diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14042 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.988280338049213,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014941401690246065,
+      "grad_norm": 2.2013847827911377,
+      "learning_rate": 4.9975e-05,
+      "loss": 6.2498,
+      "step": 1
+    },
+    {
+      "epoch": 0.002988280338049213,
+      "grad_norm": 2.3608126640319824,
+      "learning_rate": 4.995e-05,
+      "loss": 6.4856,
+      "step": 2
+    },
+    {
+      "epoch": 0.0044824205070738195,
+      "grad_norm": 2.7555134296417236,
+      "learning_rate": 4.992500000000001e-05,
+      "loss": 6.5519,
+      "step": 3
+    },
+    {
+      "epoch": 0.005976560676098426,
+      "grad_norm": 2.7937450408935547,
+      "learning_rate": 4.99e-05,
+      "loss": 6.6085,
+      "step": 4
+    },
+    {
+      "epoch": 0.0074707008451230334,
+      "grad_norm": 3.405029058456421,
+      "learning_rate": 4.9875000000000006e-05,
+      "loss": 6.4712,
+      "step": 5
+    },
+    {
+      "epoch": 0.008964841014147639,
+      "grad_norm": 2.5555474758148193,
+      "learning_rate": 4.9850000000000006e-05,
+      "loss": 6.7113,
+      "step": 6
+    },
+    {
+      "epoch": 0.010458981183172246,
+      "grad_norm": 3.9536023139953613,
+      "learning_rate": 4.9825000000000005e-05,
+      "loss": 6.3192,
+      "step": 7
+    },
+    {
+      "epoch": 0.011953121352196852,
+      "grad_norm": 3.3245086669921875,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 6.6072,
+      "step": 8
+    },
+    {
+      "epoch": 0.01344726152122146,
+      "grad_norm": 3.1858065128326416,
+      "learning_rate": 4.9775000000000004e-05,
+      "loss": 6.3415,
+      "step": 9
+    },
+    {
+      "epoch": 0.014941401690246067,
+      "grad_norm": 3.4829370975494385,
+      "learning_rate": 4.975e-05,
+      "loss": 6.5856,
+      "step": 10
+    },
+    {
+      "epoch": 0.016435541859270673,
+      "grad_norm": 3.615802526473999,
+      "learning_rate": 4.9725e-05,
+      "loss": 6.0762,
+      "step": 11
+    },
+    {
+      "epoch": 0.017929682028295278,
+      "grad_norm": 3.8015692234039307,
+      "learning_rate": 4.97e-05,
+      "loss": 6.2116,
+      "step": 12
+    },
+    {
+      "epoch": 0.019423822197319886,
+      "grad_norm": 3.587550163269043,
+      "learning_rate": 4.967500000000001e-05,
+      "loss": 6.1468,
+      "step": 13
+    },
+    {
+      "epoch": 0.02091796236634449,
+      "grad_norm": 3.3607726097106934,
+      "learning_rate": 4.965e-05,
+      "loss": 6.063,
+      "step": 14
+    },
+    {
+      "epoch": 0.0224121025353691,
+      "grad_norm": 2.8799898624420166,
+      "learning_rate": 4.962500000000001e-05,
+      "loss": 5.7877,
+      "step": 15
+    },
+    {
+      "epoch": 0.023906242704393704,
+      "grad_norm": 3.8595519065856934,
+      "learning_rate": 4.96e-05,
+      "loss": 5.6556,
+      "step": 16
+    },
+    {
+      "epoch": 0.025400382873418312,
+      "grad_norm": 3.544861316680908,
+      "learning_rate": 4.9575000000000006e-05,
+      "loss": 6.2185,
+      "step": 17
+    },
+    {
+      "epoch": 0.02689452304244292,
+      "grad_norm": 3.7592461109161377,
+      "learning_rate": 4.9550000000000005e-05,
+      "loss": 5.6731,
+      "step": 18
+    },
+    {
+      "epoch": 0.028388663211467526,
+      "grad_norm": 4.296304225921631,
+      "learning_rate": 4.9525000000000004e-05,
+      "loss": 5.7575,
+      "step": 19
+    },
+    {
+      "epoch": 0.029882803380492134,
+      "grad_norm": 3.575627088546753,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 5.6663,
+      "step": 20
+    },
+    {
+      "epoch": 0.03137694354951674,
+      "grad_norm": 3.7745871543884277,
+      "learning_rate": 4.9475e-05,
+      "loss": 5.4617,
+      "step": 21
+    },
+    {
+      "epoch": 0.03287108371854135,
+      "grad_norm": 3.608196258544922,
+      "learning_rate": 4.945e-05,
+      "loss": 5.5331,
+      "step": 22
+    },
+    {
+      "epoch": 0.03436522388756595,
+      "grad_norm": 3.33453106880188,
+      "learning_rate": 4.9425e-05,
+      "loss": 5.3907,
+      "step": 23
+    },
+    {
+      "epoch": 0.035859364056590556,
+      "grad_norm": 4.210156440734863,
+      "learning_rate": 4.94e-05,
+      "loss": 5.1661,
+      "step": 24
+    },
+    {
+      "epoch": 0.03735350422561517,
+      "grad_norm": 3.6300511360168457,
+      "learning_rate": 4.937500000000001e-05,
+      "loss": 5.0496,
+      "step": 25
+    },
+    {
+      "epoch": 0.03884764439463977,
+      "grad_norm": 4.4206671714782715,
+      "learning_rate": 4.935e-05,
+      "loss": 4.7727,
+      "step": 26
+    },
+    {
+      "epoch": 0.04034178456366438,
+      "grad_norm": 3.644752025604248,
+      "learning_rate": 4.9325000000000006e-05,
+      "loss": 4.832,
+      "step": 27
+    },
+    {
+      "epoch": 0.04183592473268898,
+      "grad_norm": 3.7340240478515625,
+      "learning_rate": 4.93e-05,
+      "loss": 4.9109,
+      "step": 28
+    },
+    {
+      "epoch": 0.043330064901713594,
+      "grad_norm": 3.7647705078125,
+      "learning_rate": 4.9275000000000005e-05,
+      "loss": 4.7356,
+      "step": 29
+    },
+    {
+      "epoch": 0.0448242050707382,
+      "grad_norm": 3.2214791774749756,
+      "learning_rate": 4.9250000000000004e-05,
+      "loss": 4.3419,
+      "step": 30
+    },
+    {
+      "epoch": 0.046318345239762804,
+      "grad_norm": 3.0918350219726562,
+      "learning_rate": 4.9225000000000004e-05,
+      "loss": 4.8978,
+      "step": 31
+    },
+    {
+      "epoch": 0.04781248540878741,
+      "grad_norm": 3.302729368209839,
+      "learning_rate": 4.92e-05,
+      "loss": 4.6097,
+      "step": 32
+    },
+    {
+      "epoch": 0.04930662557781202,
+      "grad_norm": 3.2680063247680664,
+      "learning_rate": 4.9175e-05,
+      "loss": 4.8161,
+      "step": 33
+    },
+    {
+      "epoch": 0.050800765746836625,
+      "grad_norm": 2.8400955200195312,
+      "learning_rate": 4.915e-05,
+      "loss": 4.236,
+      "step": 34
+    },
+    {
+      "epoch": 0.05229490591586123,
+      "grad_norm": 3.210604429244995,
+      "learning_rate": 4.9125e-05,
+      "loss": 4.4979,
+      "step": 35
+    },
+    {
+      "epoch": 0.05378904608488584,
+      "grad_norm": 2.803765058517456,
+      "learning_rate": 4.91e-05,
+      "loss": 4.0503,
+      "step": 36
+    },
+    {
+      "epoch": 0.055283186253910446,
+      "grad_norm": 3.3603885173797607,
+      "learning_rate": 4.907500000000001e-05,
+      "loss": 3.8568,
+      "step": 37
+    },
+    {
+      "epoch": 0.05677732642293505,
+      "grad_norm": 2.6471378803253174,
+      "learning_rate": 4.905e-05,
+      "loss": 4.2472,
+      "step": 38
+    },
+    {
+      "epoch": 0.058271466591959656,
+      "grad_norm": 3.159874439239502,
+      "learning_rate": 4.9025000000000006e-05,
+      "loss": 4.2832,
+      "step": 39
+    },
+    {
+      "epoch": 0.05976560676098427,
+      "grad_norm": 2.405097484588623,
+      "learning_rate": 4.9e-05,
+      "loss": 3.7952,
+      "step": 40
+    },
+    {
+      "epoch": 0.06125974693000887,
+      "grad_norm": 3.4419732093811035,
+      "learning_rate": 4.8975000000000005e-05,
+      "loss": 3.8268,
+      "step": 41
+    },
+    {
+      "epoch": 0.06275388709903348,
+      "grad_norm": 3.0664589405059814,
+      "learning_rate": 4.8950000000000004e-05,
+      "loss": 3.8268,
+      "step": 42
+    },
+    {
+      "epoch": 0.06424802726805809,
+      "grad_norm": 2.7936577796936035,
+      "learning_rate": 4.8925e-05,
+      "loss": 3.8516,
+      "step": 43
+    },
+    {
+      "epoch": 0.0657421674370827,
+      "grad_norm": 2.6919713020324707,
+      "learning_rate": 4.89e-05,
+      "loss": 3.5816,
+      "step": 44
+    },
+    {
+      "epoch": 0.0672363076061073,
+      "grad_norm": 2.2412493228912354,
+      "learning_rate": 4.8875e-05,
+      "loss": 3.7505,
+      "step": 45
+    },
+    {
+      "epoch": 0.0687304477751319,
+      "grad_norm": 2.645268440246582,
+      "learning_rate": 4.885e-05,
+      "loss": 3.4058,
+      "step": 46
+    },
+    {
+      "epoch": 0.07022458794415651,
+      "grad_norm": 2.3000919818878174,
+      "learning_rate": 4.8825e-05,
+      "loss": 3.328,
+      "step": 47
+    },
+    {
+      "epoch": 0.07171872811318111,
+      "grad_norm": 2.3414597511291504,
+      "learning_rate": 4.88e-05,
+      "loss": 3.246,
+      "step": 48
+    },
+    {
+      "epoch": 0.07321286828220572,
+      "grad_norm": 2.2140865325927734,
+      "learning_rate": 4.8775000000000007e-05,
+      "loss": 3.2796,
+      "step": 49
+    },
+    {
+      "epoch": 0.07470700845123034,
+      "grad_norm": 1.9956824779510498,
+      "learning_rate": 4.875e-05,
+      "loss": 3.2875,
+      "step": 50
+    },
+    {
+      "epoch": 0.07620114862025494,
+      "grad_norm": 1.9508098363876343,
+      "learning_rate": 4.8725000000000005e-05,
+      "loss": 3.0613,
+      "step": 51
+    },
+    {
+      "epoch": 0.07769528878927955,
+      "grad_norm": 1.8123174905776978,
+      "learning_rate": 4.87e-05,
+      "loss": 3.2806,
+      "step": 52
+    },
+    {
+      "epoch": 0.07918942895830415,
+      "grad_norm": 2.2940430641174316,
+      "learning_rate": 4.8675000000000004e-05,
+      "loss": 3.2443,
+      "step": 53
+    },
+    {
+      "epoch": 0.08068356912732876,
+      "grad_norm": 1.803460955619812,
+      "learning_rate": 4.8650000000000003e-05,
+      "loss": 3.17,
+      "step": 54
+    },
+    {
+      "epoch": 0.08217770929635336,
+      "grad_norm": 1.5010923147201538,
+      "learning_rate": 4.8625e-05,
+      "loss": 3.0046,
+      "step": 55
+    },
+    {
+      "epoch": 0.08367184946537796,
+      "grad_norm": 1.386048436164856,
+      "learning_rate": 4.86e-05,
+      "loss": 2.9352,
+      "step": 56
+    },
+    {
+      "epoch": 0.08516598963440258,
+      "grad_norm": 1.6107743978500366,
+      "learning_rate": 4.8575e-05,
+      "loss": 3.0539,
+      "step": 57
+    },
+    {
+      "epoch": 0.08666012980342719,
+      "grad_norm": 1.336033821105957,
+      "learning_rate": 4.855e-05,
+      "loss": 2.9155,
+      "step": 58
+    },
+    {
+      "epoch": 0.0881542699724518,
+      "grad_norm": 1.5201988220214844,
+      "learning_rate": 4.8525e-05,
+      "loss": 2.8984,
+      "step": 59
+    },
+    {
+      "epoch": 0.0896484101414764,
+      "grad_norm": 1.316045880317688,
+      "learning_rate": 4.85e-05,
+      "loss": 2.7086,
+      "step": 60
+    },
+    {
+      "epoch": 0.091142550310501,
+      "grad_norm": 1.3400336503982544,
+      "learning_rate": 4.8475000000000006e-05,
+      "loss": 2.8407,
+      "step": 61
+    },
+    {
+      "epoch": 0.09263669047952561,
+      "grad_norm": 1.3193714618682861,
+      "learning_rate": 4.845e-05,
+      "loss": 2.8589,
+      "step": 62
+    },
+    {
+      "epoch": 0.09413083064855021,
+      "grad_norm": 1.0813387632369995,
+      "learning_rate": 4.8425000000000005e-05,
+      "loss": 2.4711,
+      "step": 63
+    },
+    {
+      "epoch": 0.09562497081757482,
+      "grad_norm": 1.3248521089553833,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 2.9868,
+      "step": 64
+    },
+    {
+      "epoch": 0.09711911098659944,
+      "grad_norm": 1.227431058883667,
+      "learning_rate": 4.8375000000000004e-05,
+      "loss": 2.6897,
+      "step": 65
+    },
+    {
+      "epoch": 0.09861325115562404,
+      "grad_norm": 1.2781943082809448,
+      "learning_rate": 4.835e-05,
+      "loss": 2.6999,
+      "step": 66
+    },
+    {
+      "epoch": 0.10010739132464865,
+      "grad_norm": 1.2900972366333008,
+      "learning_rate": 4.8325e-05,
+      "loss": 2.9534,
+      "step": 67
+    },
+    {
+      "epoch": 0.10160153149367325,
+      "grad_norm": 1.0637670755386353,
+      "learning_rate": 4.83e-05,
+      "loss": 2.8384,
+      "step": 68
+    },
+    {
+      "epoch": 0.10309567166269785,
+      "grad_norm": 1.0107454061508179,
+      "learning_rate": 4.8275e-05,
+      "loss": 2.7397,
+      "step": 69
+    },
+    {
+      "epoch": 0.10458981183172246,
+      "grad_norm": 1.1608697175979614,
+      "learning_rate": 4.825e-05,
+      "loss": 2.686,
+      "step": 70
+    },
+    {
+      "epoch": 0.10608395200074706,
+      "grad_norm": 1.0701357126235962,
+      "learning_rate": 4.822500000000001e-05,
+      "loss": 2.6855,
+      "step": 71
+    },
+    {
+      "epoch": 0.10757809216977168,
+      "grad_norm": 1.022800326347351,
+      "learning_rate": 4.82e-05,
+      "loss": 2.5465,
+      "step": 72
+    },
+    {
+      "epoch": 0.10907223233879629,
+      "grad_norm": 1.0642368793487549,
+      "learning_rate": 4.8175000000000005e-05,
+      "loss": 2.6804,
+      "step": 73
+    },
+    {
+      "epoch": 0.11056637250782089,
+      "grad_norm": 0.907848596572876,
+      "learning_rate": 4.815e-05,
+      "loss": 2.4118,
+      "step": 74
+    },
+    {
+      "epoch": 0.1120605126768455,
+      "grad_norm": 1.2253451347351074,
+      "learning_rate": 4.8125000000000004e-05,
+      "loss": 2.6588,
+      "step": 75
+    },
+    {
+      "epoch": 0.1135546528458701,
+      "grad_norm": 1.0567305088043213,
+      "learning_rate": 4.8100000000000004e-05,
+      "loss": 2.615,
+      "step": 76
+    },
+    {
+      "epoch": 0.1150487930148947,
+      "grad_norm": 1.2600438594818115,
+      "learning_rate": 4.8075e-05,
+      "loss": 2.7854,
+      "step": 77
+    },
+    {
+      "epoch": 0.11654293318391931,
+      "grad_norm": 1.0818730592727661,
+      "learning_rate": 4.805e-05,
+      "loss": 2.2592,
+      "step": 78
+    },
+    {
+      "epoch": 0.11803707335294392,
+      "grad_norm": 1.1989450454711914,
+      "learning_rate": 4.8025e-05,
+      "loss": 2.6293,
+      "step": 79
+    },
+    {
+      "epoch": 0.11953121352196854,
+      "grad_norm": 1.0635451078414917,
+      "learning_rate": 4.8e-05,
+      "loss": 2.5789,
+      "step": 80
+    },
+    {
+      "epoch": 0.12102535369099314,
+      "grad_norm": 1.5961538553237915,
+      "learning_rate": 4.7975e-05,
+      "loss": 2.1708,
+      "step": 81
+    },
+    {
+      "epoch": 0.12251949386001774,
+      "grad_norm": 0.886963963508606,
+      "learning_rate": 4.795e-05,
+      "loss": 2.3345,
+      "step": 82
+    },
+    {
+      "epoch": 0.12401363402904235,
+      "grad_norm": 1.1009526252746582,
+      "learning_rate": 4.7925000000000006e-05,
+      "loss": 2.3957,
+      "step": 83
+    },
+    {
+      "epoch": 0.12550777419806697,
+      "grad_norm": 1.0787116289138794,
+      "learning_rate": 4.79e-05,
+      "loss": 2.4759,
+      "step": 84
+    },
+    {
+      "epoch": 0.12700191436709157,
+      "grad_norm": 0.9887699484825134,
+      "learning_rate": 4.7875000000000005e-05,
+      "loss": 2.4005,
+      "step": 85
+    },
+    {
+      "epoch": 0.12849605453611618,
+      "grad_norm": 0.9390841722488403,
+      "learning_rate": 4.785e-05,
+      "loss": 2.2801,
+      "step": 86
+    },
+    {
+      "epoch": 0.12999019470514078,
+      "grad_norm": 0.9958310127258301,
+      "learning_rate": 4.7825000000000004e-05,
+      "loss": 2.4892,
+      "step": 87
+    },
+    {
+      "epoch": 0.1314843348741654,
+      "grad_norm": 0.8534329533576965,
+      "learning_rate": 4.78e-05,
+      "loss": 1.9908,
+      "step": 88
+    },
+    {
+      "epoch": 0.13297847504319,
+      "grad_norm": 0.8332351446151733,
+      "learning_rate": 4.7775e-05,
+      "loss": 2.2827,
+      "step": 89
+    },
+    {
+      "epoch": 0.1344726152122146,
+      "grad_norm": 0.845981776714325,
+      "learning_rate": 4.775e-05,
+      "loss": 2.4051,
+      "step": 90
+    },
+    {
+      "epoch": 0.1359667553812392,
+      "grad_norm": 0.9278237223625183,
+      "learning_rate": 4.7725e-05,
+      "loss": 2.3518,
+      "step": 91
+    },
+    {
+      "epoch": 0.1374608955502638,
+      "grad_norm": 1.276440143585205,
+      "learning_rate": 4.77e-05,
+      "loss": 2.7287,
+      "step": 92
+    },
+    {
+      "epoch": 0.1389550357192884,
+      "grad_norm": 1.0569902658462524,
+      "learning_rate": 4.7675e-05,
+      "loss": 2.5747,
+      "step": 93
+    },
+    {
+      "epoch": 0.14044917588831302,
+      "grad_norm": 0.9122519493103027,
+      "learning_rate": 4.765e-05,
+      "loss": 2.186,
+      "step": 94
+    },
+    {
+      "epoch": 0.14194331605733762,
+      "grad_norm": 0.751939594745636,
+      "learning_rate": 4.7625000000000006e-05,
+      "loss": 2.3423,
+      "step": 95
+    },
+    {
+      "epoch": 0.14343745622636223,
+      "grad_norm": 0.95006263256073,
