "grad_norm": 2.954923563225483,
+      "learning_rate": 5.794145745790269e-08,
+      "loss": 0.5161,
+      "step": 1982
+    },
+    {
+      "epoch": 0.8947546531302877,
+      "grad_norm": 2.9240499120629373,
+      "learning_rate": 5.7452076850568186e-08,
+      "loss": 0.5764,
+      "step": 1983
+    },
+    {
+      "epoch": 0.8952058657642414,
+      "grad_norm": 3.163381313365714,
+      "learning_rate": 5.696471057164298e-08,
+      "loss": 0.4723,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8956570783981952,
+      "grad_norm": 3.001631089880879,
+      "learning_rate": 5.6479359662682246e-08,
+      "loss": 0.4736,
+      "step": 1985
+    },
+    {
+      "epoch": 0.896108291032149,
+      "grad_norm": 3.1077853978436414,
+      "learning_rate": 5.599602516093427e-08,
+      "loss": 0.4647,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8965595036661026,
+      "grad_norm": 2.9246040215255658,
+      "learning_rate": 5.551470809933756e-08,
+      "loss": 0.4681,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8970107163000564,
+      "grad_norm": 3.4387470409961223,
+      "learning_rate": 5.503540950652008e-08,
+      "loss": 0.507,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8974619289340101,
+      "grad_norm": 2.6230689040197364,
+      "learning_rate": 5.4558130406795355e-08,
+      "loss": 0.4589,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8979131415679639,
+      "grad_norm": 3.388098608880633,
+      "learning_rate": 5.408287182016091e-08,
+      "loss": 0.4525,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8983643542019176,
+      "grad_norm": 3.119545140896645,
+      "learning_rate": 5.360963476229707e-08,
+      "loss": 0.5486,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8988155668358714,
+      "grad_norm": 3.1519502919215694,
+      "learning_rate": 5.313842024456305e-08,
+      "loss": 0.4839,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8992667794698251,
+      "grad_norm": 3.7651904706666395,
+      "learning_rate": 5.2669229273996084e-08,
+      "loss": 0.6181,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8997179921037789,
+      "grad_norm": 3.058263724338039,
+      "learning_rate": 5.220206285330886e-08,
+      "loss": 0.5597,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9001692047377327,
+      "grad_norm": 3.0934748276186337,
+      "learning_rate": 5.173692198088708e-08,
+      "loss": 0.4815,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9006204173716864,
+      "grad_norm": 3.968781722790635,
+      "learning_rate": 5.1273807650788146e-08,
+      "loss": 0.5632,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9010716300056402,
+      "grad_norm": 3.131751903358897,
+      "learning_rate": 5.081272085273825e-08,
+      "loss": 0.4128,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9015228426395939,
+      "grad_norm": 2.9193006683184275,
+      "learning_rate": 5.035366257213014e-08,
+      "loss": 0.4801,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9019740552735477,
+      "grad_norm": 3.2693871968854875,
+      "learning_rate": 4.98966337900224e-08,
+      "loss": 0.5536,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9024252679075014,
+      "grad_norm": 3.2813024613985227,
+      "learning_rate": 4.944163548313496e-08,
+      "loss": 0.6593,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9028764805414552,
+      "grad_norm": 3.369192254537855,
+      "learning_rate": 4.898866862384976e-08,
+      "loss": 0.4618,
+      "step": 2001
+    },
+    {
+      "epoch": 0.903327693175409,
+      "grad_norm": 3.043012997543813,
+      "learning_rate": 4.853773418020646e-08,
+      "loss": 0.4676,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9037789058093627,
+      "grad_norm": 3.434166016020396,
+      "learning_rate": 4.8088833115901395e-08,
+      "loss": 0.3948,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9042301184433165,
+      "grad_norm": 2.7646887236809854,
+      "learning_rate": 4.764196639028572e-08,
+      "loss": 0.4774,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9046813310772701,
+      "grad_norm": 3.3361977687273927,
+      "learning_rate": 4.719713495836242e-08,
+      "loss": 0.5685,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9051325437112239,
+      "grad_norm": 3.4831449664643257,
+      "learning_rate": 4.6754339770785465e-08,
+      "loss": 0.5089,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9055837563451776,
+      "grad_norm": 3.4671397319854522,
+      "learning_rate": 4.631358177385647e-08,
+      "loss": 0.5439,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9060349689791314,
+      "grad_norm": 3.2933483172414335,
+      "learning_rate": 4.58748619095235e-08,
+      "loss": 0.5683,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9064861816130851,
+      "grad_norm": 2.985093722715634,
+      "learning_rate": 4.543818111537956e-08,
+      "loss": 0.5344,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9069373942470389,
+      "grad_norm": 3.4328919623699674,
+      "learning_rate": 4.500354032465925e-08,
+      "loss": 0.4476,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9073886068809927,
+      "grad_norm": 3.304919883947093,
+      "learning_rate": 4.457094046623755e-08,