+      "learning_rate": 4.76e-05,
+      "loss": 2.5864,
+      "step": 96
+    },
+    {
+      "epoch": 0.14493159639538683,
+      "grad_norm": 0.9910762906074524,
+      "learning_rate": 4.7575000000000004e-05,
+      "loss": 2.312,
+      "step": 97
+    },
+    {
+      "epoch": 0.14642573656441144,
+      "grad_norm": 1.133583664894104,
+      "learning_rate": 4.755e-05,
+      "loss": 2.4763,
+      "step": 98
+    },
+    {
+      "epoch": 0.14791987673343607,
+      "grad_norm": 1.2710939645767212,
+      "learning_rate": 4.7525e-05,
+      "loss": 2.605,
+      "step": 99
+    },
+    {
+      "epoch": 0.14941401690246067,
+      "grad_norm": 1.0130882263183594,
+      "learning_rate": 4.75e-05,
+      "loss": 2.416,
+      "step": 100
+    },
+    {
+      "epoch": 0.15090815707148528,
+      "grad_norm": 1.4635257720947266,
+      "learning_rate": 4.7475e-05,
+      "loss": 2.6169,
+      "step": 101
+    },
+    {
+      "epoch": 0.15240229724050988,
+      "grad_norm": 0.9598097801208496,
+      "learning_rate": 4.745e-05,
+      "loss": 2.4536,
+      "step": 102
+    },
+    {
+      "epoch": 0.1538964374095345,
+      "grad_norm": 0.8063713908195496,
+      "learning_rate": 4.7425e-05,
+      "loss": 2.2044,
+      "step": 103
+    },
+    {
+      "epoch": 0.1553905775785591,
+      "grad_norm": 0.9389742612838745,
+      "learning_rate": 4.74e-05,
+      "loss": 2.2766,
+      "step": 104
+    },
+    {
+      "epoch": 0.1568847177475837,
+      "grad_norm": 0.7887735962867737,
+      "learning_rate": 4.7375e-05,
+      "loss": 1.8076,
+      "step": 105
+    },
+    {
+      "epoch": 0.1583788579166083,
+      "grad_norm": 0.8454030156135559,
+      "learning_rate": 4.735e-05,
+      "loss": 2.3198,
+      "step": 106
+    },
+    {
+      "epoch": 0.1598729980856329,
+      "grad_norm": 1.3265002965927124,
+      "learning_rate": 4.7325000000000005e-05,
+      "loss": 2.1676,
+      "step": 107
+    },
+    {
+      "epoch": 0.1613671382546575,
+      "grad_norm": 0.8800100088119507,
+      "learning_rate": 4.73e-05,
+      "loss": 2.0926,
+      "step": 108
+    },
+    {
+      "epoch": 0.16286127842368212,
+      "grad_norm": 0.8939815163612366,
+      "learning_rate": 4.7275000000000004e-05,
+      "loss": 2.1965,
+      "step": 109
+    },
+    {
+      "epoch": 0.16435541859270672,
+      "grad_norm": 0.9632463455200195,
+      "learning_rate": 4.7249999999999997e-05,
+      "loss": 2.1399,
+      "step": 110
+    },
+    {
+      "epoch": 0.16584955876173132,
+      "grad_norm": 1.092432975769043,
+      "learning_rate": 4.7225e-05,
+      "loss": 2.2975,
+      "step": 111
+    },
+    {
+      "epoch": 0.16734369893075593,
+      "grad_norm": 0.9222836494445801,
+      "learning_rate": 4.72e-05,
+      "loss": 2.3583,
+      "step": 112
+    },
+    {
+      "epoch": 0.16883783909978053,
+      "grad_norm": 0.9905472993850708,
+      "learning_rate": 4.7175e-05,
+      "loss": 2.0784,
+      "step": 113
+    },
+    {
+      "epoch": 0.17033197926880517,
+      "grad_norm": 0.6727510094642639,
+      "learning_rate": 4.715e-05,
+      "loss": 1.9479,
+      "step": 114
+    },
+    {
+      "epoch": 0.17182611943782977,
+      "grad_norm": 1.0887380838394165,
+      "learning_rate": 4.7125e-05,
+      "loss": 2.3239,
+      "step": 115
+    },
+    {
+      "epoch": 0.17332025960685438,
+      "grad_norm": 0.7983910441398621,
+      "learning_rate": 4.71e-05,
+      "loss": 2.1595,
+      "step": 116
+    },
+    {
+      "epoch": 0.17481439977587898,
+      "grad_norm": 0.8539843559265137,
+      "learning_rate": 4.7075e-05,
+      "loss": 1.9513,
+      "step": 117
+    },
+    {
+      "epoch": 0.1763085399449036,
+      "grad_norm": 0.8280434608459473,
+      "learning_rate": 4.705e-05,
+      "loss": 2.0896,
+      "step": 118
+    },
+    {
+      "epoch": 0.1778026801139282,
+      "grad_norm": 0.9694619178771973,
+      "learning_rate": 4.7025000000000005e-05,
+      "loss": 2.3539,
+      "step": 119
+    },
+    {
+      "epoch": 0.1792968202829528,
+      "grad_norm": 1.0440824031829834,
+      "learning_rate": 4.7e-05,
+      "loss": 1.7679,
+      "step": 120
+    },
+    {
+      "epoch": 0.1807909604519774,
+      "grad_norm": 0.8723125457763672,
+      "learning_rate": 4.6975000000000003e-05,
+      "loss": 2.0644,
+      "step": 121
+    },
+    {
+      "epoch": 0.182285100621002,
+      "grad_norm": 0.726588249206543,
+      "learning_rate": 4.695e-05,
+      "loss": 1.9727,
+      "step": 122
+    },
+    {
+      "epoch": 0.1837792407900266,
+      "grad_norm": 0.6932450532913208,
+      "learning_rate": 4.6925e-05,
+      "loss": 1.8967,
+      "step": 123
+    },
+    {
+      "epoch": 0.18527338095905121,
+      "grad_norm": 0.6471284627914429,
+      "learning_rate": 4.69e-05,
+      "loss": 1.8208,
+      "step": 124
+    },
+    {
+      "epoch": 0.18676752112807582,
+      "grad_norm": 0.6124398708343506,
+      "learning_rate": 4.6875e-05,
+      "loss": 1.9418,
+      "step": 125
+    },
+    {
+      "epoch": 0.18826166129710042,
+      "grad_norm": 0.705912709236145,
+      "learning_rate": 4.685000000000001e-05,
+      "loss": 1.9967,
+      "step": 126
+    },
+    {
+      "epoch": 0.18975580146612503,
+      "grad_norm": 0.8094897866249084,
+      "learning_rate": 4.6825e-05,
+      "loss": 1.9594,
+      "step": 127
+    },
+    {
+      "epoch": 0.19124994163514963,
+      "grad_norm": 0.5901124477386475,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 2.0488,
+      "step": 128
+    },
+    {
+      "epoch": 0.19274408180417427,
+      "grad_norm": 0.7298286557197571,
+      "learning_rate": 4.6775000000000005e-05,
+      "loss": 2.1777,
+      "step": 129
+    },
+    {
+      "epoch": 0.19423822197319887,
+      "grad_norm": 0.5055814981460571,
+      "learning_rate": 4.6750000000000005e-05,
+      "loss": 1.7352,
+      "step": 130
+    },
+    {
+      "epoch": 0.19573236214222348,
+      "grad_norm": 0.47730177640914917,
+      "learning_rate": 4.6725000000000004e-05,
+      "loss": 1.8427,
+      "step": 131
+    },
+    {
+      "epoch": 0.19722650231124808,
+      "grad_norm": 0.5291617512702942,
+      "learning_rate": 4.6700000000000003e-05,
+      "loss": 1.9703,
+      "step": 132
+    },
+    {
+      "epoch": 0.19872064248027269,
+      "grad_norm": 0.6130346059799194,
+      "learning_rate": 4.6675e-05,
+      "loss": 2.0337,
+      "step": 133
+    },
+    {
+      "epoch": 0.2002147826492973,
+      "grad_norm": 0.600026547908783,
+      "learning_rate": 4.665e-05,
+      "loss": 2.1049,
+      "step": 134
+    },
+    {
+      "epoch": 0.2017089228183219,
+      "grad_norm": 0.9462055563926697,
+      "learning_rate": 4.6625e-05,
+      "loss": 1.8345,
+      "step": 135
+    },
+    {
+      "epoch": 0.2032030629873465,
+      "grad_norm": 0.6463867425918579,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 1.9301,
+      "step": 136
+    },
+    {
+      "epoch": 0.2046972031563711,
+      "grad_norm": 0.7604928016662598,
+      "learning_rate": 4.6575e-05,
+      "loss": 1.9863,
+      "step": 137
+    },
+    {
+      "epoch": 0.2061913433253957,
+      "grad_norm": 0.5646475553512573,
+      "learning_rate": 4.655000000000001e-05,
+      "loss": 2.0415,
+      "step": 138
+    },
+    {
+      "epoch": 0.20768548349442031,
+      "grad_norm": 0.5505641102790833,
+      "learning_rate": 4.6525e-05,
+      "loss": 1.9946,
+      "step": 139
+    },
+    {
+      "epoch": 0.20917962366344492,
+      "grad_norm": 0.5963394641876221,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 1.7218,
+      "step": 140
+    },
+    {
+      "epoch": 0.21067376383246952,
+      "grad_norm": 1.011972188949585,
+      "learning_rate": 4.6475000000000005e-05,
+      "loss": 2.0932,
+      "step": 141
+    },
+    {
+      "epoch": 0.21216790400149413,
+      "grad_norm": 0.5645290017127991,
+      "learning_rate": 4.6450000000000004e-05,
+      "loss": 1.8721,
+      "step": 142
+    },
+    {
+      "epoch": 0.21366204417051873,
+      "grad_norm": 0.543994128704071,
+      "learning_rate": 4.6425000000000004e-05,
+      "loss": 1.8209,
+      "step": 143
+    },
+    {
+      "epoch": 0.21515618433954337,
+      "grad_norm": 0.6430279612541199,
+      "learning_rate": 4.64e-05,
+      "loss": 1.9537,
+      "step": 144
+    },
+    {
+      "epoch": 0.21665032450856797,
+      "grad_norm": 0.8161587119102478,
+      "learning_rate": 4.6375e-05,
+      "loss": 1.9483,
+      "step": 145
+    },
+    {
+      "epoch": 0.21814446467759258,
+      "grad_norm": 0.597017765045166,
+      "learning_rate": 4.635e-05,
+      "loss": 1.9586,
+      "step": 146
+    },
+    {
+      "epoch": 0.21963860484661718,
+      "grad_norm": 0.6092536449432373,
+      "learning_rate": 4.6325e-05,
+      "loss": 1.9523,
+      "step": 147
+    },
+    {
+      "epoch": 0.22113274501564179,
+      "grad_norm": 0.8086355328559875,
+      "learning_rate": 4.630000000000001e-05,
+      "loss": 1.7828,
+      "step": 148
+    },
+    {
+      "epoch": 0.2226268851846664,
+      "grad_norm": 0.4737461507320404,
+      "learning_rate": 4.6275e-05,
+      "loss": 1.936,
+      "step": 149
+    },
+    {
+      "epoch": 0.224121025353691,
+      "grad_norm": 0.7633687853813171,
+      "learning_rate": 4.6250000000000006e-05,
+      "loss": 1.9736,
+      "step": 150
+    },
+    {
+      "epoch": 0.2256151655227156,
+      "grad_norm": 0.6465319991111755,
+      "learning_rate": 4.6225e-05,
+      "loss": 1.6353,
+      "step": 151
+    },
+    {
+      "epoch": 0.2271093056917402,
+      "grad_norm": 0.7993711233139038,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 1.8528,
+      "step": 152
+    },
+    {
+      "epoch": 0.2286034458607648,
+      "grad_norm": 0.5374677777290344,
+      "learning_rate": 4.6175000000000004e-05,
+      "loss": 1.6402,
+      "step": 153
+    },
+    {
+      "epoch": 0.2300975860297894,
+      "grad_norm": 0.8310596942901611,
+      "learning_rate": 4.6150000000000004e-05,
+      "loss": 2.1833,
+      "step": 154
+    },
+    {
+      "epoch": 0.23159172619881402,
+      "grad_norm": 0.4872928261756897,
+      "learning_rate": 4.6125e-05,
+      "loss": 1.7698,
+      "step": 155
+    },
+    {
+      "epoch": 0.23308586636783862,
+      "grad_norm": 0.587377667427063,
+      "learning_rate": 4.61e-05,
+      "loss": 2.0211,
+      "step": 156
+    },
+    {
+      "epoch": 0.23458000653686323,
+      "grad_norm": 0.5460837483406067,
+      "learning_rate": 4.6075e-05,
+      "loss": 1.721,
+      "step": 157
+    },
+    {
+      "epoch": 0.23607414670588783,
+      "grad_norm": 0.7595339417457581,
+      "learning_rate": 4.605e-05,
+      "loss": 1.6841,
+      "step": 158
+    },
+    {
+      "epoch": 0.23756828687491247,
+      "grad_norm": 0.4985441565513611,
+      "learning_rate": 4.6025e-05,
+      "loss": 1.8048,
+      "step": 159
+    },
+    {
+      "epoch": 0.23906242704393707,
+      "grad_norm": 0.8894737958908081,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.1808,
+      "step": 160
+    },
+    {
+      "epoch": 0.24055656721296167,
+      "grad_norm": 0.49112433195114136,
+      "learning_rate": 4.5975e-05,
+      "loss": 1.7699,
+      "step": 161
+    },
+    {
+      "epoch": 0.24205070738198628,
+      "grad_norm": 0.8248782157897949,
+      "learning_rate": 4.5950000000000006e-05,
+      "loss": 1.8133,
+      "step": 162
+    },
+    {
+      "epoch": 0.24354484755101088,
+      "grad_norm": 1.1163716316223145,
+      "learning_rate": 4.5925e-05,
+      "loss": 1.755,
+      "step": 163
+    },
+    {
+      "epoch": 0.2450389877200355,
+      "grad_norm": 0.7373276948928833,
+      "learning_rate": 4.5900000000000004e-05,
+      "loss": 1.6897,
+      "step": 164
+    },
+    {
+      "epoch": 0.2465331278890601,
+      "grad_norm": 0.4406241476535797,
+      "learning_rate": 4.5875000000000004e-05,
+      "loss": 1.6829,
+      "step": 165
+    },
+    {
+      "epoch": 0.2480272680580847,
+      "grad_norm": 0.6179890036582947,
+      "learning_rate": 4.585e-05,
+      "loss": 1.9277,
+      "step": 166
+    },
+    {
+      "epoch": 0.2495214082271093,
+      "grad_norm": 0.600563108921051,
+      "learning_rate": 4.5825e-05,
+      "loss": 1.8681,
+      "step": 167
+    },
+    {
+      "epoch": 0.25101554839613394,
+      "grad_norm": 0.5876383781433105,
+      "learning_rate": 4.58e-05,
+      "loss": 1.8027,
+      "step": 168
+    },
+    {
+      "epoch": 0.25250968856515854,
+      "grad_norm": 0.6557374596595764,
+      "learning_rate": 4.5775e-05,
+      "loss": 2.0037,
+      "step": 169
+    },
+    {
+      "epoch": 0.25400382873418315,
+      "grad_norm": 0.5612519383430481,
+      "learning_rate": 4.575e-05,
+      "loss": 1.8295,
+      "step": 170
+    },
+    {
+      "epoch": 0.25549796890320775,
+      "grad_norm": 0.550281822681427,
+      "learning_rate": 4.5725e-05,
+      "loss": 1.7208,
+      "step": 171
+    },
+    {
+      "epoch": 0.25699210907223236,
+      "grad_norm": 0.5334203243255615,
+      "learning_rate": 4.5700000000000006e-05,
+      "loss": 1.7105,
+      "step": 172
+    },
+    {
+      "epoch": 0.25848624924125696,
+      "grad_norm": 0.5756668448448181,
+      "learning_rate": 4.5675e-05,
+      "loss": 1.8567,
+      "step": 173
+    },
+    {
+      "epoch": 0.25998038941028156,
+      "grad_norm": 0.5913483500480652,
+      "learning_rate": 4.5650000000000005e-05,
+      "loss": 1.9032,
+      "step": 174
+    },
+    {
+      "epoch": 0.26147452957930617,
+      "grad_norm": 0.7093982100486755,
+      "learning_rate": 4.5625e-05,
+      "loss": 2.0607,
+      "step": 175
+    },
+    {
+      "epoch": 0.2629686697483308,
+      "grad_norm": 0.7194991707801819,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 1.7715,
+      "step": 176
+    },
+    {
+      "epoch": 0.2644628099173554,
+      "grad_norm": 0.5984293222427368,
+      "learning_rate": 4.5575e-05,
+      "loss": 1.8837,
+      "step": 177
+    },
+    {
+      "epoch": 0.26595695008638,
+      "grad_norm": 0.5316656827926636,
+      "learning_rate": 4.555e-05,
+      "loss": 1.8816,
+      "step": 178
+    },
+    {
+      "epoch": 0.2674510902554046,
+      "grad_norm": 0.41810792684555054,
+      "learning_rate": 4.5525e-05,
+      "loss": 1.6823,
+      "step": 179
+    },
+    {
+      "epoch": 0.2689452304244292,
+      "grad_norm": 0.6792067885398865,
+      "learning_rate": 4.55e-05,
+      "loss": 1.9474,
+      "step": 180
+    },
+    {
+      "epoch": 0.2704393705934538,
+      "grad_norm": 0.758362889289856,
+      "learning_rate": 4.5475e-05,
+      "loss": 1.8358,
+      "step": 181
+    },
+    {
+      "epoch": 0.2719335107624784,
+      "grad_norm": 0.5369607210159302,
+      "learning_rate": 4.545000000000001e-05,
+      "loss": 1.8121,
+      "step": 182
+    },
+    {
+      "epoch": 0.273427650931503,
+      "grad_norm": 0.47773298621177673,
+      "learning_rate": 4.5425e-05,
+      "loss": 1.5175,
+      "step": 183
+    },
+    {
+      "epoch": 0.2749217911005276,
+      "grad_norm": 0.6082251667976379,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 1.6287,
+      "step": 184
+    },
+    {
+      "epoch": 0.2764159312695522,
+      "grad_norm": 0.7682433724403381,
+      "learning_rate": 4.5375e-05,
+      "loss": 1.8996,
+      "step": 185
+    },
+    {
+      "epoch": 0.2779100714385768,
+      "grad_norm": 1.259735345840454,
+      "learning_rate": 4.5350000000000005e-05,
+      "loss": 1.7165,
+      "step": 186
+    },
+    {
+      "epoch": 0.2794042116076014,
+      "grad_norm": 0.7863439321517944,
+      "learning_rate": 4.5325000000000004e-05,
+      "loss": 1.8585,
+      "step": 187
+    },
+    {
+      "epoch": 0.28089835177662603,
+      "grad_norm": 0.5054369568824768,
+      "learning_rate": 4.53e-05,
+      "loss": 1.9339,
+      "step": 188
+    },
+    {
+      "epoch": 0.28239249194565064,
+      "grad_norm": 0.5214366912841797,
+      "learning_rate": 4.5275e-05,
+      "loss": 1.6759,
+      "step": 189
+    },
+    {
+      "epoch": 0.28388663211467524,
+      "grad_norm": 0.48575031757354736,
+      "learning_rate": 4.525e-05,
+      "loss": 1.897,
+      "step": 190
+    },
+    {
+      "epoch": 0.28538077228369985,
+      "grad_norm": 0.4781228303909302,
+      "learning_rate": 4.5225e-05,
+      "loss": 1.7445,
+      "step": 191
+    },
+    {
+      "epoch": 0.28687491245272445,
+      "grad_norm": 0.5672656297683716,
+      "learning_rate": 4.52e-05,
+      "loss": 1.7473,