+      "loss": 0.6034,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9078398195149464,
+      "grad_norm": 2.8099205651955206,
+      "learning_rate": 4.414038246462803e-08,
+      "loss": 0.4899,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9082910321489002,
+      "grad_norm": 3.3739143855436535,
+      "learning_rate": 4.3711867239980324e-08,
+      "loss": 0.6359,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9087422447828539,
+      "grad_norm": 3.189618086168555,
+      "learning_rate": 4.3285395708078546e-08,
+      "loss": 0.5732,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9091934574168077,
+      "grad_norm": 3.367461634350275,
+      "learning_rate": 4.286096878033929e-08,
+      "loss": 0.4531,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9096446700507614,
+      "grad_norm": 3.6483005516482767,
+      "learning_rate": 4.243858736380912e-08,
+      "loss": 0.6179,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9100958826847152,
+      "grad_norm": 3.415485122808768,
+      "learning_rate": 4.2018252361164076e-08,
+      "loss": 0.5615,
+      "step": 2017
+    },
+    {
+      "epoch": 0.910547095318669,
+      "grad_norm": 2.741772493137757,
+      "learning_rate": 4.15999646707057e-08,
+      "loss": 0.4732,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9109983079526227,
+      "grad_norm": 3.0538104199950307,
+      "learning_rate": 4.118372518636104e-08,
+      "loss": 0.5523,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9114495205865765,
+      "grad_norm": 3.0764370937470704,
+      "learning_rate": 4.076953479767964e-08,
+      "loss": 0.5979,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9119007332205302,
+      "grad_norm": 3.066765697208492,
+      "learning_rate": 4.035739438983143e-08,
+      "loss": 0.4439,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9123519458544839,
+      "grad_norm": 3.668533201777122,
+      "learning_rate": 3.994730484360609e-08,
+      "loss": 0.572,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9128031584884376,
+      "grad_norm": 3.401968524293534,
+      "learning_rate": 3.953926703540977e-08,
+      "loss": 0.5584,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9132543711223914,
+      "grad_norm": 3.4337981912223174,
+      "learning_rate": 3.9133281837264385e-08,
+      "loss": 0.513,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9137055837563451,
+      "grad_norm": 4.049276576839161,
+      "learning_rate": 3.872935011680456e-08,
+      "loss": 0.5475,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9141567963902989,
+      "grad_norm": 3.227324543340619,
+      "learning_rate": 3.832747273727699e-08,
+      "loss": 0.4974,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9146080090242527,
+      "grad_norm": 3.700152680818643,
+      "learning_rate": 3.792765055753755e-08,
+      "loss": 0.6376,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9150592216582064,
+      "grad_norm": 3.134825348605703,
+      "learning_rate": 3.7529884432050074e-08,
+      "loss": 0.5864,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9155104342921602,
+      "grad_norm": 3.287365982894785,
+      "learning_rate": 3.71341752108848e-08,
+      "loss": 0.5592,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9159616469261139,
+      "grad_norm": 3.3656815978558408,
+      "learning_rate": 3.674052373971559e-08,
+      "loss": 0.5296,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9164128595600677,
+      "grad_norm": 3.08635532381097,
+      "learning_rate": 3.634893085981872e-08,
+      "loss": 0.4835,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9168640721940214,
+      "grad_norm": 3.120263533445725,
+      "learning_rate": 3.595939740807141e-08,
+      "loss": 0.5167,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9173152848279752,
+      "grad_norm": 2.9603539149063374,
+      "learning_rate": 3.557192421694932e-08,
+      "loss": 0.5019,
+      "step": 2033
+    },
+    {
+      "epoch": 0.917766497461929,
+      "grad_norm": 3.300594188794194,
+      "learning_rate": 3.518651211452528e-08,
+      "loss": 0.6458,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9182177100958827,
+      "grad_norm": 3.122902491830948,
+      "learning_rate": 3.4803161924467196e-08,
+      "loss": 0.5104,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9186689227298365,
+      "grad_norm": 3.4528349816512103,
+      "learning_rate": 3.4421874466036285e-08,
+      "loss": 0.5535,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9191201353637902,
+      "grad_norm": 3.3890580466360913,
+      "learning_rate": 3.404265055408617e-08,
+      "loss": 0.4116,
+      "step": 2037
+    },
+    {
+      "epoch": 0.919571347997744,
+      "grad_norm": 3.2546448072566645,
+      "learning_rate": 3.36654909990598e-08,
+      "loss": 0.5527,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9200225606316977,
+      "grad_norm": 2.6288215794522314,
+      "learning_rate": 3.3290396606988405e-08,
+      "loss": 0.4729,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9204737732656514,
+      "grad_norm": 3.0588000528716344,
+      "learning_rate": 3.29173681794902e-08,
+      "loss": 0.5862,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9209249858996051,
+      "grad_norm": 3.3085745174099244,