+      "step": 192
+    },
+    {
+      "epoch": 0.28836905262174906,
+      "grad_norm": 1.3074836730957031,
+      "learning_rate": 4.5175e-05,
+      "loss": 1.7179,
+      "step": 193
+    },
+    {
+      "epoch": 0.28986319279077366,
+      "grad_norm": 0.4941707253456116,
+      "learning_rate": 4.5150000000000006e-05,
+      "loss": 1.8377,
+      "step": 194
+    },
+    {
+      "epoch": 0.29135733295979827,
+      "grad_norm": 0.6247252225875854,
+      "learning_rate": 4.5125e-05,
+      "loss": 1.5313,
+      "step": 195
+    },
+    {
+      "epoch": 0.29285147312882287,
+      "grad_norm": 0.6214970946311951,
+      "learning_rate": 4.5100000000000005e-05,
+      "loss": 1.8108,
+      "step": 196
+    },
+    {
+      "epoch": 0.29434561329784753,
+      "grad_norm": 0.589272677898407,
+      "learning_rate": 4.5075e-05,
+      "loss": 1.6296,
+      "step": 197
+    },
+    {
+      "epoch": 0.29583975346687214,
+      "grad_norm": 0.4383275508880615,
+      "learning_rate": 4.5050000000000004e-05,
+      "loss": 1.7754,
+      "step": 198
+    },
+    {
+      "epoch": 0.29733389363589674,
+      "grad_norm": 0.4272954761981964,
+      "learning_rate": 4.5025000000000003e-05,
+      "loss": 1.6274,
+      "step": 199
+    },
+    {
+      "epoch": 0.29882803380492134,
+      "grad_norm": 0.714189887046814,
+      "learning_rate": 4.5e-05,
+      "loss": 1.8001,
+      "step": 200
+    },
+    {
+      "epoch": 0.30032217397394595,
+      "grad_norm": 0.7614679932594299,
+      "learning_rate": 4.4975e-05,
+      "loss": 1.6916,
+      "step": 201
+    },
+    {
+      "epoch": 0.30181631414297055,
+      "grad_norm": 0.5107477307319641,
+      "learning_rate": 4.495e-05,
+      "loss": 1.6286,
+      "step": 202
+    },
+    {
+      "epoch": 0.30331045431199516,
+      "grad_norm": 1.2998956441879272,
+      "learning_rate": 4.4925e-05,
+      "loss": 1.9156,
+      "step": 203
+    },
+    {
+      "epoch": 0.30480459448101976,
+      "grad_norm": 0.4685801863670349,
+      "learning_rate": 4.49e-05,
+      "loss": 1.7799,
+      "step": 204
+    },
+    {
+      "epoch": 0.30629873465004437,
+      "grad_norm": 0.7038085460662842,
+      "learning_rate": 4.4875e-05,
+      "loss": 1.7224,
+      "step": 205
+    },
+    {
+      "epoch": 0.307792874819069,
+      "grad_norm": 0.5876782536506653,
+      "learning_rate": 4.4850000000000006e-05,
+      "loss": 1.6782,
+      "step": 206
+    },
+    {
+      "epoch": 0.3092870149880936,
+      "grad_norm": 0.7861223220825195,
+      "learning_rate": 4.4825e-05,
+      "loss": 1.7567,
+      "step": 207
+    },
+    {
+      "epoch": 0.3107811551571182,
+      "grad_norm": 0.7584080100059509,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.7193,
+      "step": 208
+    },
+    {
+      "epoch": 0.3122752953261428,
+      "grad_norm": 0.6750642657279968,
+      "learning_rate": 4.4775e-05,
+      "loss": 1.6588,
+      "step": 209
+    },
+    {
+      "epoch": 0.3137694354951674,
+      "grad_norm": 0.6683038473129272,
+      "learning_rate": 4.4750000000000004e-05,
+      "loss": 1.6351,
+      "step": 210
+    },
+    {
+      "epoch": 0.315263575664192,
+      "grad_norm": 0.7034425735473633,
+      "learning_rate": 4.4725e-05,
+      "loss": 1.6917,
+      "step": 211
+    },
+    {
+      "epoch": 0.3167577158332166,
+      "grad_norm": 0.6255552172660828,
+      "learning_rate": 4.47e-05,
+      "loss": 1.7779,
+      "step": 212
+    },
+    {
+      "epoch": 0.3182518560022412,
+      "grad_norm": 0.7990907430648804,
+      "learning_rate": 4.4675e-05,
+      "loss": 1.6069,
+      "step": 213
+    },
+    {
+      "epoch": 0.3197459961712658,
+      "grad_norm": 1.1993693113327026,
+      "learning_rate": 4.465e-05,
+      "loss": 1.9121,
+      "step": 214
+    },
+    {
+      "epoch": 0.3212401363402904,
+      "grad_norm": 0.5611770749092102,
+      "learning_rate": 4.4625e-05,
+      "loss": 1.62,
+      "step": 215
+    },
+    {
+      "epoch": 0.322734276509315,
+      "grad_norm": 0.7897306680679321,
+      "learning_rate": 4.46e-05,
+      "loss": 1.5772,
+      "step": 216
+    },
+    {
+      "epoch": 0.3242284166783396,
+      "grad_norm": 0.7402398586273193,
+      "learning_rate": 4.4575e-05,
+      "loss": 1.7087,
+      "step": 217
+    },
+    {
+      "epoch": 0.32572255684736423,
+      "grad_norm": 0.5716530680656433,
+      "learning_rate": 4.4550000000000005e-05,
+      "loss": 1.6714,
+      "step": 218
+    },
+    {
+      "epoch": 0.32721669701638884,
+      "grad_norm": 0.6520130634307861,
+      "learning_rate": 4.4525e-05,
+      "loss": 1.6386,
+      "step": 219
+    },
+    {
+      "epoch": 0.32871083718541344,
+      "grad_norm": 0.5196349024772644,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 1.6034,
+      "step": 220
+    },
+    {
+      "epoch": 0.33020497735443805,
+      "grad_norm": 0.6718409657478333,
+      "learning_rate": 4.4475e-05,
+      "loss": 1.6676,
+      "step": 221
+    },
+    {
+      "epoch": 0.33169911752346265,
+      "grad_norm": 1.0844227075576782,
+      "learning_rate": 4.445e-05,
+      "loss": 1.6749,
+      "step": 222
+    },
+    {
+      "epoch": 0.33319325769248725,
+      "grad_norm": 0.6650180816650391,
+      "learning_rate": 4.4425e-05,
+      "loss": 1.6038,
+      "step": 223
+    },
+    {
+      "epoch": 0.33468739786151186,
+      "grad_norm": 0.5425866842269897,
+      "learning_rate": 4.44e-05,
+      "loss": 1.8488,
+      "step": 224
+    },
+    {
+      "epoch": 0.33618153803053646,
+      "grad_norm": 0.4744863510131836,
+      "learning_rate": 4.4375e-05,
+      "loss": 1.6235,
+      "step": 225
+    },
+    {
+      "epoch": 0.33767567819956107,
+      "grad_norm": 0.564362645149231,
+      "learning_rate": 4.435e-05,
+      "loss": 1.6729,
+      "step": 226
+    },
+    {
+      "epoch": 0.33916981836858573,
+      "grad_norm": 0.3389017879962921,
+      "learning_rate": 4.4325e-05,
+      "loss": 1.5262,
+      "step": 227
+    },
+    {
+      "epoch": 0.34066395853761033,
+      "grad_norm": 0.7532176971435547,
+      "learning_rate": 4.43e-05,
+      "loss": 1.7369,
+      "step": 228
+    },
+    {
+      "epoch": 0.34215809870663494,
+      "grad_norm": 0.6776916980743408,
+      "learning_rate": 4.4275e-05,
+      "loss": 1.7183,
+      "step": 229
+    },
+    {
+      "epoch": 0.34365223887565954,
+      "grad_norm": 0.5167742371559143,
+      "learning_rate": 4.4250000000000005e-05,
+      "loss": 1.7791,
+      "step": 230
+    },
+    {
+      "epoch": 0.34514637904468415,
+      "grad_norm": 0.6937631368637085,
+      "learning_rate": 4.4225e-05,
+      "loss": 1.7255,
+      "step": 231
+    },
+    {
+      "epoch": 0.34664051921370875,
+      "grad_norm": 0.5599237084388733,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 1.6517,
+      "step": 232
+    },
+    {
+      "epoch": 0.34813465938273336,
+      "grad_norm": 0.6251773238182068,
+      "learning_rate": 4.4174999999999996e-05,
+      "loss": 1.6929,
+      "step": 233
+    },
+    {
+      "epoch": 0.34962879955175796,
+      "grad_norm": 0.9755148887634277,
+      "learning_rate": 4.415e-05,
+      "loss": 1.6177,
+      "step": 234
+    },
+    {
+      "epoch": 0.35112293972078257,
+      "grad_norm": 0.951604962348938,
+      "learning_rate": 4.4125e-05,
+      "loss": 1.5687,
+      "step": 235
+    },
+    {
+      "epoch": 0.3526170798898072,
+      "grad_norm": 0.45756927132606506,
+      "learning_rate": 4.41e-05,
+      "loss": 1.6375,
+      "step": 236
+    },
+    {
+      "epoch": 0.3541112200588318,
+      "grad_norm": 0.4150543808937073,
+      "learning_rate": 4.4075e-05,
+      "loss": 1.5537,
+      "step": 237
+    },
+    {
+      "epoch": 0.3556053602278564,
+      "grad_norm": 0.4883075952529907,
+      "learning_rate": 4.405e-05,
+      "loss": 1.6506,
+      "step": 238
+    },
+    {
+      "epoch": 0.357099500396881,
+      "grad_norm": 0.5154853463172913,
+      "learning_rate": 4.4025e-05,
+      "loss": 1.7254,
+      "step": 239
+    },
+    {
+      "epoch": 0.3585936405659056,
+      "grad_norm": 1.2003599405288696,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.6396,
+      "step": 240
+    },
+    {
+      "epoch": 0.3600877807349302,
+      "grad_norm": 0.5465105175971985,
+      "learning_rate": 4.3975e-05,
+      "loss": 1.5758,
+      "step": 241
+    },
+    {
+      "epoch": 0.3615819209039548,
+      "grad_norm": 0.5422613620758057,
+      "learning_rate": 4.3950000000000004e-05,
+      "loss": 1.5582,
+      "step": 242
+    },
+    {
+      "epoch": 0.3630760610729794,
+      "grad_norm": 0.44329527020454407,
+      "learning_rate": 4.3925e-05,
+      "loss": 1.6287,
+      "step": 243
+    },
+    {
+      "epoch": 0.364570201242004,
+      "grad_norm": 0.7963645458221436,
+      "learning_rate": 4.39e-05,
+      "loss": 1.736,
+      "step": 244
+    },
+    {
+      "epoch": 0.3660643414110286,
+      "grad_norm": 0.490032434463501,
+      "learning_rate": 4.3875e-05,
+      "loss": 1.6711,
+      "step": 245
+    },
+    {
+      "epoch": 0.3675584815800532,
+      "grad_norm": 1.1757748126983643,
+      "learning_rate": 4.385e-05,
+      "loss": 1.6462,
+      "step": 246
+    },
+    {
+      "epoch": 0.3690526217490778,
+      "grad_norm": 0.39032885432243347,
+      "learning_rate": 4.3825e-05,
+      "loss": 1.7074,
+      "step": 247
+    },
+    {
+      "epoch": 0.37054676191810243,
+      "grad_norm": 0.5092072486877441,
+      "learning_rate": 4.38e-05,
+      "loss": 1.4719,
+      "step": 248
+    },
+    {
+      "epoch": 0.37204090208712703,
+      "grad_norm": 0.6537796258926392,
+      "learning_rate": 4.3775e-05,
+      "loss": 1.7254,
+      "step": 249
+    },
+    {
+      "epoch": 0.37353504225615164,
+      "grad_norm": 0.9034501314163208,
+      "learning_rate": 4.375e-05,
+      "loss": 1.5082,
+      "step": 250
+    },
+    {
+      "epoch": 0.37502918242517624,
+      "grad_norm": 1.2041537761688232,
+      "learning_rate": 4.3725000000000006e-05,
+      "loss": 1.8014,
+      "step": 251
+    },
+    {
+      "epoch": 0.37652332259420085,
+      "grad_norm": 0.6997256278991699,
+      "learning_rate": 4.3700000000000005e-05,
+      "loss": 1.7564,
+      "step": 252
+    },
+    {
+      "epoch": 0.37801746276322545,
+      "grad_norm": 0.6815714240074158,
+      "learning_rate": 4.3675000000000005e-05,
+      "loss": 1.5752,
+      "step": 253
+    },
+    {
+      "epoch": 0.37951160293225006,
+      "grad_norm": 2.3094871044158936,
+      "learning_rate": 4.3650000000000004e-05,
+      "loss": 1.6464,
+      "step": 254
+    },
+    {
+      "epoch": 0.38100574310127466,
+      "grad_norm": 0.5100992918014526,
+      "learning_rate": 4.3625e-05,
+      "loss": 1.8269,
+      "step": 255
+    },
+    {
+      "epoch": 0.38249988327029927,
+      "grad_norm": 0.4920763373374939,
+      "learning_rate": 4.36e-05,
+      "loss": 1.5146,
+      "step": 256
+    },
+    {
+      "epoch": 0.38399402343932393,
+      "grad_norm": 0.5346136093139648,
+      "learning_rate": 4.3575e-05,
+      "loss": 1.6748,
+      "step": 257
+    },
+    {
+      "epoch": 0.38548816360834853,
+      "grad_norm": 0.5572606921195984,
+      "learning_rate": 4.355e-05,
+      "loss": 1.6669,
+      "step": 258
+    },
+    {
+      "epoch": 0.38698230377737314,
+      "grad_norm": 0.344314843416214,
+      "learning_rate": 4.352500000000001e-05,
+      "loss": 1.6193,
+      "step": 259
+    },
+    {
+      "epoch": 0.38847644394639774,
+      "grad_norm": 0.47694844007492065,
+      "learning_rate": 4.35e-05,
+      "loss": 1.6713,
+      "step": 260
+    },
+    {
+      "epoch": 0.38997058411542235,
+      "grad_norm": 0.5320355892181396,
+      "learning_rate": 4.3475000000000006e-05,
+      "loss": 1.5918,
+      "step": 261
+    },
+    {
+      "epoch": 0.39146472428444695,
+      "grad_norm": 0.5045621395111084,
+      "learning_rate": 4.345e-05,
+      "loss": 1.5785,
+      "step": 262
+    },
+    {
+      "epoch": 0.39295886445347156,
+      "grad_norm": 0.48489850759506226,
+      "learning_rate": 4.3425000000000005e-05,
+      "loss": 1.8115,
+      "step": 263
+    },
+    {
+      "epoch": 0.39445300462249616,
+      "grad_norm": 0.8452106714248657,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 1.6116,
+      "step": 264
+    },
+    {
+      "epoch": 0.39594714479152077,
+      "grad_norm": 0.6507668495178223,
+      "learning_rate": 4.3375000000000004e-05,
+      "loss": 1.5557,
+      "step": 265
+    },
+    {
+      "epoch": 0.39744128496054537,
+      "grad_norm": 0.5312817692756653,
+      "learning_rate": 4.335e-05,
+      "loss": 1.5833,
+      "step": 266
+    },
+    {
+      "epoch": 0.39893542512957,
+      "grad_norm": 1.0674476623535156,
+      "learning_rate": 4.3325e-05,
+      "loss": 1.6319,
+      "step": 267
+    },
+    {
+      "epoch": 0.4004295652985946,
+      "grad_norm": 0.4685145616531372,
+      "learning_rate": 4.33e-05,
+      "loss": 1.6653,
+      "step": 268
+    },
+    {
+      "epoch": 0.4019237054676192,
+      "grad_norm": 0.6625531315803528,
+      "learning_rate": 4.3275e-05,
+      "loss": 1.5749,
+      "step": 269
+    },
+    {
+      "epoch": 0.4034178456366438,
+      "grad_norm": 0.46707406640052795,
+      "learning_rate": 4.325e-05,
+      "loss": 1.4153,
+      "step": 270
+    },
+    {
+      "epoch": 0.4049119858056684,
+      "grad_norm": 0.7175610065460205,
+      "learning_rate": 4.322500000000001e-05,
+      "loss": 1.6583,
+      "step": 271
+    },
+    {
+      "epoch": 0.406406125974693,
+      "grad_norm": 1.0776153802871704,
+      "learning_rate": 4.32e-05,
+      "loss": 1.6923,
+      "step": 272
+    },
+    {
+      "epoch": 0.4079002661437176,
+      "grad_norm": 0.453715443611145,
+      "learning_rate": 4.3175000000000006e-05,
+      "loss": 1.6136,
+      "step": 273
+    },
+    {
+      "epoch": 0.4093944063127422,
+      "grad_norm": 0.5465681552886963,
+      "learning_rate": 4.315e-05,
+      "loss": 1.5902,
+      "step": 274
+    },
+    {
+      "epoch": 0.4108885464817668,
+      "grad_norm": 0.49540913105010986,
+      "learning_rate": 4.3125000000000005e-05,
+      "loss": 1.4397,
+      "step": 275
+    },
+    {
+      "epoch": 0.4123826866507914,
+      "grad_norm": 0.39481091499328613,
+      "learning_rate": 4.3100000000000004e-05,
+      "loss": 1.4795,
+      "step": 276
+    },
+    {
+      "epoch": 0.413876826819816,
+      "grad_norm": 0.8381078839302063,
+      "learning_rate": 4.3075000000000003e-05,
+      "loss": 1.5042,
+      "step": 277
+    },
+    {
+      "epoch": 0.41537096698884063,
+      "grad_norm": 0.366020143032074,
+      "learning_rate": 4.305e-05,
+      "loss": 1.4619,
+      "step": 278
+    },
+    {
+      "epoch": 0.41686510715786523,
+      "grad_norm": 0.3974578082561493,
+      "learning_rate": 4.3025e-05,
+      "loss": 1.6508,
+      "step": 279
+    },
+    {
+      "epoch": 0.41835924732688984,
+      "grad_norm": 0.38582220673561096,
+      "learning_rate": 4.3e-05,
+      "loss": 1.7035,
+      "step": 280
+    },
+    {
+      "epoch": 0.41985338749591444,
+      "grad_norm": 0.5376545786857605,
+      "learning_rate": 4.2975e-05,
+      "loss": 1.5625,
+      "step": 281
+    },
+    {
+      "epoch": 0.42134752766493905,
+      "grad_norm": 0.3950735032558441,
+      "learning_rate": 4.295e-05,
+      "loss": 1.58,
+      "step": 282
+    },
+    {
+      "epoch": 0.42284166783396365,
+      "grad_norm": 0.496115505695343,
+      "learning_rate": 4.2925000000000007e-05,
+      "loss": 1.7686,
+      "step": 283
+    },
+    {
+      "epoch": 0.42433580800298826,
+      "grad_norm": 0.5910571813583374,
+      "learning_rate": 4.29e-05,
+      "loss": 1.6702,
+      "step": 284
+    },
+    {
+      "epoch": 0.42582994817201286,
+      "grad_norm": 0.3945789039134979,
+      "learning_rate": 4.2875000000000005e-05,
+      "loss": 1.7765,
+      "step": 285
+    },
+    {
+      "epoch": 0.42732408834103747,
+      "grad_norm": 0.5639134645462036,
+      "learning_rate": 4.285e-05,
+      "loss": 1.4734,
+      "step": 286
+    },
+    {
+      "epoch": 0.4288182285100621,
+      "grad_norm": 0.3879479765892029,
+      "learning_rate": 4.2825000000000004e-05,
+      "loss": 1.4693,
+      "step": 287
+    },
+    {
+      "epoch": 0.43031236867908673,
+      "grad_norm": 0.5454258322715759,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 1.3784,
+      "step": 288
+    },
+    {
+      "epoch": 0.43180650884811134,