+      "learning_rate": 3.25464065137675e-08,
+      "loss": 0.4663,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9213761985335589,
+      "grad_norm": 3.3049910550091286,
+      "learning_rate": 3.217751240260647e-08,
+      "loss": 0.5277,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9218274111675127,
+      "grad_norm": 3.1470391275241405,
+      "learning_rate": 3.1810686634374253e-08,
+      "loss": 0.5739,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9222786238014664,
+      "grad_norm": 2.8926904316169897,
+      "learning_rate": 3.144592999301754e-08,
+      "loss": 0.5079,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9227298364354202,
+      "grad_norm": 3.3359025166747753,
+      "learning_rate": 3.1083243258061666e-08,
+      "loss": 0.6565,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9231810490693739,
+      "grad_norm": 3.2267530813326193,
+      "learning_rate": 3.072262720460783e-08,
+      "loss": 0.511,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9236322617033277,
+      "grad_norm": 3.172137007655229,
+      "learning_rate": 3.036408260333223e-08,
+      "loss": 0.5653,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9240834743372814,
+      "grad_norm": 2.9652257724423183,
+      "learning_rate": 3.000761022048393e-08,
+      "loss": 0.5494,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9245346869712352,
+      "grad_norm": 3.162831602182436,
+      "learning_rate": 2.9653210817883634e-08,
+      "loss": 0.5811,
+      "step": 2049
+    },
+    {
+      "epoch": 0.924985899605189,
+      "grad_norm": 3.3548734459357195,
+      "learning_rate": 2.930088515292173e-08,
+      "loss": 0.4703,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9254371122391427,
+      "grad_norm": 2.971253690703245,
+      "learning_rate": 2.8950633978556906e-08,
+      "loss": 0.6375,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9258883248730965,
+      "grad_norm": 3.282908277488212,
+      "learning_rate": 2.860245804331429e-08,
+      "loss": 0.5976,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9263395375070502,
+      "grad_norm": 3.4398673398293833,
+      "learning_rate": 2.8256358091284238e-08,
+      "loss": 0.5767,
+      "step": 2053
+    },
+    {
+      "epoch": 0.926790750141004,
+      "grad_norm": 3.0750885972506654,
+      "learning_rate": 2.79123348621203e-08,
+      "loss": 0.492,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9272419627749577,
+      "grad_norm": 3.295821068137443,
+      "learning_rate": 2.7570389091037926e-08,
+      "loss": 0.5278,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9276931754089115,
+      "grad_norm": 3.54320248311843,
+      "learning_rate": 2.7230521508812553e-08,
+      "loss": 0.5603,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9281443880428653,
+      "grad_norm": 3.6284081250605063,
+      "learning_rate": 2.689273284177873e-08,
+      "loss": 0.6389,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9285956006768189,
+      "grad_norm": 3.136805769994715,
+      "learning_rate": 2.6557023811827894e-08,
+      "loss": 0.5818,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9290468133107727,
+      "grad_norm": 2.943590397644658,
+      "learning_rate": 2.6223395136407145e-08,
+      "loss": 0.5318,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9294980259447264,
+      "grad_norm": 3.126539259551608,
+      "learning_rate": 2.5891847528517476e-08,
+      "loss": 0.4908,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9299492385786802,
+      "grad_norm": 3.2669493617048198,
+      "learning_rate": 2.5562381696712654e-08,
+      "loss": 0.5458,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9304004512126339,
+      "grad_norm": 3.0754859463953803,
+      "learning_rate": 2.5234998345097237e-08,
+      "loss": 0.4965,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9308516638465877,
+      "grad_norm": 3.6839611131375336,
+      "learning_rate": 2.4909698173325443e-08,
+      "loss": 0.5531,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9313028764805414,
+      "grad_norm": 3.22462453822676,
+      "learning_rate": 2.458648187659962e-08,
+      "loss": 0.4566,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9317540891144952,
+      "grad_norm": 3.4060647980598,
+      "learning_rate": 2.4265350145668106e-08,
+      "loss": 0.7031,
+      "step": 2065
+    },
+    {
+      "epoch": 0.932205301748449,
+      "grad_norm": 3.089929652023956,
+      "learning_rate": 2.394630366682493e-08,
+      "loss": 0.4991,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9326565143824027,
+      "grad_norm": 3.0972124530500427,
+      "learning_rate": 2.3629343121907562e-08,
+      "loss": 0.5814,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9331077270163565,
+      "grad_norm": 2.956122127539399,
+      "learning_rate": 2.3314469188295272e-08,
+      "loss": 0.447,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9335589396503102,
+      "grad_norm": 3.4655230691165775,
+      "learning_rate": 2.300168253890833e-08,
+      "loss": 0.5864,
+      "step": 2069
+    },
+    {
+      "epoch": 0.934010152284264,
+      "grad_norm": 3.2759006855749764,
+      "learning_rate": 2.2690983842205914e-08,
+      "loss": 0.5745,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9344613649182177,