+      "grad_norm": 0.4519095718860626,
+      "learning_rate": 4.2775e-05,
+      "loss": 1.6286,
+      "step": 289
+    },
+    {
+      "epoch": 0.43330064901713594,
+      "grad_norm": 0.49812281131744385,
+      "learning_rate": 4.275e-05,
+      "loss": 1.6295,
+      "step": 290
+    },
+    {
+      "epoch": 0.43479478918616055,
+      "grad_norm": 0.5681044459342957,
+      "learning_rate": 4.2725e-05,
+      "loss": 1.5457,
+      "step": 291
+    },
+    {
+      "epoch": 0.43628892935518515,
+      "grad_norm": 0.3775969445705414,
+      "learning_rate": 4.27e-05,
+      "loss": 1.4316,
+      "step": 292
+    },
+    {
+      "epoch": 0.43778306952420976,
+      "grad_norm": 0.5119141340255737,
+      "learning_rate": 4.2675e-05,
+      "loss": 1.5647,
+      "step": 293
+    },
+    {
+      "epoch": 0.43927720969323436,
+      "grad_norm": 0.5567271113395691,
+      "learning_rate": 4.265e-05,
+      "loss": 1.6077,
+      "step": 294
+    },
+    {
+      "epoch": 0.44077134986225897,
+      "grad_norm": 2.239450216293335,
+      "learning_rate": 4.2625000000000006e-05,
+      "loss": 1.4764,
+      "step": 295
+    },
+    {
+      "epoch": 0.44226549003128357,
+      "grad_norm": 0.41082441806793213,
+      "learning_rate": 4.26e-05,
+      "loss": 1.6279,
+      "step": 296
+    },
+    {
+      "epoch": 0.4437596302003082,
+      "grad_norm": 0.3343771994113922,
+      "learning_rate": 4.2575000000000005e-05,
+      "loss": 1.5487,
+      "step": 297
+    },
+    {
+      "epoch": 0.4452537703693328,
+      "grad_norm": 0.35104796290397644,
+      "learning_rate": 4.2550000000000004e-05,
+      "loss": 1.5431,
+      "step": 298
+    },
+    {
+      "epoch": 0.4467479105383574,
+      "grad_norm": 0.7141225337982178,
+      "learning_rate": 4.2525000000000004e-05,
+      "loss": 1.5845,
+      "step": 299
+    },
+    {
+      "epoch": 0.448242050707382,
+      "grad_norm": 1.133215069770813,
+      "learning_rate": 4.25e-05,
+      "loss": 1.5753,
+      "step": 300
+    },
+    {
+      "epoch": 0.4497361908764066,
+      "grad_norm": 0.419767826795578,
+      "learning_rate": 4.2475e-05,
+      "loss": 1.5059,
+      "step": 301
+    },
+    {
+      "epoch": 0.4512303310454312,
+      "grad_norm": 0.3364299535751343,
+      "learning_rate": 4.245e-05,
+      "loss": 1.5082,
+      "step": 302
+    },
+    {
+      "epoch": 0.4527244712144558,
+      "grad_norm": 2.21221661567688,
+      "learning_rate": 4.2425e-05,
+      "loss": 1.7031,
+      "step": 303
+    },
+    {
+      "epoch": 0.4542186113834804,
+      "grad_norm": 1.6208406686782837,
+      "learning_rate": 4.24e-05,
+      "loss": 1.657,
+      "step": 304
+    },
+    {
+      "epoch": 0.455712751552505,
+      "grad_norm": 0.6496986150741577,
+      "learning_rate": 4.237500000000001e-05,
+      "loss": 1.8416,
+      "step": 305
+    },
+    {
+      "epoch": 0.4572068917215296,
+      "grad_norm": 0.3318963050842285,
+      "learning_rate": 4.235e-05,
+      "loss": 1.7409,
+      "step": 306
+    },
+    {
+      "epoch": 0.4587010318905542,
+      "grad_norm": 0.5762292742729187,
+      "learning_rate": 4.2325000000000006e-05,
+      "loss": 1.5737,
+      "step": 307
+    },
+    {
+      "epoch": 0.4601951720595788,
+      "grad_norm": 0.7929126024246216,
+      "learning_rate": 4.23e-05,
+      "loss": 1.6096,
+      "step": 308
+    },
+    {
+      "epoch": 0.46168931222860343,
+      "grad_norm": 0.3540865480899811,
+      "learning_rate": 4.2275000000000004e-05,
+      "loss": 1.4596,
+      "step": 309
+    },
+    {
+      "epoch": 0.46318345239762804,
+      "grad_norm": 0.3595794141292572,
+      "learning_rate": 4.2250000000000004e-05,
+      "loss": 1.6072,
+      "step": 310
+    },
+    {
+      "epoch": 0.46467759256665264,
+      "grad_norm": 0.8183721899986267,
+      "learning_rate": 4.2225e-05,
+      "loss": 1.4945,
+      "step": 311
+    },
+    {
+      "epoch": 0.46617173273567725,
+      "grad_norm": 0.36932075023651123,
+      "learning_rate": 4.22e-05,
+      "loss": 1.6256,
+      "step": 312
+    },
+    {
+      "epoch": 0.46766587290470185,
+      "grad_norm": 1.8945622444152832,
+      "learning_rate": 4.2175e-05,
+      "loss": 1.7877,
+      "step": 313
+    },
+    {
+      "epoch": 0.46916001307372646,
+      "grad_norm": 0.4660819172859192,
+      "learning_rate": 4.215e-05,
+      "loss": 1.3404,
+      "step": 314
+    },
+    {
+      "epoch": 0.47065415324275106,
+      "grad_norm": 0.3449573814868927,
+      "learning_rate": 4.2125e-05,
+      "loss": 1.4818,
+      "step": 315
+    },
+    {
+      "epoch": 0.47214829341177567,
+      "grad_norm": 0.5242890119552612,
+      "learning_rate": 4.21e-05,
+      "loss": 1.6351,
+      "step": 316
+    },
+    {
+      "epoch": 0.4736424335808003,
+      "grad_norm": 0.709052324295044,
+      "learning_rate": 4.2075000000000006e-05,
+      "loss": 1.4658,
+      "step": 317
+    },
+    {
+      "epoch": 0.47513657374982493,
+      "grad_norm": 0.4316338300704956,
+      "learning_rate": 4.205e-05,
+      "loss": 1.4741,
+      "step": 318
+    },
+    {
+      "epoch": 0.47663071391884954,
+      "grad_norm": 0.6699191331863403,
+      "learning_rate": 4.2025000000000005e-05,
+      "loss": 1.5725,
+      "step": 319
+    },
+    {
+      "epoch": 0.47812485408787414,
+      "grad_norm": 0.5343687534332275,
+      "learning_rate": 4.2e-05,
+      "loss": 1.5691,
+      "step": 320
+    },
+    {
+      "epoch": 0.47961899425689875,
+      "grad_norm": 0.35519710183143616,
+      "learning_rate": 4.1975000000000004e-05,
+      "loss": 1.2731,
+      "step": 321
+    },
+    {
+      "epoch": 0.48111313442592335,
+      "grad_norm": 0.385275274515152,
+      "learning_rate": 4.195e-05,
+      "loss": 1.345,
+      "step": 322
+    },
+    {
+      "epoch": 0.48260727459494795,
+      "grad_norm": 0.6153557300567627,
+      "learning_rate": 4.1925e-05,
+      "loss": 1.5702,
+      "step": 323
+    },
+    {
+      "epoch": 0.48410141476397256,
+      "grad_norm": 0.6316625475883484,
+      "learning_rate": 4.19e-05,
+      "loss": 1.4381,
+      "step": 324
+    },
+    {
+      "epoch": 0.48559555493299716,
+      "grad_norm": 2.4540088176727295,
+      "learning_rate": 4.1875e-05,
+      "loss": 1.3345,
+      "step": 325
+    },
+    {
+      "epoch": 0.48708969510202177,
+      "grad_norm": 0.3455633819103241,
+      "learning_rate": 4.185e-05,
+      "loss": 1.2002,
+      "step": 326
+    },
+    {
+      "epoch": 0.4885838352710464,
+      "grad_norm": 0.6012340188026428,
+      "learning_rate": 4.1825e-05,
+      "loss": 1.5849,
+      "step": 327
+    },
+    {
+      "epoch": 0.490077975440071,
+      "grad_norm": 0.6323389410972595,
+      "learning_rate": 4.18e-05,
+      "loss": 1.5271,
+      "step": 328
+    },
+    {
+      "epoch": 0.4915721156090956,
+      "grad_norm": 0.5738844871520996,
+      "learning_rate": 4.1775000000000006e-05,
+      "loss": 1.6768,
+      "step": 329
+    },
+    {
+      "epoch": 0.4930662557781202,
+      "grad_norm": 0.5335976481437683,
+      "learning_rate": 4.175e-05,
+      "loss": 1.625,
+      "step": 330
+    },
+    {
+      "epoch": 0.4945603959471448,
+      "grad_norm": 0.40596872568130493,
+      "learning_rate": 4.1725000000000005e-05,
+      "loss": 1.573,
+      "step": 331
+    },
+    {
+      "epoch": 0.4960545361161694,
+      "grad_norm": 0.46490761637687683,
+      "learning_rate": 4.17e-05,
+      "loss": 1.3669,
+      "step": 332
+    },
+    {
+      "epoch": 0.497548676285194,
+      "grad_norm": 0.42321160435676575,
+      "learning_rate": 4.1675e-05,
+      "loss": 1.4261,
+      "step": 333
+    },
+    {
+      "epoch": 0.4990428164542186,
+      "grad_norm": 0.8209381699562073,
+      "learning_rate": 4.165e-05,
+      "loss": 1.3413,
+      "step": 334
+    },
+    {
+      "epoch": 0.5005369566232433,
+      "grad_norm": 0.3648524880409241,
+      "learning_rate": 4.1625e-05,
+      "loss": 1.3683,
+      "step": 335
+    },
+    {
+      "epoch": 0.5020310967922679,
+      "grad_norm": 0.4434033930301666,
+      "learning_rate": 4.16e-05,
+      "loss": 1.5501,
+      "step": 336
+    },
+    {
+      "epoch": 0.5035252369612925,
+      "grad_norm": 0.44914835691452026,
+      "learning_rate": 4.1575e-05,
+      "loss": 1.7285,
+      "step": 337
+    },
+    {
+      "epoch": 0.5050193771303171,
+      "grad_norm": 0.4023708701133728,
+      "learning_rate": 4.155e-05,
+      "loss": 1.4152,
+      "step": 338
+    },
+    {
+      "epoch": 0.5065135172993417,
+      "grad_norm": 0.39679622650146484,
+      "learning_rate": 4.1525e-05,
+      "loss": 1.5925,
+      "step": 339
+    },
+    {
+      "epoch": 0.5080076574683663,
+      "grad_norm": 0.41201382875442505,
+      "learning_rate": 4.15e-05,
+      "loss": 1.4728,
+      "step": 340
+    },
+    {
+      "epoch": 0.5095017976373909,
+      "grad_norm": 0.37450897693634033,
+      "learning_rate": 4.1475000000000005e-05,
+      "loss": 1.3821,
+      "step": 341
+    },
+    {
+      "epoch": 0.5109959378064155,
+      "grad_norm": 3.6442813873291016,
+      "learning_rate": 4.145e-05,
+      "loss": 1.4883,
+      "step": 342
+    },
+    {
+      "epoch": 0.5124900779754401,
+      "grad_norm": 0.529839813709259,
+      "learning_rate": 4.1425000000000004e-05,
+      "loss": 1.5838,
+      "step": 343
+    },
+    {
+      "epoch": 0.5139842181444647,
+      "grad_norm": 0.47011345624923706,
+      "learning_rate": 4.14e-05,
+      "loss": 1.4973,
+      "step": 344
+    },
+    {
+      "epoch": 0.5154783583134893,
+      "grad_norm": 0.7854222059249878,
+      "learning_rate": 4.1375e-05,
+      "loss": 1.526,
+      "step": 345
+    },
+    {
+      "epoch": 0.5169724984825139,
+      "grad_norm": 0.3422839641571045,
+      "learning_rate": 4.135e-05,
+      "loss": 1.7585,
+      "step": 346
+    },
+    {
+      "epoch": 0.5184666386515385,
+      "grad_norm": 0.3906049132347107,
+      "learning_rate": 4.1325e-05,
+      "loss": 1.5914,
+      "step": 347
+    },
+    {
+      "epoch": 0.5199607788205631,
+      "grad_norm": 0.8085627555847168,
+      "learning_rate": 4.13e-05,
+      "loss": 1.5909,
+      "step": 348
+    },
+    {
+      "epoch": 0.5214549189895877,
+      "grad_norm": 0.571151614189148,
+      "learning_rate": 4.1275e-05,
+      "loss": 1.5494,
+      "step": 349
+    },
+    {
+      "epoch": 0.5229490591586123,
+      "grad_norm": 1.0025105476379395,
+      "learning_rate": 4.125e-05,
+      "loss": 1.7711,
+      "step": 350
+    },
+    {
+      "epoch": 0.5244431993276369,
+      "grad_norm": 0.361357182264328,
+      "learning_rate": 4.1225e-05,
+      "loss": 1.494,
+      "step": 351
+    },
+    {
+      "epoch": 0.5259373394966615,
+      "grad_norm": 0.3900647759437561,
+      "learning_rate": 4.12e-05,
+      "loss": 1.5864,
+      "step": 352
+    },
+    {
+      "epoch": 0.5274314796656862,
+      "grad_norm": 0.4274671673774719,
+      "learning_rate": 4.1175000000000005e-05,
+      "loss": 1.4905,
+      "step": 353
+    },
+    {
+      "epoch": 0.5289256198347108,
+      "grad_norm": 0.3727322220802307,
+      "learning_rate": 4.115e-05,
+      "loss": 1.5137,
+      "step": 354
+    },
+    {
+      "epoch": 0.5304197600037354,
+      "grad_norm": 0.49744248390197754,
+      "learning_rate": 4.1125000000000004e-05,
+      "loss": 1.509,
+      "step": 355
+    },
+    {
+      "epoch": 0.53191390017276,
+      "grad_norm": 0.39131686091423035,
+      "learning_rate": 4.11e-05,
+      "loss": 1.5525,
+      "step": 356
+    },
+    {
+      "epoch": 0.5334080403417846,
+      "grad_norm": 1.4326519966125488,
+      "learning_rate": 4.1075e-05,
+      "loss": 1.5982,
+      "step": 357
+    },
+    {
+      "epoch": 0.5349021805108092,
+      "grad_norm": 0.4607619047164917,
+      "learning_rate": 4.105e-05,
+      "loss": 1.3984,
+      "step": 358
+    },
+    {
+      "epoch": 0.5363963206798338,
+      "grad_norm": 0.5902805328369141,
+      "learning_rate": 4.1025e-05,
+      "loss": 1.4801,
+      "step": 359
+    },
+    {
+      "epoch": 0.5378904608488584,
+      "grad_norm": 0.34372955560684204,
+      "learning_rate": 4.1e-05,
+      "loss": 1.4714,
+      "step": 360
+    },
+    {
+      "epoch": 0.539384601017883,
+      "grad_norm": 0.32892340421676636,
+      "learning_rate": 4.0975e-05,
+      "loss": 1.5977,
+      "step": 361
+    },
+    {
+      "epoch": 0.5408787411869076,
+      "grad_norm": 0.45786258578300476,
+      "learning_rate": 4.095e-05,
+      "loss": 1.6396,
+      "step": 362
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.6414769887924194,
+      "learning_rate": 4.0925000000000005e-05,
+      "loss": 1.3731,
+      "step": 363
+    },
+    {
+      "epoch": 0.5438670215249568,
+      "grad_norm": 0.4823474586009979,
+      "learning_rate": 4.09e-05,
+      "loss": 1.4974,
+      "step": 364
+    },
+    {
+      "epoch": 0.5453611616939814,
+      "grad_norm": 1.1086201667785645,
+      "learning_rate": 4.0875000000000004e-05,
+      "loss": 1.6457,
+      "step": 365
+    },
+    {
+      "epoch": 0.546855301863006,
+      "grad_norm": 0.9799003005027771,
+      "learning_rate": 4.085e-05,
+      "loss": 1.4262,
+      "step": 366
+    },
+    {
+      "epoch": 0.5483494420320306,
+      "grad_norm": 0.5253656506538391,
+      "learning_rate": 4.0825e-05,
+      "loss": 1.3437,
+      "step": 367
+    },
+    {
+      "epoch": 0.5498435822010552,
+      "grad_norm": 0.6868051290512085,
+      "learning_rate": 4.08e-05,
+      "loss": 1.464,
+      "step": 368
+    },
+    {
+      "epoch": 0.5513377223700798,
+      "grad_norm": 0.33976659178733826,
+      "learning_rate": 4.0775e-05,
+      "loss": 1.693,
+      "step": 369
+    },
+    {
+      "epoch": 0.5528318625391044,
+      "grad_norm": 0.3966132700443268,
+      "learning_rate": 4.075e-05,
+      "loss": 1.6771,
+      "step": 370
+    },
+    {
+      "epoch": 0.554326002708129,
+      "grad_norm": 0.39192917943000793,
+      "learning_rate": 4.0725e-05,
+      "loss": 1.5507,
+      "step": 371
+    },
+    {
+      "epoch": 0.5558201428771536,
+      "grad_norm": 1.481609582901001,
+      "learning_rate": 4.07e-05,
+      "loss": 1.4611,
+      "step": 372
+    },
+    {
+      "epoch": 0.5573142830461782,
+      "grad_norm": 0.6329764723777771,
+      "learning_rate": 4.0675e-05,
+      "loss": 1.542,
+      "step": 373
+    },
+    {
+      "epoch": 0.5588084232152029,
+      "grad_norm": 0.41724011301994324,
+      "learning_rate": 4.065e-05,
+      "loss": 1.6033,
+      "step": 374
+    },
+    {
+      "epoch": 0.5603025633842275,
+      "grad_norm": 0.32675087451934814,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 1.5204,
+      "step": 375
+    },
+    {
+      "epoch": 0.5617967035532521,
+      "grad_norm": 0.35794204473495483,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 1.4091,
+      "step": 376
+    },
+    {
+      "epoch": 0.5632908437222767,
+      "grad_norm": 0.6309773921966553,
+      "learning_rate": 4.0575000000000004e-05,
+      "loss": 1.4087,
+      "step": 377
+    },
+    {
+      "epoch": 0.5647849838913013,
+      "grad_norm": 0.4585449993610382,
+      "learning_rate": 4.055e-05,
+      "loss": 1.5238,
+      "step": 378
+    },
+    {
+      "epoch": 0.5662791240603259,
+      "grad_norm": 0.3686904311180115,
+      "learning_rate": 4.0525e-05,
+      "loss": 1.5567,
+      "step": 379
+    },
+    {
+      "epoch": 0.5677732642293505,
+      "grad_norm": 0.4154909551143646,
+      "learning_rate": 4.05e-05,
+      "loss": 1.4935,
+      "step": 380
+    },
+    {
+      "epoch": 0.5692674043983751,
+      "grad_norm": 0.302909255027771,
+      "learning_rate": 4.0475e-05,
+      "loss": 1.5521,
+      "step": 381
+    },
+    {
+      "epoch": 0.5707615445673997,
+      "grad_norm": 0.40441974997520447,
+      "learning_rate": 4.045000000000001e-05,
+      "loss": 1.6756,
+      "step": 382
+    },
+    {
+      "epoch": 0.5722556847364243,
+      "grad_norm": 0.4657610058784485,
+      "learning_rate": 4.0425e-05,
+      "loss": 1.5313,
+      "step": 383
+    },
+    {
+      "epoch": 0.5737498249054489,
+      "grad_norm": 1.0451929569244385,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 1.5619,
+      "step": 384
+    },
+    {
+      "epoch": 0.5752439650744735,
+      "grad_norm": 0.4887607991695404,
+      "learning_rate": 4.0375e-05,
+      "loss": 1.5529,
+      "step": 385
+    },
+    {
+      "epoch": 0.5767381052434981,