+      "grad_norm": 3.0615174312777014,
+      "learning_rate": 2.2382373762185658e-08,
+      "loss": 0.4796,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9349125775521715,
+      "grad_norm": 3.432283728728365,
+      "learning_rate": 2.207585295838099e-08,
+      "loss": 0.6578,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9353637901861253,
+      "grad_norm": 3.3123138166785275,
+      "learning_rate": 2.177142208586047e-08,
+      "loss": 0.5887,
+      "step": 2073
+    },
+    {
+      "epoch": 0.935815002820079,
+      "grad_norm": 3.3385012625964143,
+      "learning_rate": 2.146908179522644e-08,
+      "loss": 0.5244,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9362662154540327,
+      "grad_norm": 3.412059349668194,
+      "learning_rate": 2.116883273261316e-08,
+      "loss": 0.4841,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9367174280879864,
+      "grad_norm": 3.0584437698376314,
+      "learning_rate": 2.087067553968602e-08,
+      "loss": 0.3759,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9371686407219402,
+      "grad_norm": 3.2188511717745123,
+      "learning_rate": 2.057461085363954e-08,
+      "loss": 0.5688,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9376198533558939,
+      "grad_norm": 2.9976242270120594,
+      "learning_rate": 2.028063930719637e-08,
+      "loss": 0.4009,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9380710659898477,
+      "grad_norm": 2.892808153960682,
+      "learning_rate": 1.9988761528606178e-08,
+      "loss": 0.4392,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9385222786238014,
+      "grad_norm": 3.2358320547434443,
+      "learning_rate": 1.9698978141643784e-08,
+      "loss": 0.503,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9389734912577552,
+      "grad_norm": 3.3962163022150653,
+      "learning_rate": 1.941128976560791e-08,
+      "loss": 0.5694,
+      "step": 2081
+    },
+    {
+      "epoch": 0.939424703891709,
+      "grad_norm": 2.998665449003642,
+      "learning_rate": 1.912569701532063e-08,
+      "loss": 0.5902,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9398759165256627,
+      "grad_norm": 3.3801760197505835,
+      "learning_rate": 1.8842200501124615e-08,
+      "loss": 0.5383,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9403271291596165,
+      "grad_norm": 2.9348713239408593,
+      "learning_rate": 1.8560800828883227e-08,
+      "loss": 0.4877,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9407783417935702,
+      "grad_norm": 3.097739950065729,
+      "learning_rate": 1.8281498599978407e-08,
+      "loss": 0.6081,
+      "step": 2085
+    },
+    {
+      "epoch": 0.941229554427524,
+      "grad_norm": 3.2982172408128814,
+      "learning_rate": 1.8004294411309685e-08,
+      "loss": 0.5686,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9416807670614777,
+      "grad_norm": 2.9622880055723084,
+      "learning_rate": 1.7729188855292954e-08,
+      "loss": 0.578,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9421319796954315,
+      "grad_norm": 2.947330771863818,
+      "learning_rate": 1.7456182519858808e-08,
+      "loss": 0.5588,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9425831923293853,
+      "grad_norm": 2.937817283751574,
+      "learning_rate": 1.7185275988451986e-08,
+      "loss": 0.4514,
+      "step": 2089
+    },
+    {
+      "epoch": 0.943034404963339,
+      "grad_norm": 3.3308032462704644,
+      "learning_rate": 1.691646984002937e-08,
+      "loss": 0.5546,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9434856175972928,
+      "grad_norm": 2.8609200678838875,
+      "learning_rate": 1.6649764649059094e-08,
+      "loss": 0.4772,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9439368302312465,
+      "grad_norm": 3.3257921371498256,
+      "learning_rate": 1.6385160985519564e-08,
+      "loss": 0.4277,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9443880428652002,
+      "grad_norm": 3.280225796167774,
+      "learning_rate": 1.6122659414897876e-08,
+      "loss": 0.6392,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9448392554991539,
+      "grad_norm": 3.8085560156358054,
+      "learning_rate": 1.5862260498188728e-08,
+      "loss": 0.7179,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9452904681331077,
+      "grad_norm": 3.3335448343938268,
+      "learning_rate": 1.56039647918933e-08,
+      "loss": 0.6197,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9457416807670614,
+      "grad_norm": 3.031288983603637,
+      "learning_rate": 1.5347772848017583e-08,
+      "loss": 0.4715,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9461928934010152,
+      "grad_norm": 3.164918639328314,
+      "learning_rate": 1.509368521407217e-08,
+      "loss": 0.5569,
+      "step": 2097
+    },
+    {
+      "epoch": 0.946644106034969,
+      "grad_norm": 3.210287170541672,
+      "learning_rate": 1.4841702433070037e-08,
+      "loss": 0.6519,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9470953186689227,
+      "grad_norm": 2.6369570547287977,
+      "learning_rate": 1.4591825043526073e-08,
+      "loss": 0.4529,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9475465313028765,
+      "grad_norm": 3.391418189703669,
+      "learning_rate": 1.4344053579455894e-08,