+      "grad_norm": 0.4439973533153534,
+      "learning_rate": 4.0350000000000005e-05,
+      "loss": 1.5859,
+      "step": 386
+    },
+    {
+      "epoch": 0.5782322454125227,
+      "grad_norm": 0.36021965742111206,
+      "learning_rate": 4.0325000000000004e-05,
+      "loss": 1.5467,
+      "step": 387
+    },
+    {
+      "epoch": 0.5797263855815473,
+      "grad_norm": 0.9734621047973633,
+      "learning_rate": 4.0300000000000004e-05,
+      "loss": 1.6785,
+      "step": 388
+    },
+    {
+      "epoch": 0.5812205257505719,
+      "grad_norm": 0.4874991774559021,
+      "learning_rate": 4.0275e-05,
+      "loss": 1.462,
+      "step": 389
+    },
+    {
+      "epoch": 0.5827146659195965,
+      "grad_norm": 0.3743913769721985,
+      "learning_rate": 4.025e-05,
+      "loss": 1.4352,
+      "step": 390
+    },
+    {
+      "epoch": 0.5842088060886211,
+      "grad_norm": 0.5307121276855469,
+      "learning_rate": 4.0225e-05,
+      "loss": 1.5183,
+      "step": 391
+    },
+    {
+      "epoch": 0.5857029462576457,
+      "grad_norm": 0.616095244884491,
+      "learning_rate": 4.02e-05,
+      "loss": 1.4568,
+      "step": 392
+    },
+    {
+      "epoch": 0.5871970864266705,
+      "grad_norm": 0.7378297448158264,
+      "learning_rate": 4.0175e-05,
+      "loss": 1.6607,
+      "step": 393
+    },
+    {
+      "epoch": 0.5886912265956951,
+      "grad_norm": 0.38973790407180786,
+      "learning_rate": 4.015000000000001e-05,
+      "loss": 1.5029,
+      "step": 394
+    },
+    {
+      "epoch": 0.5901853667647197,
+      "grad_norm": 0.7084895968437195,
+      "learning_rate": 4.0125e-05,
+      "loss": 1.6412,
+      "step": 395
+    },
+    {
+      "epoch": 0.5916795069337443,
+      "grad_norm": 0.615085244178772,
+      "learning_rate": 4.0100000000000006e-05,
+      "loss": 1.48,
+      "step": 396
+    },
+    {
+      "epoch": 0.5931736471027689,
+      "grad_norm": 0.5237393379211426,
+      "learning_rate": 4.0075e-05,
+      "loss": 1.5518,
+      "step": 397
+    },
+    {
+      "epoch": 0.5946677872717935,
+      "grad_norm": 0.5331382155418396,
+      "learning_rate": 4.0050000000000004e-05,
+      "loss": 1.5439,
+      "step": 398
+    },
+    {
+      "epoch": 0.5961619274408181,
+      "grad_norm": 0.36154451966285706,
+      "learning_rate": 4.0025000000000004e-05,
+      "loss": 1.7163,
+      "step": 399
+    },
+    {
+      "epoch": 0.5976560676098427,
+      "grad_norm": 0.3548503816127777,
+      "learning_rate": 4e-05,
+      "loss": 1.4117,
+      "step": 400
+    },
+    {
+      "epoch": 0.5991502077788673,
+      "grad_norm": 0.5904132127761841,
+      "learning_rate": 3.9975e-05,
+      "loss": 1.4805,
+      "step": 401
+    },
+    {
+      "epoch": 0.6006443479478919,
+      "grad_norm": 0.2864095866680145,
+      "learning_rate": 3.995e-05,
+      "loss": 1.4976,
+      "step": 402
+    },
+    {
+      "epoch": 0.6021384881169165,
+      "grad_norm": 1.153709888458252,
+      "learning_rate": 3.9925e-05,
+      "loss": 1.3897,
+      "step": 403
+    },
+    {
+      "epoch": 0.6036326282859411,
+      "grad_norm": 0.48970621824264526,
+      "learning_rate": 3.99e-05,
+      "loss": 1.5441,
+      "step": 404
+    },
+    {
+      "epoch": 0.6051267684549657,
+      "grad_norm": 0.8419874906539917,
+      "learning_rate": 3.9875e-05,
+      "loss": 1.5113,
+      "step": 405
+    },
+    {
+      "epoch": 0.6066209086239903,
+      "grad_norm": 0.429151713848114,
+      "learning_rate": 3.9850000000000006e-05,
+      "loss": 1.5316,
+      "step": 406
+    },
+    {
+      "epoch": 0.6081150487930149,
+      "grad_norm": 0.24945476651191711,
+      "learning_rate": 3.9825e-05,
+      "loss": 1.3466,
+      "step": 407
+    },
+    {
+      "epoch": 0.6096091889620395,
+      "grad_norm": 0.8487308621406555,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 1.5499,
+      "step": 408
+    },
+    {
+      "epoch": 0.6111033291310641,
+      "grad_norm": 0.5790063738822937,
+      "learning_rate": 3.9775e-05,
+      "loss": 1.5398,
+      "step": 409
+    },
+    {
+      "epoch": 0.6125974693000887,
+      "grad_norm": 0.3966883718967438,
+      "learning_rate": 3.9750000000000004e-05,
+      "loss": 1.7663,
+      "step": 410
+    },
+    {
+      "epoch": 0.6140916094691133,
+      "grad_norm": 0.3764961361885071,
+      "learning_rate": 3.9725e-05,
+      "loss": 1.6983,
+      "step": 411
+    },
+    {
+      "epoch": 0.615585749638138,
+      "grad_norm": 0.3106807470321655,
+      "learning_rate": 3.97e-05,
+      "loss": 1.4007,
+      "step": 412
+    },
+    {
+      "epoch": 0.6170798898071626,
+      "grad_norm": 0.5766478776931763,
+      "learning_rate": 3.9675e-05,
+      "loss": 1.8415,
+      "step": 413
+    },
+    {
+      "epoch": 0.6185740299761872,
+      "grad_norm": 0.662083089351654,
+      "learning_rate": 3.965e-05,
+      "loss": 1.4447,
+      "step": 414
+    },
+    {
+      "epoch": 0.6200681701452118,
+      "grad_norm": 0.7968102097511292,
+      "learning_rate": 3.9625e-05,
+      "loss": 1.7127,
+      "step": 415
+    },
+    {
+      "epoch": 0.6215623103142364,
+      "grad_norm": 0.3866075575351715,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 1.6128,
+      "step": 416
+    },
+    {
+      "epoch": 0.623056450483261,
+      "grad_norm": 0.40726110339164734,
+      "learning_rate": 3.9575e-05,
+      "loss": 1.4841,
+      "step": 417
+    },
+    {
+      "epoch": 0.6245505906522856,
+      "grad_norm": 0.3251897990703583,
+      "learning_rate": 3.9550000000000006e-05,
+      "loss": 1.4696,
+      "step": 418
+    },
+    {
+      "epoch": 0.6260447308213102,
+      "grad_norm": 0.5213460326194763,
+      "learning_rate": 3.9525e-05,
+      "loss": 1.6224,
+      "step": 419
+    },
+    {
+      "epoch": 0.6275388709903348,
+      "grad_norm": 0.33423733711242676,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 1.5016,
+      "step": 420
+    },
+    {
+      "epoch": 0.6290330111593594,
+      "grad_norm": 0.39543136954307556,
+      "learning_rate": 3.9475000000000004e-05,
+      "loss": 1.5712,
+      "step": 421
+    },
+    {
+      "epoch": 0.630527151328384,
+      "grad_norm": 0.4221692383289337,
+      "learning_rate": 3.9450000000000003e-05,
+      "loss": 1.3305,
+      "step": 422
+    },
+    {
+      "epoch": 0.6320212914974086,
+      "grad_norm": 0.3848581314086914,
+      "learning_rate": 3.9425e-05,
+      "loss": 1.5393,
+      "step": 423
+    },
+    {
+      "epoch": 0.6335154316664332,
+      "grad_norm": 0.510204553604126,
+      "learning_rate": 3.94e-05,
+      "loss": 1.4739,
+      "step": 424
+    },
+    {
+      "epoch": 0.6350095718354578,
+      "grad_norm": 0.37247434258461,
+      "learning_rate": 3.9375e-05,
+      "loss": 1.3264,
+      "step": 425
+    },
+    {
+      "epoch": 0.6365037120044824,
+      "grad_norm": 0.6320198774337769,
+      "learning_rate": 3.935e-05,
+      "loss": 1.6585,
+      "step": 426
+    },
+    {
+      "epoch": 0.637997852173507,
+      "grad_norm": 0.4277353584766388,
+      "learning_rate": 3.9325e-05,
+      "loss": 1.5301,
+      "step": 427
+    },
+    {
+      "epoch": 0.6394919923425316,
+      "grad_norm": 0.33874911069869995,
+      "learning_rate": 3.9300000000000007e-05,
+      "loss": 1.4133,
+      "step": 428
+    },
+    {
+      "epoch": 0.6409861325115562,
+      "grad_norm": 0.5412879586219788,
+      "learning_rate": 3.9275e-05,
+      "loss": 1.4922,
+      "step": 429
+    },
+    {
+      "epoch": 0.6424802726805808,
+      "grad_norm": 0.8273118138313293,
+      "learning_rate": 3.9250000000000005e-05,
+      "loss": 1.4497,
+      "step": 430
+    },
+    {
+      "epoch": 0.6439744128496054,
+      "grad_norm": 0.42817798256874084,
+      "learning_rate": 3.9225e-05,
+      "loss": 1.6066,
+      "step": 431
+    },
+    {
+      "epoch": 0.64546855301863,
+      "grad_norm": 0.5471529960632324,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.4121,
+      "step": 432
+    },
+    {
+      "epoch": 0.6469626931876546,
+      "grad_norm": 0.48357531428337097,
+      "learning_rate": 3.9175000000000004e-05,
+      "loss": 1.4167,
+      "step": 433
+    },
+    {
+      "epoch": 0.6484568333566793,
+      "grad_norm": 0.561120331287384,
+      "learning_rate": 3.915e-05,
+      "loss": 1.3161,
+      "step": 434
+    },
+    {
+      "epoch": 0.6499509735257039,
+      "grad_norm": 0.37435629963874817,
+      "learning_rate": 3.9125e-05,
+      "loss": 1.4888,
+      "step": 435
+    },
+    {
+      "epoch": 0.6514451136947285,
+      "grad_norm": 0.302143931388855,
+      "learning_rate": 3.91e-05,
+      "loss": 1.3934,
+      "step": 436
+    },
+    {
+      "epoch": 0.6529392538637531,
+      "grad_norm": 0.6095237135887146,
+      "learning_rate": 3.9075e-05,
+      "loss": 1.4218,
+      "step": 437
+    },
+    {
+      "epoch": 0.6544333940327777,
+      "grad_norm": 0.6860696077346802,
+      "learning_rate": 3.905e-05,
+      "loss": 1.4912,
+      "step": 438
+    },
+    {
+      "epoch": 0.6559275342018023,
+      "grad_norm": 0.34080833196640015,
+      "learning_rate": 3.9025e-05,
+      "loss": 1.3564,
+      "step": 439
+    },
+    {
+      "epoch": 0.6574216743708269,
+      "grad_norm": 0.568854808807373,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.5372,
+      "step": 440
+    },
+    {
+      "epoch": 0.6589158145398515,
+      "grad_norm": 0.3843095898628235,
+      "learning_rate": 3.8975e-05,
+      "loss": 1.6467,
+      "step": 441
+    },
+    {
+      "epoch": 0.6604099547088761,
+      "grad_norm": 0.7436652183532715,
+      "learning_rate": 3.8950000000000005e-05,
+      "loss": 1.5655,
+      "step": 442
+    },
+    {
+      "epoch": 0.6619040948779007,
+      "grad_norm": 0.6002988219261169,
+      "learning_rate": 3.8925e-05,
+      "loss": 1.6488,
+      "step": 443
+    },
+    {
+      "epoch": 0.6633982350469253,
+      "grad_norm": 0.9423897862434387,
+      "learning_rate": 3.8900000000000004e-05,
+      "loss": 1.4125,
+      "step": 444
+    },
+    {
+      "epoch": 0.6648923752159499,
+      "grad_norm": 0.6755229234695435,
+      "learning_rate": 3.8875e-05,
+      "loss": 1.4681,
+      "step": 445
+    },
+    {
+      "epoch": 0.6663865153849745,
+      "grad_norm": 0.7312940955162048,
+      "learning_rate": 3.885e-05,
+      "loss": 1.8124,
+      "step": 446
+    },
+    {
+      "epoch": 0.6678806555539991,
+      "grad_norm": 0.48998695611953735,
+      "learning_rate": 3.8825e-05,
+      "loss": 1.5058,
+      "step": 447
+    },
+    {
+      "epoch": 0.6693747957230237,
+      "grad_norm": 0.41998910903930664,
+      "learning_rate": 3.88e-05,
+      "loss": 1.3735,
+      "step": 448
+    },
+    {
+      "epoch": 0.6708689358920483,
+      "grad_norm": 0.5682637095451355,
+      "learning_rate": 3.8775e-05,
+      "loss": 1.5505,
+      "step": 449
+    },
+    {
+      "epoch": 0.6723630760610729,
+      "grad_norm": 0.6611406803131104,
+      "learning_rate": 3.875e-05,
+      "loss": 1.6073,
+      "step": 450
+    },
+    {
+      "epoch": 0.6738572162300975,
+      "grad_norm": 0.3332504332065582,
+      "learning_rate": 3.8725e-05,
+      "loss": 1.4912,
+      "step": 451
+    },
+    {
+      "epoch": 0.6753513563991221,
+      "grad_norm": 0.3210924565792084,
+      "learning_rate": 3.8700000000000006e-05,
+      "loss": 1.4406,
+      "step": 452
+    },
+    {
+      "epoch": 0.6768454965681469,
+      "grad_norm": 0.5756349563598633,
+      "learning_rate": 3.8675e-05,
+      "loss": 1.5155,
+      "step": 453
+    },
+    {
+      "epoch": 0.6783396367371715,
+      "grad_norm": 0.33762139081954956,
+      "learning_rate": 3.8650000000000004e-05,
+      "loss": 1.577,
+      "step": 454
+    },
+    {
+      "epoch": 0.6798337769061961,
+      "grad_norm": 0.39244896173477173,
+      "learning_rate": 3.8625e-05,
+      "loss": 1.4908,
+      "step": 455
+    },
+    {
+      "epoch": 0.6813279170752207,
+      "grad_norm": 0.270628422498703,
+      "learning_rate": 3.86e-05,
+      "loss": 1.409,
+      "step": 456
+    },
+    {
+      "epoch": 0.6828220572442453,
+      "grad_norm": 0.3090542256832123,
+      "learning_rate": 3.8575e-05,
+      "loss": 1.5653,
+      "step": 457
+    },
+    {
+      "epoch": 0.6843161974132699,
+      "grad_norm": 0.46093299984931946,
+      "learning_rate": 3.855e-05,
+      "loss": 1.6372,
+      "step": 458
+    },
+    {
+      "epoch": 0.6858103375822945,
+      "grad_norm": 0.6079492568969727,
+      "learning_rate": 3.8525e-05,
+      "loss": 1.4893,
+      "step": 459
+    },
+    {
+      "epoch": 0.6873044777513191,
+      "grad_norm": 1.2198094129562378,
+      "learning_rate": 3.85e-05,
+      "loss": 1.5613,
+      "step": 460
+    },
+    {
+      "epoch": 0.6887986179203437,
+      "grad_norm": 0.3242771029472351,
+      "learning_rate": 3.8475e-05,
+      "loss": 1.4498,
+      "step": 461
+    },
+    {
+      "epoch": 0.6902927580893683,
+      "grad_norm": 0.2971858084201813,
+      "learning_rate": 3.845e-05,
+      "loss": 1.5638,
+      "step": 462
+    },
+    {
+      "epoch": 0.6917868982583929,
+      "grad_norm": 0.2709638178348541,
+      "learning_rate": 3.8425e-05,
+      "loss": 1.3314,
+      "step": 463
+    },
+    {
+      "epoch": 0.6932810384274175,
+      "grad_norm": 0.25972089171409607,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.3006,
+      "step": 464
+    },
+    {
+      "epoch": 0.6947751785964421,
+      "grad_norm": 0.2617763876914978,
+      "learning_rate": 3.8375e-05,
+      "loss": 1.3331,
+      "step": 465
+    },
+    {
+      "epoch": 0.6962693187654667,
+      "grad_norm": 0.5027948021888733,
+      "learning_rate": 3.8350000000000004e-05,
+      "loss": 1.524,
+      "step": 466
+    },
+    {
+      "epoch": 0.6977634589344913,
+      "grad_norm": 0.637778639793396,
+      "learning_rate": 3.8324999999999996e-05,
+      "loss": 1.376,
+      "step": 467
+    },
+    {
+      "epoch": 0.6992575991035159,
+      "grad_norm": 0.315097451210022,
+      "learning_rate": 3.83e-05,
+      "loss": 1.4328,
+      "step": 468
+    },
+    {
+      "epoch": 0.7007517392725405,
+      "grad_norm": 0.311083048582077,
+      "learning_rate": 3.8275e-05,
+      "loss": 1.3065,
+      "step": 469
+    },
+    {
+      "epoch": 0.7022458794415651,
+      "grad_norm": 0.39416611194610596,
+      "learning_rate": 3.825e-05,
+      "loss": 1.5988,
+      "step": 470
+    },
+    {
+      "epoch": 0.7037400196105897,
+      "grad_norm": 0.33352068066596985,
+      "learning_rate": 3.8225e-05,
+      "loss": 1.5433,
+      "step": 471
+    },
+    {
+      "epoch": 0.7052341597796143,
+      "grad_norm": 0.9079835414886475,
+      "learning_rate": 3.82e-05,
+      "loss": 1.6284,
+      "step": 472
+    },
+    {
+      "epoch": 0.706728299948639,
+      "grad_norm": 0.370746910572052,
+      "learning_rate": 3.8175e-05,
+      "loss": 1.3519,
+      "step": 473
+    },
+    {
+      "epoch": 0.7082224401176636,
+      "grad_norm": 0.6235456466674805,
+      "learning_rate": 3.8150000000000006e-05,
+      "loss": 1.6093,
+      "step": 474
+    },
+    {
+      "epoch": 0.7097165802866882,
+      "grad_norm": 0.3454919457435608,
+      "learning_rate": 3.8125e-05,
+      "loss": 1.4376,
+      "step": 475
+    },
+    {
+      "epoch": 0.7112107204557128,
+      "grad_norm": 1.0801193714141846,
+      "learning_rate": 3.8100000000000005e-05,
+      "loss": 1.4585,
+      "step": 476
+    },
+    {
+      "epoch": 0.7127048606247374,
+      "grad_norm": 0.30193617939949036,
+      "learning_rate": 3.8075e-05,
+      "loss": 1.5557,
+      "step": 477
+    },
+    {
+      "epoch": 0.714199000793762,
+      "grad_norm": 0.8482101559638977,
+      "learning_rate": 3.805e-05,
+      "loss": 1.4223,
+      "step": 478
+    },
+    {
+      "epoch": 0.7156931409627866,
+      "grad_norm": 0.35040053725242615,
+      "learning_rate": 3.8025e-05,
+      "loss": 1.5185,
+      "step": 479
+    },
+    {
+      "epoch": 0.7171872811318112,
+      "grad_norm": 0.468925803899765,
+      "learning_rate": 3.8e-05,
+      "loss": 1.4051,
+      "step": 480
+    },
+    {
+      "epoch": 0.7186814213008358,
+      "grad_norm": 0.2951725423336029,
+      "learning_rate": 3.7975e-05,
+      "loss": 1.4082,
+      "step": 481
+    },
+    {
+      "epoch": 0.7201755614698604,
+      "grad_norm": 0.2716180086135864,
+      "learning_rate": 3.795e-05,
+      "loss": 1.3357,
+      "step": 482
+    },
+    {
+      "epoch": 0.721669701638885,
+      "grad_norm": 0.6728554368019104,
+      "learning_rate": 3.7925e-05,
+      "loss": 1.3553,
+      "step": 483
+    },
+    {
+      "epoch": 0.7231638418079096,
+      "grad_norm": 0.42035940289497375,
+      "learning_rate": 3.79e-05,
+      "loss": 1.382,
+      "step": 484
+    },
+    {
+      "epoch": 0.7246579819769342,
+      "grad_norm": 0.5181907415390015,
+      "learning_rate": 3.7875e-05,
+      "loss": 1.5328,
+      "step": 485
+    },
+    {
+      "epoch": 0.7261521221459588,
+      "grad_norm": 0.30366700887680054,
+      "learning_rate": 3.7850000000000005e-05,
+      "loss": 1.4639,
+      "step": 486
+    },
+    {
+      "epoch": 0.7276462623149834,
+      "grad_norm": 0.30150046944618225,
+      "learning_rate": 3.7825e-05,
+      "loss": 1.336,
+      "step": 487
+    },
+    {
+      "epoch": 0.729140402484008,
+      "grad_norm": 1.3998833894729614,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 1.7007,
+      "step": 488
+    },
+    {
+      "epoch": 0.7306345426530326,
+      "grad_norm": 0.3373676836490631,
+      "learning_rate": 3.7775e-05,
+      "loss": 1.289,
+      "step": 489
+    },
+    {
+      "epoch": 0.7321286828220572,
+      "grad_norm": 0.43975895643234253,
+      "learning_rate": 3.775e-05,
+      "loss": 1.6339,
+      "step": 490
+    },
+    {
+      "epoch": 0.7336228229910818,
+      "grad_norm": 0.2881298363208771,
+      "learning_rate": 3.7725e-05,
+      "loss": 1.6197,
+      "step": 491
+    },
+    {
+      "epoch": 0.7351169631601064,
+      "grad_norm": 0.2758522033691406,
+      "learning_rate": 3.77e-05,
+      "loss": 1.4038,
+      "step": 492
+    },
+    {
+      "epoch": 0.736611103329131,
+      "grad_norm": 0.2488110512495041,
+      "learning_rate": 3.7675e-05,
+      "loss": 1.4352,
+      "step": 493
+    },
+    {
+      "epoch": 0.7381052434981556,
+      "grad_norm": 0.2574450671672821,
+      "learning_rate": 3.765e-05,
+      "loss": 1.2815,
+      "step": 494
+    },
+    {
+      "epoch": 0.7395993836671803,
+      "grad_norm": 0.38335368037223816,
+      "learning_rate": 3.7625e-05,
+      "loss": 1.2665,
+      "step": 495
+    },
+    {
+      "epoch": 0.7410935238362049,
+      "grad_norm": 0.3784824311733246,
+      "learning_rate": 3.76e-05,
+      "loss": 1.6348,
+      "step": 496
+    },
+    {
+      "epoch": 0.7425876640052295,
+      "grad_norm": 0.32331913709640503,
+      "learning_rate": 3.7575e-05,
+      "loss": 1.4485,
+      "step": 497
+    },
+    {
+      "epoch": 0.7440818041742541,
+      "grad_norm": 0.636945903301239,
+      "learning_rate": 3.7550000000000005e-05,
+      "loss": 1.474,
+      "step": 498
+    },
+    {
+      "epoch": 0.7455759443432787,
+      "grad_norm": 0.4890618920326233,
+      "learning_rate": 3.7525e-05,
+      "loss": 1.4858,
+      "step": 499
+    },
+    {
+      "epoch": 0.7470700845123033,
+      "grad_norm": 1.4802592992782593,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.5492,
+      "step": 500
+    },
+    {
+      "epoch": 0.7485642246813279,
+      "grad_norm": 0.8308455348014832,
+      "learning_rate": 3.7475e-05,
+      "loss": 1.4273,
+      "step": 501
+    },
+    {
+      "epoch": 0.7500583648503525,
+      "grad_norm": 0.3105123043060303,
+      "learning_rate": 3.745e-05,
+      "loss": 1.6754,
+      "step": 502
+    },
+    {
+      "epoch": 0.7515525050193771,
+      "grad_norm": 0.3114786744117737,
+      "learning_rate": 3.7425e-05,
+      "loss": 1.4259,
+      "step": 503
+    },
+    {
+      "epoch": 0.7530466451884017,
+      "grad_norm": 0.4946189224720001,
+      "learning_rate": 3.74e-05,
+      "loss": 1.5702,
+      "step": 504
+    },
+    {
+      "epoch": 0.7545407853574263,
+      "grad_norm": 0.3033464550971985,
+      "learning_rate": 3.737500000000001e-05,
+      "loss": 1.509,
+      "step": 505
+    },
+    {
+      "epoch": 0.7560349255264509,
+      "grad_norm": 0.3518037796020508,
+      "learning_rate": 3.735e-05,
+      "loss": 1.555,
+      "step": 506
+    },
+    {
+      "epoch": 0.7575290656954755,
+      "grad_norm": 4.805240631103516,
+      "learning_rate": 3.7325000000000006e-05,
+      "loss": 1.4278,
+      "step": 507
+    },
+    {
+      "epoch": 0.7590232058645001,
+      "grad_norm": 0.2884902358055115,
+      "learning_rate": 3.73e-05,
+      "loss": 1.3305,
+      "step": 508
+    },
+    {
+      "epoch": 0.7605173460335247,
+      "grad_norm": 0.48153796792030334,
+      "learning_rate": 3.7275000000000005e-05,
+      "loss": 1.5316,
+      "step": 509
+    },
+    {
+      "epoch": 0.7620114862025493,
+      "grad_norm": 0.4066573977470398,
+      "learning_rate": 3.7250000000000004e-05,
+      "loss": 1.7447,
+      "step": 510
+    },
+    {
+      "epoch": 0.7635056263715739,
+      "grad_norm": 0.35123422741889954,
+      "learning_rate": 3.7225000000000004e-05,
+      "loss": 1.4725,
+      "step": 511
+    },
+    {
+      "epoch": 0.7649997665405985,
+      "grad_norm": 4.914884090423584,
+      "learning_rate": 3.72e-05,
+      "loss": 1.4226,
+      "step": 512
+    },
+    {
+      "epoch": 0.7664939067096233,
+      "grad_norm": 0.6892772912979126,
+      "learning_rate": 3.7175e-05,
+      "loss": 1.4583,
+      "step": 513
+    },
+    {
+      "epoch": 0.7679880468786479,
+      "grad_norm": 0.7510517239570618,
+      "learning_rate": 3.715e-05,
+      "loss": 1.5202,
+      "step": 514
+    },
+    {
+      "epoch": 0.7694821870476725,
+      "grad_norm": 0.3141719102859497,
+      "learning_rate": 3.7125e-05,
+      "loss": 1.4267,
+      "step": 515
+    },
+    {
+      "epoch": 0.7709763272166971,
+      "grad_norm": 0.36383432149887085,
+      "learning_rate": 3.71e-05,
+      "loss": 1.6657,
+      "step": 516
+    },
+    {
+      "epoch": 0.7724704673857217,
+      "grad_norm": 0.30458924174308777,
+      "learning_rate": 3.707500000000001e-05,
+      "loss": 1.3829,
+      "step": 517
+    },
+    {
+      "epoch": 0.7739646075547463,
+      "grad_norm": 0.4784793257713318,
+      "learning_rate": 3.705e-05,
+      "loss": 1.3675,
+      "step": 518
+    },
+    {
+      "epoch": 0.7754587477237709,
+      "grad_norm": 0.6077926754951477,
+      "learning_rate": 3.7025000000000005e-05,
+      "loss": 1.4977,
+      "step": 519
+    },
+    {
+      "epoch": 0.7769528878927955,
+      "grad_norm": 0.5152701139450073,
+      "learning_rate": 3.7e-05,
+      "loss": 1.6965,
+      "step": 520
+    },
+    {
+      "epoch": 0.7784470280618201,
+      "grad_norm": 0.2685502767562866,
+      "learning_rate": 3.6975000000000004e-05,
+      "loss": 1.3124,
+      "step": 521
+    },
+    {
+      "epoch": 0.7799411682308447,
+      "grad_norm": 0.3213263750076294,
+      "learning_rate": 3.6950000000000004e-05,
+      "loss": 1.4155,
+      "step": 522
+    },
+    {
+      "epoch": 0.7814353083998693,
+      "grad_norm": 0.5548874735832214,
+      "learning_rate": 3.6925e-05,
+      "loss": 1.302,
+      "step": 523
+    },
+    {
+      "epoch": 0.7829294485688939,
+      "grad_norm": 0.3476009666919708,
+      "learning_rate": 3.69e-05,
+      "loss": 1.593,
+      "step": 524
+    },
+    {
+      "epoch": 0.7844235887379185,
+      "grad_norm": 0.45263391733169556,
+      "learning_rate": 3.6875e-05,
+      "loss": 1.5412,
+      "step": 525
+    },
+    {
+      "epoch": 0.7859177289069431,
+      "grad_norm": 0.3879849910736084,
+      "learning_rate": 3.685e-05,
+      "loss": 1.4468,
+      "step": 526
+    },
+    {
+      "epoch": 0.7874118690759677,
+      "grad_norm": 0.3185078203678131,
+      "learning_rate": 3.6825e-05,
+      "loss": 1.4325,
+      "step": 527
+    },
+    {
+      "epoch": 0.7889060092449923,
+      "grad_norm": 0.3549663722515106,
+      "learning_rate": 3.68e-05,
+      "loss": 1.5194,
+      "step": 528
+    },
+    {
+      "epoch": 0.7904001494140169,
+      "grad_norm": 0.7724360823631287,
+      "learning_rate": 3.6775000000000006e-05,
+      "loss": 1.5654,
+      "step": 529
+    },
+    {
+      "epoch": 0.7918942895830415,
+      "grad_norm": 0.48684829473495483,
+      "learning_rate": 3.675e-05,
+      "loss": 1.4374,
+      "step": 530
+    },
+    {
+      "epoch": 0.7933884297520661,
+      "grad_norm": 0.35977497696876526,
+      "learning_rate": 3.6725000000000005e-05,
+      "loss": 1.4841,
+      "step": 531
+    },
+    {
+      "epoch": 0.7948825699210907,
+      "grad_norm": 0.5556880831718445,
+      "learning_rate": 3.6700000000000004e-05,
+      "loss": 1.575,
+      "step": 532
+    },
+    {
+      "epoch": 0.7963767100901153,
+      "grad_norm": 0.3836474120616913,
+      "learning_rate": 3.6675000000000004e-05,
+      "loss": 1.4394,
+      "step": 533
+    },
+    {
+      "epoch": 0.79787085025914,
+      "grad_norm": 0.606639564037323,
+      "learning_rate": 3.665e-05,
+      "loss": 1.4676,
+      "step": 534
+    },
+    {
+      "epoch": 0.7993649904281646,
+      "grad_norm": 0.37593191862106323,
+      "learning_rate": 3.6625e-05,
+      "loss": 1.4674,
+      "step": 535
+    },
+    {
+      "epoch": 0.8008591305971892,
+      "grad_norm": 0.33318042755126953,
+      "learning_rate": 3.66e-05,
+      "loss": 1.3682,
+      "step": 536
+    },
+    {
+      "epoch": 0.8023532707662138,
+      "grad_norm": 0.427814781665802,
+      "learning_rate": 3.6575e-05,
+      "loss": 1.5771,
+      "step": 537
+    },
+    {
+      "epoch": 0.8038474109352384,
+      "grad_norm": 0.5013397336006165,
+      "learning_rate": 3.655e-05,
+      "loss": 1.5273,
+      "step": 538
+    },
+    {
+      "epoch": 0.805341551104263,
+      "grad_norm": 0.333675742149353,
+      "learning_rate": 3.652500000000001e-05,
+      "loss": 1.3863,
+      "step": 539
+    },
+    {
+      "epoch": 0.8068356912732876,
+      "grad_norm": 0.27623650431632996,
+      "learning_rate": 3.65e-05,
+      "loss": 1.4684,
+      "step": 540
+    },
+    {
+      "epoch": 0.8083298314423122,
+      "grad_norm": 0.8387918472290039,
+      "learning_rate": 3.6475000000000006e-05,
+      "loss": 1.5025,
+      "step": 541
+    },
+    {
+      "epoch": 0.8098239716113368,
+      "grad_norm": 0.39421287178993225,
+      "learning_rate": 3.645e-05,
+      "loss": 1.5331,
+      "step": 542
+    },
+    {
+      "epoch": 0.8113181117803614,
+      "grad_norm": 0.44059473276138306,
+      "learning_rate": 3.6425000000000004e-05,
+      "loss": 1.5072,
+      "step": 543
+    },
+    {
+      "epoch": 0.812812251949386,
+      "grad_norm": 0.4156414866447449,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 1.3086,
+      "step": 544
+    },
+    {
+      "epoch": 0.8143063921184106,
+      "grad_norm": 0.4378097355365753,
+      "learning_rate": 3.6375e-05,
+      "loss": 1.2068,
+      "step": 545
+    },
+    {
+      "epoch": 0.8158005322874352,
+      "grad_norm": 0.45686447620391846,
+      "learning_rate": 3.635e-05,
+      "loss": 1.4368,
+      "step": 546
+    },
+    {
+      "epoch": 0.8172946724564598,
+      "grad_norm": 0.3088967502117157,
+      "learning_rate": 3.6325e-05,
+      "loss": 1.5133,
+      "step": 547
+    },
+    {
+      "epoch": 0.8187888126254844,
+      "grad_norm": 0.28649911284446716,
+      "learning_rate": 3.63e-05,
+      "loss": 1.5194,
+      "step": 548
+    },
+    {
+      "epoch": 0.820282952794509,
+      "grad_norm": 0.28436142206192017,
+      "learning_rate": 3.6275e-05,
+      "loss": 1.3408,
+      "step": 549
+    },
+    {
+      "epoch": 0.8217770929635336,
+      "grad_norm": 0.4080042243003845,
+      "learning_rate": 3.625e-05,
+      "loss": 1.5412,
+      "step": 550
+    },
+    {
+      "epoch": 0.8232712331325582,
+      "grad_norm": 0.47546809911727905,
+      "learning_rate": 3.6225000000000006e-05,
+      "loss": 1.6426,
+      "step": 551
+    },
+    {
+      "epoch": 0.8247653733015828,
+      "grad_norm": 0.30255812406539917,
+      "learning_rate": 3.62e-05,
+      "loss": 1.2971,
+      "step": 552
+    },
+    {
+      "epoch": 0.8262595134706074,
+      "grad_norm": 0.526885986328125,
+      "learning_rate": 3.6175000000000005e-05,
+      "loss": 1.4314,
+      "step": 553
+    },
+    {
+      "epoch": 0.827753653639632,
+      "grad_norm": 0.3695448040962219,
+      "learning_rate": 3.615e-05,
+      "loss": 1.5206,
+      "step": 554
+    },
+    {
+      "epoch": 0.8292477938086567,
+      "grad_norm": 0.36277472972869873,
+      "learning_rate": 3.6125000000000004e-05,
+      "loss": 1.4134,
+      "step": 555
+    },
+    {
+      "epoch": 0.8307419339776813,
+      "grad_norm": 0.2916204631328583,
+      "learning_rate": 3.61e-05,
+      "loss": 1.4018,
+      "step": 556
+    },
+    {
+      "epoch": 0.8322360741467059,
+      "grad_norm": 0.26475563645362854,
+      "learning_rate": 3.6075e-05,
+      "loss": 1.3982,
+      "step": 557
+    },
+    {
+      "epoch": 0.8337302143157305,
+      "grad_norm": 0.35685673356056213,
+      "learning_rate": 3.605e-05,
+      "loss": 1.4475,
+      "step": 558
+    },
+    {
+      "epoch": 0.8352243544847551,
+      "grad_norm": 0.3294183909893036,
+      "learning_rate": 3.6025e-05,
+      "loss": 1.263,
+      "step": 559
+    },
+    {
+      "epoch": 0.8367184946537797,
+      "grad_norm": 0.3987225890159607,
+      "learning_rate": 3.6e-05,
+      "loss": 1.3704,
+      "step": 560
+    },
+    {
+      "epoch": 0.8382126348228043,
+      "grad_norm": 0.40317440032958984,
+      "learning_rate": 3.5975e-05,
+      "loss": 1.5723,
+      "step": 561
+    },
+    {
+      "epoch": 0.8397067749918289,
+      "grad_norm": 0.3420385718345642,
+      "learning_rate": 3.595e-05,
+      "loss": 1.5984,
+      "step": 562
+    },
+    {
+      "epoch": 0.8412009151608535,
+      "grad_norm": 0.3810725510120392,
+      "learning_rate": 3.5925000000000006e-05,
+      "loss": 1.4563,
+      "step": 563
+    },
+    {
+      "epoch": 0.8426950553298781,
+      "grad_norm": 0.5123723745346069,
+      "learning_rate": 3.59e-05,
+      "loss": 1.4666,
+      "step": 564
+    },
+    {
+      "epoch": 0.8441891954989027,
+      "grad_norm": 0.21530090272426605,
+      "learning_rate": 3.5875000000000005e-05,
+      "loss": 1.3619,
+      "step": 565
+    },
+    {
+      "epoch": 0.8456833356679273,
+      "grad_norm": 1.0108494758605957,
+      "learning_rate": 3.585e-05,
+      "loss": 1.4085,
+      "step": 566
+    },
+    {
+      "epoch": 0.8471774758369519,
+      "grad_norm": 0.47152891755104065,
+      "learning_rate": 3.5825000000000003e-05,
+      "loss": 1.4704,
+      "step": 567
+    },
+    {
+      "epoch": 0.8486716160059765,
+      "grad_norm": 0.3395189046859741,
+      "learning_rate": 3.58e-05,
+      "loss": 1.468,
+      "step": 568
+    },
+    {
+      "epoch": 0.8501657561750011,
+      "grad_norm": 0.2822551727294922,
+      "learning_rate": 3.5775e-05,
+      "loss": 1.2872,
+      "step": 569
+    },
+    {
+      "epoch": 0.8516598963440257,
+      "grad_norm": 0.6422669887542725,
+      "learning_rate": 3.575e-05,
+      "loss": 1.3155,
+      "step": 570
+    },
+    {
+      "epoch": 0.8531540365130503,
+      "grad_norm": 0.45001545548439026,
+      "learning_rate": 3.5725e-05,
+      "loss": 1.4375,
+      "step": 571
+    },
+    {
+      "epoch": 0.8546481766820749,
+      "grad_norm": 0.26399505138397217,
+      "learning_rate": 3.57e-05,
+      "loss": 1.394,
+      "step": 572
+    },
+    {
+      "epoch": 0.8561423168510995,
+      "grad_norm": 0.3477565050125122,
+      "learning_rate": 3.5675e-05,
+      "loss": 1.3054,
+      "step": 573
+    },
+    {
+      "epoch": 0.8576364570201243,
+      "grad_norm": 0.33781954646110535,
+      "learning_rate": 3.565e-05,
+      "loss": 1.3096,
+      "step": 574
+    },
+    {
+      "epoch": 0.8591305971891489,
+      "grad_norm": 0.2942800521850586,
+      "learning_rate": 3.5625000000000005e-05,
+      "loss": 1.4921,
+      "step": 575
+    },
+    {
+      "epoch": 0.8606247373581735,
+      "grad_norm": 0.47856467962265015,
+      "learning_rate": 3.56e-05,
+      "loss": 1.2843,
+      "step": 576
+    },
+    {
+      "epoch": 0.8621188775271981,
+      "grad_norm": 0.369923859834671,
+      "learning_rate": 3.5575000000000004e-05,
+      "loss": 1.5034,
+      "step": 577
+    },
+    {
+      "epoch": 0.8636130176962227,
+      "grad_norm": 0.27429136633872986,
+      "learning_rate": 3.555e-05,
+      "loss": 1.3235,
+      "step": 578
+    },
+    {
+      "epoch": 0.8651071578652473,
+      "grad_norm": 0.42868244647979736,
+      "learning_rate": 3.5525e-05,
+      "loss": 1.437,