+      "loss": 0.4644,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9479977439368302,
+      "grad_norm": 2.9940931472415793,
+      "learning_rate": 1.4098388570374154e-08,
+      "loss": 0.4532,
+      "step": 2101
+    },
+    {
+      "epoch": 0.948448956570784,
+      "grad_norm": 3.1943317280291077,
+      "learning_rate": 1.3854830541294105e-08,
+      "loss": 0.5737,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9489001692047377,
+      "grad_norm": 3.463904492669196,
+      "learning_rate": 1.3613380012725717e-08,
+      "loss": 0.5669,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9493513818386915,
+      "grad_norm": 2.950631481291773,
+      "learning_rate": 1.337403750067545e-08,
+      "loss": 0.5732,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9498025944726453,
+      "grad_norm": 3.2538975233288303,
+      "learning_rate": 1.3136803516644701e-08,
+      "loss": 0.5587,
+      "step": 2105
+    },
+    {
+      "epoch": 0.950253807106599,
+      "grad_norm": 3.3928844420275692,
+      "learning_rate": 1.2901678567628249e-08,
+      "loss": 0.5758,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9507050197405528,
+      "grad_norm": 3.402553144487364,
+      "learning_rate": 1.2668663156114035e-08,
+      "loss": 0.5521,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9511562323745065,
+      "grad_norm": 3.1090460601388688,
+      "learning_rate": 1.2437757780081715e-08,
+      "loss": 0.4712,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9516074450084603,
+      "grad_norm": 3.157111834878996,
+      "learning_rate": 1.2208962933001332e-08,
+      "loss": 0.7341,
+      "step": 2109
+    },
+    {
+      "epoch": 0.952058657642414,
+      "grad_norm": 3.6264603472952692,
+      "learning_rate": 1.1982279103832539e-08,
+      "loss": 0.6544,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9525098702763677,
+      "grad_norm": 2.997104592016241,
+      "learning_rate": 1.175770677702359e-08,
+      "loss": 0.6325,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9529610829103214,
+      "grad_norm": 3.428867574227346,
+      "learning_rate": 1.1535246432510249e-08,
+      "loss": 0.6784,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9534122955442752,
+      "grad_norm": 3.215528072634851,
+      "learning_rate": 1.1314898545714768e-08,
+      "loss": 0.5196,
+      "step": 2113
+    },
+    {
+      "epoch": 0.953863508178229,
+      "grad_norm": 3.0135546799565507,
+      "learning_rate": 1.1096663587544574e-08,
+      "loss": 0.4638,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9543147208121827,
+      "grad_norm": 3.4816204927235277,
+      "learning_rate": 1.0880542024391926e-08,
+      "loss": 0.5687,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9547659334461365,
+      "grad_norm": 3.2568560809861333,
+      "learning_rate": 1.0666534318132248e-08,
+      "loss": 0.521,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9552171460800902,
+      "grad_norm": 3.1510176627650037,
+      "learning_rate": 1.0454640926123581e-08,
+      "loss": 0.6516,
+      "step": 2117
+    },
+    {
+      "epoch": 0.955668358714044,
+      "grad_norm": 3.3030358895502543,
+      "learning_rate": 1.0244862301205248e-08,
+      "loss": 0.4968,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9561195713479977,
+      "grad_norm": 3.204100014366423,
+      "learning_rate": 1.0037198891697297e-08,
+      "loss": 0.5478,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9565707839819515,
+      "grad_norm": 3.180566457304421,
+      "learning_rate": 9.831651141399167e-09,
+      "loss": 0.6166,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9570219966159053,
+      "grad_norm": 2.9835868653318114,
+      "learning_rate": 9.62821948958914e-09,
+      "loss": 0.5578,
+      "step": 2121
+    },
+    {
+      "epoch": 0.957473209249859,
+      "grad_norm": 3.335223111473605,
+      "learning_rate": 9.42690437102267e-09,
+      "loss": 0.6616,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9579244218838128,
+      "grad_norm": 3.0269577095198965,
+      "learning_rate": 9.227706215932718e-09,
+      "loss": 0.4504,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9583756345177665,
+      "grad_norm": 3.034533006063733,
+      "learning_rate": 9.030625450027197e-09,
+      "loss": 0.535,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9588268471517203,
+      "grad_norm": 3.4956858458681195,
+      "learning_rate": 8.835662494489638e-09,
+      "loss": 0.5865,
+      "step": 2125
+    },
+    {
+      "epoch": 0.959278059785674,
+      "grad_norm": 3.2430278438524804,
+      "learning_rate": 8.642817765977084e-09,
+      "loss": 0.568,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9597292724196278,
+      "grad_norm": 3.259168460054936,
+      "learning_rate": 8.452091676619976e-09,
+      "loss": 0.4907,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9601804850535814,
+      "grad_norm": 2.9506525517388447,
+      "learning_rate": 8.263484634020934e-09,
+      "loss": 0.5421,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9606316976875352,
+      "grad_norm": 3.3734708681832846,
+      "learning_rate": 8.076997041253864e-09,
+      "loss": 0.6559,
+      "step": 2129
+    },
+    {
+      "epoch": 0.961082910321489,