+      "step": 579
+    },
+    {
+      "epoch": 0.8666012980342719,
+      "grad_norm": 0.24422957003116608,
+      "learning_rate": 3.55e-05,
+      "loss": 1.2913,
+      "step": 580
+    },
+    {
+      "epoch": 0.8680954382032965,
+      "grad_norm": 0.34310224652290344,
+      "learning_rate": 3.5475e-05,
+      "loss": 1.3075,
+      "step": 581
+    },
+    {
+      "epoch": 0.8695895783723211,
+      "grad_norm": 0.436494380235672,
+      "learning_rate": 3.545e-05,
+      "loss": 1.5981,
+      "step": 582
+    },
+    {
+      "epoch": 0.8710837185413457,
+      "grad_norm": 0.2664638161659241,
+      "learning_rate": 3.5425e-05,
+      "loss": 1.3887,
+      "step": 583
+    },
+    {
+      "epoch": 0.8725778587103703,
+      "grad_norm": 0.24781467020511627,
+      "learning_rate": 3.54e-05,
+      "loss": 1.3543,
+      "step": 584
+    },
+    {
+      "epoch": 0.8740719988793949,
+      "grad_norm": 0.2947968542575836,
+      "learning_rate": 3.5375e-05,
+      "loss": 1.3933,
+      "step": 585
+    },
+    {
+      "epoch": 0.8755661390484195,
+      "grad_norm": 0.3375110328197479,
+      "learning_rate": 3.535e-05,
+      "loss": 1.4935,
+      "step": 586
+    },
+    {
+      "epoch": 0.8770602792174441,
+      "grad_norm": 0.3563228249549866,
+      "learning_rate": 3.5325000000000005e-05,
+      "loss": 1.5311,
+      "step": 587
+    },
+    {
+      "epoch": 0.8785544193864687,
+      "grad_norm": 0.38266658782958984,
+      "learning_rate": 3.53e-05,
+      "loss": 1.4301,
+      "step": 588
+    },
+    {
+      "epoch": 0.8800485595554933,
+      "grad_norm": 0.2319614440202713,
+      "learning_rate": 3.5275000000000004e-05,
+      "loss": 1.3498,
+      "step": 589
+    },
+    {
+      "epoch": 0.8815426997245179,
+      "grad_norm": 0.4004661440849304,
+      "learning_rate": 3.525e-05,
+      "loss": 1.5444,
+      "step": 590
+    },
+    {
+      "epoch": 0.8830368398935425,
+      "grad_norm": 0.25451788306236267,
+      "learning_rate": 3.5225e-05,
+      "loss": 1.4218,
+      "step": 591
+    },
+    {
+      "epoch": 0.8845309800625671,
+      "grad_norm": 0.46508535742759705,
+      "learning_rate": 3.52e-05,
+      "loss": 1.426,
+      "step": 592
+    },
+    {
+      "epoch": 0.8860251202315917,
+      "grad_norm": 0.45118725299835205,
+      "learning_rate": 3.5175e-05,
+      "loss": 1.4752,
+      "step": 593
+    },
+    {
+      "epoch": 0.8875192604006163,
+      "grad_norm": 0.30142486095428467,
+      "learning_rate": 3.515e-05,
+      "loss": 1.4359,
+      "step": 594
+    },
+    {
+      "epoch": 0.889013400569641,
+      "grad_norm": 0.38366007804870605,
+      "learning_rate": 3.5125e-05,
+      "loss": 1.3149,
+      "step": 595
+    },
+    {
+      "epoch": 0.8905075407386656,
+      "grad_norm": 0.275544136762619,
+      "learning_rate": 3.51e-05,
+      "loss": 1.2721,
+      "step": 596
+    },
+    {
+      "epoch": 0.8920016809076902,
+      "grad_norm": 0.3999973237514496,
+      "learning_rate": 3.5075000000000006e-05,
+      "loss": 1.4521,
+      "step": 597
+    },
+    {
+      "epoch": 0.8934958210767148,
+      "grad_norm": 0.22361697256565094,
+      "learning_rate": 3.505e-05,
+      "loss": 1.3303,
+      "step": 598
+    },
+    {
+      "epoch": 0.8949899612457394,
+      "grad_norm": 0.3005754351615906,
+      "learning_rate": 3.5025000000000004e-05,
+      "loss": 1.2179,
+      "step": 599
+    },
+    {
+      "epoch": 0.896484101414764,
+      "grad_norm": 0.3509037494659424,
+      "learning_rate": 3.5e-05,
+      "loss": 1.3804,
+      "step": 600
+    },
+    {
+      "epoch": 0.8979782415837886,
+      "grad_norm": 0.32248327136039734,
+      "learning_rate": 3.4975e-05,
+      "loss": 1.4697,
+      "step": 601
+    },
+    {
+      "epoch": 0.8994723817528132,
+      "grad_norm": 0.6695634126663208,
+      "learning_rate": 3.495e-05,
+      "loss": 1.5914,
+      "step": 602
+    },
+    {
+      "epoch": 0.9009665219218378,
+      "grad_norm": 0.39635902643203735,
+      "learning_rate": 3.4925e-05,
+      "loss": 1.1656,
+      "step": 603
+    },
+    {
+      "epoch": 0.9024606620908624,
+      "grad_norm": 0.44676870107650757,
+      "learning_rate": 3.49e-05,
+      "loss": 1.277,
+      "step": 604
+    },
+    {
+      "epoch": 0.903954802259887,
+      "grad_norm": 0.2875045835971832,
+      "learning_rate": 3.4875e-05,
+      "loss": 1.464,
+      "step": 605
+    },
+    {
+      "epoch": 0.9054489424289116,
+      "grad_norm": 0.4377475082874298,
+      "learning_rate": 3.485e-05,
+      "loss": 1.3933,
+      "step": 606
+    },
+    {
+      "epoch": 0.9069430825979362,
+      "grad_norm": 0.38153204321861267,
+      "learning_rate": 3.4825e-05,
+      "loss": 1.3474,
+      "step": 607
+    },
+    {
+      "epoch": 0.9084372227669608,
+      "grad_norm": 0.3615608215332031,
+      "learning_rate": 3.48e-05,
+      "loss": 1.3261,
+      "step": 608
+    },
+    {
+      "epoch": 0.9099313629359854,
+      "grad_norm": 0.4355512857437134,
+      "learning_rate": 3.4775000000000005e-05,
+      "loss": 1.4538,
+      "step": 609
+    },
+    {
+      "epoch": 0.91142550310501,
+      "grad_norm": 0.30448365211486816,
+      "learning_rate": 3.475e-05,
+      "loss": 1.5085,
+      "step": 610
+    },
+    {
+      "epoch": 0.9129196432740346,
+      "grad_norm": 0.2584160566329956,
+      "learning_rate": 3.4725000000000004e-05,
+      "loss": 1.2786,
+      "step": 611
+    },
+    {
+      "epoch": 0.9144137834430592,
+      "grad_norm": 0.2645299434661865,
+      "learning_rate": 3.4699999999999996e-05,
+      "loss": 1.4541,
+      "step": 612
+    },
+    {
+      "epoch": 0.9159079236120838,
+      "grad_norm": 0.29261913895606995,
+      "learning_rate": 3.4675e-05,
+      "loss": 1.4672,
+      "step": 613
+    },
+    {
+      "epoch": 0.9174020637811084,
+      "grad_norm": 0.33550161123275757,
+      "learning_rate": 3.465e-05,
+      "loss": 1.3547,
+      "step": 614
+    },
+    {
+      "epoch": 0.918896203950133,
+      "grad_norm": 1.3263543844223022,
+      "learning_rate": 3.4625e-05,
+      "loss": 1.4612,
+      "step": 615
+    },
+    {
+      "epoch": 0.9203903441191577,
+      "grad_norm": 0.28431329131126404,
+      "learning_rate": 3.46e-05,
+      "loss": 1.5397,
+      "step": 616
+    },
+    {
+      "epoch": 0.9218844842881823,
+      "grad_norm": 0.6457627415657043,
+      "learning_rate": 3.4575e-05,
+      "loss": 1.4713,
+      "step": 617
+    },
+    {
+      "epoch": 0.9233786244572069,
+      "grad_norm": 1.0010414123535156,
+      "learning_rate": 3.455e-05,
+      "loss": 1.5088,
+      "step": 618
+    },
+    {
+      "epoch": 0.9248727646262315,
+      "grad_norm": 0.3725411891937256,
+      "learning_rate": 3.4525e-05,
+      "loss": 1.3313,
+      "step": 619
+    },
+    {
+      "epoch": 0.9263669047952561,
+      "grad_norm": 0.4637717604637146,
+      "learning_rate": 3.45e-05,
+      "loss": 1.3926,
+      "step": 620
+    },
+    {
+      "epoch": 0.9278610449642807,
+      "grad_norm": 0.591336190700531,
+      "learning_rate": 3.4475000000000005e-05,
+      "loss": 1.3868,
+      "step": 621
+    },
+    {
+      "epoch": 0.9293551851333053,
+      "grad_norm": 0.44118255376815796,
+      "learning_rate": 3.445e-05,
+      "loss": 1.3341,
+      "step": 622
+    },
+    {
+      "epoch": 0.9308493253023299,
+      "grad_norm": 0.8001243472099304,
+      "learning_rate": 3.4425e-05,
+      "loss": 1.4864,
+      "step": 623
+    },
+    {
+      "epoch": 0.9323434654713545,
+      "grad_norm": 0.23716165125370026,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 1.2912,
+      "step": 624
+    },
+    {
+      "epoch": 0.9338376056403791,
+      "grad_norm": 0.2823966443538666,
+      "learning_rate": 3.4375e-05,
+      "loss": 1.3462,
+      "step": 625
+    },
+    {
+      "epoch": 0.9353317458094037,
+      "grad_norm": 0.4168063700199127,
+      "learning_rate": 3.435e-05,
+      "loss": 1.383,
+      "step": 626
+    },
+    {
+      "epoch": 0.9368258859784283,
+      "grad_norm": 0.3376123905181885,
+      "learning_rate": 3.4325e-05,
+      "loss": 1.2524,
+      "step": 627
+    },
+    {
+      "epoch": 0.9383200261474529,
+      "grad_norm": 0.5403313040733337,
+      "learning_rate": 3.430000000000001e-05,
+      "loss": 1.3068,
+      "step": 628
+    },
+    {
+      "epoch": 0.9398141663164775,
+      "grad_norm": 0.36801594495773315,
+      "learning_rate": 3.4275e-05,
+      "loss": 1.3774,
+      "step": 629
+    },
+    {
+      "epoch": 0.9413083064855021,
+      "grad_norm": 0.31388500332832336,
+      "learning_rate": 3.4250000000000006e-05,
+      "loss": 1.3882,
+      "step": 630
+    },
+    {
+      "epoch": 0.9428024466545267,
+      "grad_norm": 0.6890659332275391,
+      "learning_rate": 3.4225e-05,
+      "loss": 1.3958,
+      "step": 631
+    },
+    {
+      "epoch": 0.9442965868235513,
+      "grad_norm": 0.3605225086212158,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 1.4728,
+      "step": 632
+    },
+    {
+      "epoch": 0.9457907269925759,
+      "grad_norm": 0.3409678041934967,
+      "learning_rate": 3.4175000000000004e-05,
+      "loss": 1.2731,
+      "step": 633
+    },
+    {
+      "epoch": 0.9472848671616007,
+      "grad_norm": 0.22081942856311798,
+      "learning_rate": 3.415e-05,
+      "loss": 1.4976,
+      "step": 634
+    },
+    {
+      "epoch": 0.9487790073306253,
+      "grad_norm": 0.2751425504684448,
+      "learning_rate": 3.4125e-05,
+      "loss": 1.395,
+      "step": 635
+    },
+    {
+      "epoch": 0.9502731474996499,
+      "grad_norm": 0.4033777117729187,
+      "learning_rate": 3.41e-05,
+      "loss": 1.4546,
+      "step": 636
+    },
+    {
+      "epoch": 0.9517672876686745,
+      "grad_norm": 0.7273349761962891,
+      "learning_rate": 3.4075e-05,
+      "loss": 1.4307,
+      "step": 637
+    },
+    {
+      "epoch": 0.9532614278376991,
+      "grad_norm": 0.4339013993740082,
+      "learning_rate": 3.405e-05,
+      "loss": 1.4301,
+      "step": 638
+    },
+    {
+      "epoch": 0.9547555680067237,
+      "grad_norm": 0.3764870762825012,
+      "learning_rate": 3.4025e-05,
+      "loss": 1.2367,
+      "step": 639
+    },
+    {
+      "epoch": 0.9562497081757483,
+      "grad_norm": 0.468927264213562,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.1901,
+      "step": 640
+    },
+    {
+      "epoch": 0.9577438483447729,
+      "grad_norm": 0.25675371289253235,
+      "learning_rate": 3.3975e-05,
+      "loss": 1.1584,
+      "step": 641
+    },
+    {
+      "epoch": 0.9592379885137975,
+      "grad_norm": 0.40362444519996643,
+      "learning_rate": 3.3950000000000005e-05,
+      "loss": 1.4858,
+      "step": 642
+    },
+    {
+      "epoch": 0.9607321286828221,
+      "grad_norm": 0.38865911960601807,
+      "learning_rate": 3.3925e-05,
+      "loss": 1.4717,
+      "step": 643
+    },
+    {
+      "epoch": 0.9622262688518467,
+      "grad_norm": 0.4919620156288147,
+      "learning_rate": 3.3900000000000004e-05,
+      "loss": 1.3452,
+      "step": 644
+    },
+    {
+      "epoch": 0.9637204090208713,
+      "grad_norm": 0.2999139130115509,
+      "learning_rate": 3.3875000000000003e-05,
+      "loss": 1.3253,
+      "step": 645
+    },
+    {
+      "epoch": 0.9652145491898959,
+      "grad_norm": 0.3142748177051544,
+      "learning_rate": 3.385e-05,
+      "loss": 1.4305,
+      "step": 646
+    },
+    {
+      "epoch": 0.9667086893589205,
+      "grad_norm": 0.36084461212158203,
+      "learning_rate": 3.3825e-05,
+      "loss": 1.4515,
+      "step": 647
+    },
+    {
+      "epoch": 0.9682028295279451,
+      "grad_norm": 0.29961463809013367,
+      "learning_rate": 3.38e-05,
+      "loss": 1.5675,
+      "step": 648
+    },
+    {
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.26652055978775024,
+      "learning_rate": 3.3775e-05,
+      "loss": 1.4403,
+      "step": 649
+    },
+    {
+      "epoch": 0.9711911098659943,
+      "grad_norm": 0.601382315158844,
+      "learning_rate": 3.375000000000001e-05,
+      "loss": 1.4065,
+      "step": 650
+    },
+    {
+      "epoch": 0.9726852500350189,
+      "grad_norm": 0.2838383913040161,
+      "learning_rate": 3.3725e-05,
+      "loss": 1.3804,
+      "step": 651
+    },
+    {
+      "epoch": 0.9741793902040435,
+      "grad_norm": 0.3205418586730957,
+      "learning_rate": 3.3700000000000006e-05,
+      "loss": 1.2481,
+      "step": 652
+    },
+    {
+      "epoch": 0.9756735303730681,
+      "grad_norm": 0.28523319959640503,
+      "learning_rate": 3.3675e-05,
+      "loss": 1.5713,
+      "step": 653
+    },
+    {
+      "epoch": 0.9771676705420927,
+      "grad_norm": 0.3010360598564148,
+      "learning_rate": 3.3650000000000005e-05,
+      "loss": 1.5305,
+      "step": 654
+    },
+    {
+      "epoch": 0.9786618107111174,
+      "grad_norm": 0.22972212731838226,
+      "learning_rate": 3.3625000000000004e-05,
+      "loss": 1.362,
+      "step": 655
+    },
+    {
+      "epoch": 0.980155950880142,
+      "grad_norm": 0.31966134905815125,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.3665,
+      "step": 656
+    },
+    {
+      "epoch": 0.9816500910491666,
+      "grad_norm": 0.39883023500442505,
+      "learning_rate": 3.3575e-05,
+      "loss": 1.4166,
+      "step": 657
+    },
+    {
+      "epoch": 0.9831442312181912,
+      "grad_norm": 0.33253997564315796,
+      "learning_rate": 3.355e-05,
+      "loss": 1.3789,
+      "step": 658
+    },
+    {
+      "epoch": 0.9846383713872158,
+      "grad_norm": 0.2807278633117676,
+      "learning_rate": 3.3525e-05,
+      "loss": 1.5887,
+      "step": 659
+    },
+    {
+      "epoch": 0.9861325115562404,
+      "grad_norm": 0.29328641295433044,
+      "learning_rate": 3.35e-05,
+      "loss": 1.2736,
+      "step": 660
+    },
+    {
+      "epoch": 0.987626651725265,
+      "grad_norm": 0.3489589989185333,
+      "learning_rate": 3.3475e-05,
+      "loss": 1.4426,
+      "step": 661
+    },
+    {
+      "epoch": 0.9891207918942896,
+      "grad_norm": 0.26897239685058594,
+      "learning_rate": 3.345000000000001e-05,
+      "loss": 1.5113,
+      "step": 662
+    },
+    {
+      "epoch": 0.9906149320633142,
+      "grad_norm": 0.3970909118652344,
+      "learning_rate": 3.3425e-05,
+      "loss": 1.4399,
+      "step": 663
+    },
+    {
+      "epoch": 0.9921090722323388,
+      "grad_norm": 0.31349503993988037,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 1.4667,
+      "step": 664
+    },
+    {
+      "epoch": 0.9936032124013634,
+      "grad_norm": 0.4700935482978821,
+      "learning_rate": 3.3375e-05,
+      "loss": 1.3598,
+      "step": 665
+    },
+    {
+      "epoch": 0.995097352570388,
+      "grad_norm": 0.25302940607070923,
+      "learning_rate": 3.3350000000000004e-05,
+      "loss": 1.2526,
+      "step": 666
+    },
+    {
+      "epoch": 0.9965914927394126,
+      "grad_norm": 0.34383395314216614,
+      "learning_rate": 3.3325000000000004e-05,
+      "loss": 1.4012,
+      "step": 667
+    },
+    {
+      "epoch": 0.9980856329084372,
+      "grad_norm": 0.34091341495513916,
+      "learning_rate": 3.33e-05,
+      "loss": 1.3117,
+      "step": 668
+    },
+    {
+      "epoch": 0.9995797730774618,
+      "grad_norm": 0.3936751186847687,
+      "learning_rate": 3.3275e-05,
+      "loss": 1.3979,
+      "step": 669
+    },
+    {
+      "epoch": 1.0010739132464865,
+      "grad_norm": 0.3573870360851288,
+      "learning_rate": 3.325e-05,
+      "loss": 1.3728,
+      "step": 670
+    },
+    {
+      "epoch": 1.002568053415511,
+      "grad_norm": 0.23025058209896088,
+      "learning_rate": 3.3225e-05,
+      "loss": 1.4986,
+      "step": 671
+    },
+    {
+      "epoch": 1.0040621935845357,
+      "grad_norm": 0.30586880445480347,
+      "learning_rate": 3.32e-05,
+      "loss": 1.391,
+      "step": 672
+    },
+    {
+      "epoch": 1.0055563337535602,
+      "grad_norm": 0.27675920724868774,
+      "learning_rate": 3.3175e-05,
+      "loss": 1.2885,
+      "step": 673
+    },
+    {
+      "epoch": 1.007050473922585,
+      "grad_norm": 0.24607954919338226,
+      "learning_rate": 3.3150000000000006e-05,
+      "loss": 1.4229,
+      "step": 674
+    },
+    {
+      "epoch": 1.0085446140916094,
+      "grad_norm": 0.2698909640312195,
+      "learning_rate": 3.3125e-05,
+      "loss": 1.3594,
+      "step": 675
+    },
+    {
+      "epoch": 1.0100387542606342,
+      "grad_norm": 0.33533975481987,
+      "learning_rate": 3.3100000000000005e-05,