+      "grad_norm": 3.239791595659205,
+      "learning_rate": 7.892629296863296e-09,
+      "loss": 0.6207,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9615341229554427,
+      "grad_norm": 2.9754850080940325,
+      "learning_rate": 7.710381794863275e-09,
+      "loss": 0.4749,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9619853355893965,
+      "grad_norm": 3.1775519947710387,
+      "learning_rate": 7.53025492473669e-09,
+      "loss": 0.6391,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9624365482233502,
+      "grad_norm": 3.066624979505507,
+      "learning_rate": 7.352249071434613e-09,
+      "loss": 0.6327,
+      "step": 2133
+    },
+    {
+      "epoch": 0.962887760857304,
+      "grad_norm": 3.1723120674863563,
+      "learning_rate": 7.176364615374964e-09,
+      "loss": 0.4504,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9633389734912577,
+      "grad_norm": 3.026276568695946,
+      "learning_rate": 7.002601932442176e-09,
+      "loss": 0.5376,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9637901861252115,
+      "grad_norm": 3.0281976135848914,
+      "learning_rate": 6.830961393986201e-09,
+      "loss": 0.5515,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9642413987591653,
+      "grad_norm": 3.5847532083995337,
+      "learning_rate": 6.661443366821618e-09,
+      "loss": 0.5691,
+      "step": 2137
+    },
+    {
+      "epoch": 0.964692611393119,
+      "grad_norm": 3.1294010668010026,
+      "learning_rate": 6.4940482132272985e-09,
+      "loss": 0.5833,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9651438240270728,
+      "grad_norm": 3.387530884932056,
+      "learning_rate": 6.3287762909447486e-09,
+      "loss": 0.5154,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9655950366610265,
+      "grad_norm": 3.646146989054173,
+      "learning_rate": 6.165627953178432e-09,
+      "loss": 0.6854,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9660462492949803,
+      "grad_norm": 3.083896198719111,
+      "learning_rate": 6.0046035485941114e-09,
+      "loss": 0.64,
+      "step": 2141
+    },
+    {
+      "epoch": 0.966497461928934,
+      "grad_norm": 3.1945218965338165,
+      "learning_rate": 5.845703421318849e-09,
+      "loss": 0.5138,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9669486745628878,
+      "grad_norm": 2.9725665961761156,
+      "learning_rate": 5.688927910939445e-09,
+      "loss": 0.5811,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9673998871968416,
+      "grad_norm": 3.065001768071281,
+      "learning_rate": 5.534277352502448e-09,
+      "loss": 0.4917,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9678510998307953,
+      "grad_norm": 2.9575837796796147,
+      "learning_rate": 5.381752076513146e-09,
+      "loss": 0.4729,
+      "step": 2145
+    },
+    {
+      "epoch": 0.968302312464749,
+      "grad_norm": 3.490044797232381,
+      "learning_rate": 5.231352408934686e-09,
+      "loss": 0.5855,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9687535250987027,
+      "grad_norm": 3.2834883714280823,
+      "learning_rate": 5.083078671187846e-09,
+      "loss": 0.603,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9692047377326565,
+      "grad_norm": 3.6227714783334855,
+      "learning_rate": 4.936931180149706e-09,
+      "loss": 0.5146,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9696559503666102,
+      "grad_norm": 3.360342852449879,
+      "learning_rate": 4.792910248153537e-09,
+      "loss": 0.6926,
+      "step": 2149
+    },
+    {
+      "epoch": 0.970107163000564,
+      "grad_norm": 2.9745190180859997,
+      "learning_rate": 4.6510161829880215e-09,
+      "loss": 0.4959,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9705583756345177,
+      "grad_norm": 2.8409802136356115,
+      "learning_rate": 4.511249287896257e-09,
+      "loss": 0.4792,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9710095882684715,
+      "grad_norm": 2.9527499447768024,
+      "learning_rate": 4.373609861575422e-09,
+      "loss": 0.5262,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9714608009024253,
+      "grad_norm": 2.8234329667683995,
+      "learning_rate": 4.238098198175999e-09,
+      "loss": 0.453,
+      "step": 2153
+    },
+    {
+      "epoch": 0.971912013536379,
+      "grad_norm": 3.1955934495623177,
+      "learning_rate": 4.1047145873015494e-09,
+      "loss": 0.5705,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9723632261703328,
+      "grad_norm": 3.2186315351530177,
+      "learning_rate": 3.9734593140072766e-09,
+      "loss": 0.5774,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9728144388042865,
+      "grad_norm": 3.2781816568797115,
+      "learning_rate": 3.844332658800131e-09,
+      "loss": 0.534,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9732656514382403,
+      "grad_norm": 3.1271563804768516,
+      "learning_rate": 3.717334897638147e-09,
+      "loss": 0.4134,
+      "step": 2157
+    },
+    {
+      "epoch": 0.973716864072194,
+      "grad_norm": 3.1072908879945045,
+      "learning_rate": 3.59246630192922e-09,
+      "loss": 0.6476,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9741680767061478,
+      "grad_norm": 3.0501654737467048,
+      "learning_rate": 3.469727138531442e-09,
+      "loss": 0.517,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9746192893401016,
+      "grad_norm": 3.1296507770053585,
+      "learning_rate": 3.3491176697517663e-09,
+      "loss": 0.5104,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9750705019740553,
+      "grad_norm": 3.452718374388686,
+      "learning_rate": 3.2306381533460103e-09,
+      "loss": 0.572,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9755217146080091,
+      "grad_norm": 3.002928921865537,
+      "learning_rate": 3.1142888425177428e-09,
+      "loss": 0.5037,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9759729272419628,
+      "grad_norm": 3.1485029119178862,
+      "learning_rate": 3.0000699859183965e-09,
+      "loss": 0.5625,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9764241398759165,
+      "grad_norm": 3.0075732113850138,
+      "learning_rate": 2.8879818276459357e-09,
+      "loss": 0.6007,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9768753525098702,
+      "grad_norm": 3.5339981344018234,
+      "learning_rate": 2.7780246072454103e-09,
+      "loss": 0.641,
+      "step": 2165
+    },
+    {
+      "epoch": 0.977326565143824,
+      "grad_norm": 2.9938714355039138,
+      "learning_rate": 2.6701985597071817e-09,
+      "loss": 0.6232,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 3.5748734649397735,
+      "learning_rate": 2.5645039154675863e-09,
+      "loss": 0.5054,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9782289904117315,
+      "grad_norm": 2.821490730111447,
+      "learning_rate": 2.4609409004074934e-09,
+      "loss": 0.3994,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9786802030456853,
+      "grad_norm": 3.2813541645652347,
+      "learning_rate": 2.3595097358525275e-09,
+      "loss": 0.5495,
+      "step": 2169
+    },
+    {
+      "epoch": 0.979131415679639,
+      "grad_norm": 3.1682355394847694,
+      "learning_rate": 2.26021063857218e-09,
+      "loss": 0.4805,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9795826283135928,
+      "grad_norm": 3.940594922294758,
+      "learning_rate": 2.1630438207795864e-09,
+      "loss": 0.5344,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9800338409475465,
+      "grad_norm": 2.8957814567912292,
+      "learning_rate": 2.068009490130862e-09,
+      "loss": 0.4688,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9804850535815003,
+      "grad_norm": 3.2705425141515643,
+      "learning_rate": 1.9751078497248776e-09,
+      "loss": 0.5038,
+      "step": 2173
+    },
+    {
+      "epoch": 0.980936266215454,
+      "grad_norm": 3.2831548812730134,
+      "learning_rate": 1.884339098102483e-09,
+      "loss": 0.4829,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9813874788494078,
+      "grad_norm": 3.0060524899921393,
+      "learning_rate": 1.79570342924662e-09,
+      "loss": 0.5734,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9818386914833616,
+      "grad_norm": 3.2596135761529332,
+      "learning_rate": 1.709201032581431e-09,
+      "loss": 0.4523,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9822899041173153,
+      "grad_norm": 2.987843850321189,
+      "learning_rate": 1.6248320929719283e-09,
+      "loss": 0.5607,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9827411167512691,
+      "grad_norm": 2.9922698251969893,
+      "learning_rate": 1.5425967907239933e-09,
+      "loss": 0.5095,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9831923293852228,
+      "grad_norm": 3.1200632797078933,
+      "learning_rate": 1.4624953015832663e-09,
+      "loss": 0.5743,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9836435420191766,
+      "grad_norm": 3.0528829193084914,
+      "learning_rate": 1.3845277967355905e-09,
+      "loss": 0.4337,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9840947546531302,
+      "grad_norm": 2.9733715654870614,
+      "learning_rate": 1.3086944428060132e-09,
+      "loss": 0.3556,
+      "step": 2181
+    },
+    {
+      "epoch": 0.984545967287084,
+      "grad_norm": 2.9735610218703026,
+      "learning_rate": 1.234995401858785e-09,
+      "loss": 0.5374,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9849971799210377,
+      "grad_norm": 3.3255247596982183,
+      "learning_rate": 1.1634308313966944e-09,
+      "loss": 0.5705,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9854483925549915,
+      "grad_norm": 3.379807226305083,
+      "learning_rate": 1.0940008843612903e-09,
+      "loss": 0.6434,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9858996051889453,
+      "grad_norm": 3.2662985750115894,
+      "learning_rate": 1.026705709131992e-09,
+      "loss": 0.5889,
+      "step": 2185
+    },
+    {
+      "epoch": 0.986350817822899,
+      "grad_norm": 3.4188254818839927,
+      "learning_rate": 9.61545449525758e-10,
+      "loss": 0.4827,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9868020304568528,
+      "grad_norm": 3.3483494345972713,
+      "learning_rate": 8.985202447974183e-10,
+      "loss": 0.5504,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9872532430908065,
+      "grad_norm": 3.2171372630578676,
+      "learning_rate": 8.376302296387861e-10,
+      "loss": 0.5418,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9877044557247603,
+      "grad_norm": 3.6688593817328794,
+      "learning_rate": 7.788755341783249e-10,
+      "loss": 0.5123,
+      "step": 2189
+    },
+    {
+      "epoch": 0.988155668358714,
+      "grad_norm": 3.6499281007897153,
+      "learning_rate": 7.222562839813706e-10,
+      "loss": 0.6186,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9886068809926678,
+      "grad_norm": 3.39857073628445,
+      "learning_rate": 6.677726000494655e-10,
+      "loss": 0.5808,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9890580936266216,
+      "grad_norm": 3.5449673321180204,
+      "learning_rate": 6.154245988202466e-10,
+      "loss": 0.7051,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9895093062605753,
+      "grad_norm": 3.248475022142295,
+      "learning_rate": 5.652123921672247e-10,
+      "loss": 0.5403,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9899605188945291,
+      "grad_norm": 3.3216368867145736,
+      "learning_rate": 5.171360873991171e-10,
+      "loss": 0.5873,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9904117315284828,
+      "grad_norm": 3.0891117032715267,
+      "learning_rate": 4.711957872606254e-10,
+      "loss": 0.5611,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9908629441624366,
+      "grad_norm": 3.340990779373879,
+      "learning_rate": 4.273915899309921e-10,
+      "loss": 0.5511,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9913141567963903,
+      "grad_norm": 3.062648006172101,
+      "learning_rate": 3.857235890245558e-10,
+      "loss": 0.4507,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9917653694303441,
+      "grad_norm": 3.168752716935158,
+      "learning_rate": 3.461918735905289e-10,
+      "loss": 0.5381,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9922165820642977,
+      "grad_norm": 3.215884817371043,
+      "learning_rate": 3.0879652811255376e-10,
+      "loss": 0.5669,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9926677946982515,
+      "grad_norm": 3.3933100109929524,
+      "learning_rate": 2.735376325084804e-10,
+      "loss": 0.5129,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9931190073322053,
+      "grad_norm": 3.2902465372899896,
+      "learning_rate": 2.404152621305888e-10,
+      "loss": 0.5563,
+      "step": 2201
+    },
+    {
+      "epoch": 0.993570219966159,
+      "grad_norm": 3.2517872002551265,
+      "learning_rate": 2.0942948776481173e-10,
+      "loss": 0.4794,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9940214326001128,
+      "grad_norm": 3.0936052923048227,
+      "learning_rate": 1.805803756314006e-10,
+      "loss": 0.53,
+      "step": 2203
+    },
+    {
+      "epoch": 0.9944726452340665,
+      "grad_norm": 3.3848557451197743,
+      "learning_rate": 1.5386798738381557e-10,
+      "loss": 0.5361,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9949238578680203,
+      "grad_norm": 2.5261957634383094,
+      "learning_rate": 1.292923801096135e-10,
+      "loss": 0.3873,
+      "step": 2205
+    },
+    {
+      "epoch": 0.995375070501974,
+      "grad_norm": 3.2421625947937587,
+      "learning_rate": 1.0685360632933793e-10,
+      "loss": 0.5038,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9958262831359278,
+      "grad_norm": 3.283331275472657,
+      "learning_rate": 8.655171399718497e-11,
+      "loss": 0.568,
+      "step": 2207
+    },
+    {
+      "epoch": 0.9962774957698816,
+      "grad_norm": 2.9967877566234473,
+      "learning_rate": 6.838674650067044e-11,
+      "loss": 0.4354,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9967287084038353,
+      "grad_norm": 3.2686080869856142,
+      "learning_rate": 5.235874266018569e-11,
+      "loss": 0.4931,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9971799210377891,
+      "grad_norm": 2.926365790766811,
+      "learning_rate": 3.846773672933068e-11,
+      "loss": 0.6223,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9976311336717428,
+      "grad_norm": 3.0651076068440113,
+      "learning_rate": 2.6713758394802943e-11,
+      "loss": 0.6735,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9980823463056966,
+      "grad_norm": 3.317938299846236,
+      "learning_rate": 1.709683277606455e-11,
+      "loss": 0.5608,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9985335589396503,
+      "grad_norm": 3.170578430333791,
+      "learning_rate": 9.616980425453113e-12,
+      "loss": 0.569,
+      "step": 2213
+    },
+    {
+      "epoch": 0.9989847715736041,
+      "grad_norm": 2.805405937442547,
+      "learning_rate": 4.274217328514851e-12,
+      "loss": 0.5771,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9994359842075579,
+      "grad_norm": 3.2152320900605185,
+      "learning_rate": 1.068554903005392e-12,
+      "loss": 0.4857,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9998871968415116,
+      "grad_norm": 3.2758141104619267,
+      "learning_rate": 0.0,
+      "loss": 0.479,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9998871968415116,
+      "step": 2216,
+      "total_flos": 1474645333573632.0,
+      "train_loss": 0.580713667661382,
+      "train_runtime": 104901.7766,
+      "train_samples_per_second": 1.352,
+      "train_steps_per_second": 0.021
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2216,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1474645333573632.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}