+      "loss": 1.3268,
+      "step": 676
+    },
+    {
+      "epoch": 1.0115328944296587,
+      "grad_norm": 0.3288819491863251,
+      "learning_rate": 3.3075e-05,
+      "loss": 1.5193,
+      "step": 677
+    },
+    {
+      "epoch": 1.0130270345986834,
+      "grad_norm": 0.49441808462142944,
+      "learning_rate": 3.3050000000000004e-05,
+      "loss": 1.4717,
+      "step": 678
+    },
+    {
+      "epoch": 1.0145211747677079,
+      "grad_norm": 0.47882768511772156,
+      "learning_rate": 3.3025e-05,
+      "loss": 1.3084,
+      "step": 679
+    },
+    {
+      "epoch": 1.0160153149367326,
+      "grad_norm": 0.3494391143321991,
+      "learning_rate": 3.3e-05,
+      "loss": 1.3419,
+      "step": 680
+    },
+    {
+      "epoch": 1.017509455105757,
+      "grad_norm": 0.41978195309638977,
+      "learning_rate": 3.2975e-05,
+      "loss": 1.5436,
+      "step": 681
+    },
+    {
+      "epoch": 1.0190035952747818,
+      "grad_norm": 0.25467464327812195,
+      "learning_rate": 3.295e-05,
+      "loss": 1.3994,
+      "step": 682
+    },
+    {
+      "epoch": 1.0204977354438063,
+      "grad_norm": 0.6259259581565857,
+      "learning_rate": 3.2925e-05,
+      "loss": 1.2119,
+      "step": 683
+    },
+    {
+      "epoch": 1.021991875612831,
+      "grad_norm": 0.3374781012535095,
+      "learning_rate": 3.29e-05,
+      "loss": 1.3365,
+      "step": 684
+    },
+    {
+      "epoch": 1.0234860157818555,
+      "grad_norm": 1.0253171920776367,
+      "learning_rate": 3.2875e-05,
+      "loss": 1.3974,
+      "step": 685
+    },
+    {
+      "epoch": 1.0249801559508802,
+      "grad_norm": 0.27831631898880005,
+      "learning_rate": 3.2850000000000006e-05,
+      "loss": 1.3793,
+      "step": 686
+    },
+    {
+      "epoch": 1.0264742961199047,
+      "grad_norm": 0.2690315842628479,
+      "learning_rate": 3.2825e-05,
+      "loss": 1.4334,
+      "step": 687
+    },
+    {
+      "epoch": 1.0279684362889294,
+      "grad_norm": 0.26815739274024963,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 1.4584,
+      "step": 688
+    },
+    {
+      "epoch": 1.029462576457954,
+      "grad_norm": 0.829439640045166,
+      "learning_rate": 3.2775e-05,
+      "loss": 1.3885,
+      "step": 689
+    },
+    {
+      "epoch": 1.0309567166269786,
+      "grad_norm": 0.29329952597618103,
+      "learning_rate": 3.275e-05,
+      "loss": 1.5287,
+      "step": 690
+    },
+    {
+      "epoch": 1.0324508567960031,
+      "grad_norm": 0.2592604458332062,
+      "learning_rate": 3.2725e-05,
+      "loss": 1.3663,
+      "step": 691
+    },
+    {
+      "epoch": 1.0339449969650278,
+      "grad_norm": 0.4112248718738556,
+      "learning_rate": 3.27e-05,
+      "loss": 1.409,
+      "step": 692
+    },
+    {
+      "epoch": 1.0354391371340523,
+      "grad_norm": 0.4711347222328186,
+      "learning_rate": 3.2675e-05,
+      "loss": 1.5283,
+      "step": 693
+    },
+    {
+      "epoch": 1.036933277303077,
+      "grad_norm": 0.2854495048522949,
+      "learning_rate": 3.265e-05,
+      "loss": 1.1964,
+      "step": 694
+    },
+    {
+      "epoch": 1.0384274174721015,
+      "grad_norm": 0.3027067184448242,
+      "learning_rate": 3.2625e-05,
+      "loss": 1.3459,
+      "step": 695
+    },
+    {
+      "epoch": 1.0399215576411263,
+      "grad_norm": 0.2927369177341461,
+      "learning_rate": 3.26e-05,
+      "loss": 1.3208,
+      "step": 696
+    },
+    {
+      "epoch": 1.0414156978101508,
+      "grad_norm": 0.26334816217422485,
+      "learning_rate": 3.2575e-05,
+      "loss": 1.4442,
+      "step": 697
+    },
+    {
+      "epoch": 1.0429098379791755,
+      "grad_norm": 0.6484261155128479,
+      "learning_rate": 3.2550000000000005e-05,
+      "loss": 1.3031,
+      "step": 698
+    },
+    {
+      "epoch": 1.0444039781482,
+      "grad_norm": 0.32218441367149353,
+      "learning_rate": 3.2525e-05,
+      "loss": 1.2947,
+      "step": 699
+    },
+    {
+      "epoch": 1.0458981183172247,
+      "grad_norm": 1.0116149187088013,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.3189,
+      "step": 700
+    },
+    {
+      "epoch": 1.0473922584862492,
+      "grad_norm": 0.35297858715057373,
+      "learning_rate": 3.2474999999999997e-05,
+      "loss": 1.5605,
+      "step": 701
+    },
+    {
+      "epoch": 1.0488863986552739,
+      "grad_norm": 0.3126513659954071,
+      "learning_rate": 3.245e-05,
+      "loss": 1.5413,
+      "step": 702
+    },
+    {
+      "epoch": 1.0503805388242984,
+      "grad_norm": 0.25015923380851746,
+      "learning_rate": 3.2425e-05,
+      "loss": 1.3003,
+      "step": 703
+    },
+    {
+      "epoch": 1.051874678993323,
+      "grad_norm": 0.2285999357700348,
+      "learning_rate": 3.24e-05,
+      "loss": 1.3829,
+      "step": 704
+    },
+    {
+      "epoch": 1.0533688191623476,
+      "grad_norm": 0.2972055673599243,
+      "learning_rate": 3.2375e-05,
+      "loss": 1.5097,
+      "step": 705
+    },
+    {
+      "epoch": 1.0548629593313723,
+      "grad_norm": 0.3634691834449768,
+      "learning_rate": 3.235e-05,
+      "loss": 1.4746,
+      "step": 706
+    },
+    {
+      "epoch": 1.0563570995003968,
+      "grad_norm": 2.4061403274536133,
+      "learning_rate": 3.2325e-05,
+      "loss": 1.4028,
+      "step": 707
+    },
+    {
+      "epoch": 1.0578512396694215,
+      "grad_norm": 0.5329898595809937,
+      "learning_rate": 3.2300000000000006e-05,
+      "loss": 1.4579,
+      "step": 708
+    },
+    {
+      "epoch": 1.059345379838446,
+      "grad_norm": 0.28464484214782715,
+      "learning_rate": 3.2275e-05,
+      "loss": 1.5056,
+      "step": 709
+    },
+    {
+      "epoch": 1.0608395200074707,
+      "grad_norm": 0.39857912063598633,
+      "learning_rate": 3.2250000000000005e-05,
+      "loss": 1.4314,
+      "step": 710
+    },
+    {
+      "epoch": 1.0623336601764952,
+      "grad_norm": 0.2124282419681549,
+      "learning_rate": 3.2225e-05,
+      "loss": 1.3051,
+      "step": 711
+    },
+    {
+      "epoch": 1.06382780034552,
+      "grad_norm": 0.31750163435935974,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 1.5166,
+      "step": 712
+    },
+    {
+      "epoch": 1.0653219405145444,
+      "grad_norm": 0.25398388504981995,
+      "learning_rate": 3.2175e-05,
+      "loss": 1.3385,
+      "step": 713
+    },
+    {
+      "epoch": 1.0668160806835691,
+      "grad_norm": 0.3630509674549103,
+      "learning_rate": 3.215e-05,
+      "loss": 1.2873,
+      "step": 714
+    },
+    {
+      "epoch": 1.0683102208525936,
+      "grad_norm": 0.3998902440071106,
+      "learning_rate": 3.2125e-05,
+      "loss": 1.4107,
+      "step": 715
+    },
+    {
+      "epoch": 1.0698043610216184,
+      "grad_norm": 0.2822936475276947,
+      "learning_rate": 3.21e-05,
+      "loss": 1.4899,
+      "step": 716
+    },
+    {
+      "epoch": 1.0712985011906428,
+      "grad_norm": 0.2608018219470978,
+      "learning_rate": 3.2075e-05,
+      "loss": 1.2648,
+      "step": 717
+    },
+    {
+      "epoch": 1.0727926413596676,
+      "grad_norm": 0.24499449133872986,
+      "learning_rate": 3.205e-05,
+      "loss": 1.51,
+      "step": 718
+    },
+    {
+      "epoch": 1.074286781528692,
+      "grad_norm": 0.29815685749053955,
+      "learning_rate": 3.2025e-05,
+      "loss": 1.1699,
+      "step": 719
+    },
+    {
+      "epoch": 1.0757809216977168,
+      "grad_norm": 0.38822293281555176,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.4869,
+      "step": 720
+    },
+    {
+      "epoch": 1.0772750618667413,
+      "grad_norm": 0.2638522684574127,
+      "learning_rate": 3.1975e-05,
+      "loss": 1.2973,
+      "step": 721
+    },
+    {
+      "epoch": 1.078769202035766,
+      "grad_norm": 0.5482388734817505,
+      "learning_rate": 3.1950000000000004e-05,
+      "loss": 1.5617,
+      "step": 722
+    },
+    {
+      "epoch": 1.0802633422047907,
+      "grad_norm": 0.3795502185821533,
+      "learning_rate": 3.1925e-05,
+      "loss": 1.3082,
+      "step": 723
+    },
+    {
+      "epoch": 1.0817574823738152,
+      "grad_norm": 0.30303916335105896,
+      "learning_rate": 3.19e-05,
+      "loss": 1.3434,
+      "step": 724
+    },
+    {
+      "epoch": 1.0832516225428397,
+      "grad_norm": 0.44353151321411133,
+      "learning_rate": 3.1875e-05,
+      "loss": 1.211,
+      "step": 725
+    },
+    {
+      "epoch": 1.0847457627118644,
+      "grad_norm": 0.47806358337402344,
+      "learning_rate": 3.185e-05,
+      "loss": 1.3748,
+      "step": 726
+    },
+    {
+      "epoch": 1.0862399028808891,
+      "grad_norm": 0.32267263531684875,
+      "learning_rate": 3.1825e-05,
+      "loss": 1.3756,
+      "step": 727
+    },
+    {
+      "epoch": 1.0877340430499136,
+      "grad_norm": 0.33392220735549927,
+      "learning_rate": 3.18e-05,
+      "loss": 1.4036,
+      "step": 728
+    },
+    {
+      "epoch": 1.0892281832189383,
+      "grad_norm": 0.38585975766181946,
+      "learning_rate": 3.1775e-05,
+      "loss": 1.4745,
+      "step": 729
+    },
+    {
+      "epoch": 1.0907223233879628,
+      "grad_norm": 0.23577117919921875,
+      "learning_rate": 3.175e-05,
+      "loss": 1.273,
+      "step": 730
+    },
+    {
+      "epoch": 1.0922164635569875,
+      "grad_norm": 0.3241201937198639,
+      "learning_rate": 3.1725e-05,
+      "loss": 1.3136,
+      "step": 731
+    },
+    {
+      "epoch": 1.093710603726012,
+      "grad_norm": 0.3119312524795532,
+      "learning_rate": 3.1700000000000005e-05,
+      "loss": 1.3497,
+      "step": 732
+    },
+    {
+      "epoch": 1.0952047438950367,
+      "grad_norm": 0.18843188881874084,
+      "learning_rate": 3.1675e-05,
+      "loss": 1.3864,
+      "step": 733
+    },
+    {
+      "epoch": 1.0966988840640612,
+      "grad_norm": 0.21842527389526367,
+      "learning_rate": 3.1650000000000004e-05,
+      "loss": 1.3344,
+      "step": 734
+    },
+    {
+      "epoch": 1.098193024233086,
+      "grad_norm": 0.28005000948905945,
+      "learning_rate": 3.1624999999999996e-05,
+      "loss": 1.2891,
+      "step": 735
+    },
+    {
+      "epoch": 1.0996871644021105,
+      "grad_norm": 0.23576630651950836,
+      "learning_rate": 3.16e-05,
+      "loss": 1.5126,
+      "step": 736
+    },
+    {
+      "epoch": 1.1011813045711352,
+      "grad_norm": 2.74788236618042,
+      "learning_rate": 3.1575e-05,
+      "loss": 1.365,
+      "step": 737
+    },
+    {
+      "epoch": 1.1026754447401597,
+      "grad_norm": 0.42025816440582275,
+      "learning_rate": 3.155e-05,
+      "loss": 1.3187,
+      "step": 738
+    },
+    {
+      "epoch": 1.1041695849091844,
+      "grad_norm": 0.3011794984340668,
+      "learning_rate": 3.1525e-05,
+      "loss": 1.2736,
+      "step": 739
+    },
+    {
+      "epoch": 1.1056637250782089,
+      "grad_norm": 0.2144736796617508,
+      "learning_rate": 3.15e-05,
+      "loss": 1.4265,
+      "step": 740
+    },
+    {
+      "epoch": 1.1071578652472336,
+      "grad_norm": 1.246840000152588,
+      "learning_rate": 3.1475e-05,
+      "loss": 1.5073,
+      "step": 741
+    },
+    {
+      "epoch": 1.108652005416258,
+      "grad_norm": 0.20965440571308136,
+      "learning_rate": 3.145e-05,
+      "loss": 1.2651,
+      "step": 742
+    },
+    {
+      "epoch": 1.1101461455852828,
+      "grad_norm": 0.3805752694606781,
+      "learning_rate": 3.1425e-05,
+      "loss": 1.4304,
+      "step": 743
+    },
+    {
+      "epoch": 1.1116402857543073,
+      "grad_norm": 0.3760318458080292,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 1.2205,
+      "step": 744
+    },
+    {
+      "epoch": 1.113134425923332,
+      "grad_norm": 0.302948534488678,
+      "learning_rate": 3.1375e-05,
+      "loss": 1.3653,
+      "step": 745
+    },
+    {
+      "epoch": 1.1146285660923565,
+      "grad_norm": 0.5752202868461609,
+      "learning_rate": 3.135e-05,
+      "loss": 1.2555,
+      "step": 746
+    },
+    {
+      "epoch": 1.1161227062613812,
+      "grad_norm": 0.26672235131263733,
+      "learning_rate": 3.1324999999999996e-05,
+      "loss": 1.453,
+      "step": 747
+    },
+    {
+      "epoch": 1.1176168464304057,
+      "grad_norm": 0.261255145072937,
+      "learning_rate": 3.13e-05,
+      "loss": 1.3766,
+      "step": 748
+    },
+    {
+      "epoch": 1.1191109865994304,
+      "grad_norm": 0.3653060495853424,
+      "learning_rate": 3.1275e-05,
+      "loss": 1.4956,
+      "step": 749
+    },
+    {
+      "epoch": 1.120605126768455,
+      "grad_norm": 0.35471010208129883,
+      "learning_rate": 3.125e-05,
+      "loss": 1.3901,
+      "step": 750
+    },
+    {
+      "epoch": 1.1220992669374796,
+      "grad_norm": 0.2914360463619232,
+      "learning_rate": 3.122500000000001e-05,
+      "loss": 1.3893,
+      "step": 751
+    },
+    {
+      "epoch": 1.1235934071065041,
+      "grad_norm": 0.3225819170475006,
+      "learning_rate": 3.12e-05,
+      "loss": 1.2372,
+      "step": 752
+    },
+    {
+      "epoch": 1.1250875472755288,
+      "grad_norm": 0.3456740975379944,
+      "learning_rate": 3.1175000000000006e-05,
+      "loss": 1.4392,
+      "step": 753
+    },
+    {
+      "epoch": 1.1265816874445533,
+      "grad_norm": 0.37935250997543335,
+      "learning_rate": 3.115e-05,
+      "loss": 1.2221,
+      "step": 754
+    },
+    {
+      "epoch": 1.128075827613578,
+      "grad_norm": 0.34086886048316956,
+      "learning_rate": 3.1125000000000004e-05,
+      "loss": 1.4491,
+      "step": 755
+    },
+    {
+      "epoch": 1.1295699677826025,
+      "grad_norm": 0.3890403211116791,
+      "learning_rate": 3.1100000000000004e-05,
+      "loss": 1.3886,
+      "step": 756
+    },
+    {
+      "epoch": 1.1310641079516273,
+      "grad_norm": 0.19871480762958527,
+      "learning_rate": 3.1075e-05,
+      "loss": 1.2324,
+      "step": 757
+    },
+    {
+      "epoch": 1.1325582481206518,
+      "grad_norm": 0.4740760922431946,
+      "learning_rate": 3.105e-05,
+      "loss": 1.5408,
+      "step": 758
+    },
+    {
+      "epoch": 1.1340523882896765,
+      "grad_norm": 0.25011759996414185,
+      "learning_rate": 3.1025e-05,
+      "loss": 1.5476,
+      "step": 759
+    },
+    {
+      "epoch": 1.135546528458701,
+      "grad_norm": 0.23516173660755157,
+      "learning_rate": 3.1e-05,
+      "loss": 1.4722,
+      "step": 760
+    },
+    {
+      "epoch": 1.1370406686277257,
+      "grad_norm": 0.20128604769706726,
+      "learning_rate": 3.0975e-05,
+      "loss": 1.3795,
+      "step": 761
+    },
+    {
+      "epoch": 1.1385348087967502,
+      "grad_norm": 0.21477344632148743,
+      "learning_rate": 3.095e-05,
+      "loss": 1.4503,
+      "step": 762
+    },
+    {
+      "epoch": 1.140028948965775,
+      "grad_norm": 0.3315068483352661,
+      "learning_rate": 3.0925000000000006e-05,
+      "loss": 1.4391,
+      "step": 763
+    },
+    {
+      "epoch": 1.1415230891347994,
+      "grad_norm": 0.30626919865608215,
+      "learning_rate": 3.09e-05,
+      "loss": 1.362,
+      "step": 764
+    },
+    {
+      "epoch": 1.143017229303824,
+      "grad_norm": 0.5180973410606384,
+      "learning_rate": 3.0875000000000005e-05,
+      "loss": 1.3855,
+      "step": 765
+    },
+    {
+      "epoch": 1.1445113694728486,
+      "grad_norm": 0.4765807092189789,
+      "learning_rate": 3.0850000000000004e-05,
+      "loss": 1.4129,
+      "step": 766
+    },
+    {
+      "epoch": 1.1460055096418733,
+      "grad_norm": 0.3589872121810913,
+      "learning_rate": 3.0825000000000004e-05,
+      "loss": 1.4661,
+      "step": 767
+    },
+    {
+      "epoch": 1.1474996498108978,
+      "grad_norm": 0.23404400050640106,
+      "learning_rate": 3.08e-05,
+      "loss": 1.3543,
+      "step": 768
+    },
+    {
+      "epoch": 1.1489937899799225,
+      "grad_norm": 0.34380847215652466,
+      "learning_rate": 3.0775e-05,
+      "loss": 1.4952,
+      "step": 769
+    },
+    {
+      "epoch": 1.150487930148947,
+      "grad_norm": 0.3612429201602936,
+      "learning_rate": 3.075e-05,
+      "loss": 1.1269,
+      "step": 770
+    },
+    {
+      "epoch": 1.1519820703179717,
+      "grad_norm": 0.17686842381954193,
+      "learning_rate": 3.0725e-05,
+      "loss": 1.3392,
+      "step": 771
+    },
+    {
+      "epoch": 1.1534762104869962,
+      "grad_norm": 0.21031558513641357,
+      "learning_rate": 3.07e-05,
+      "loss": 1.3877,
+      "step": 772
+    },
+    {
+      "epoch": 1.154970350656021,
+      "grad_norm": 0.24665507674217224,
+      "learning_rate": 3.067500000000001e-05,
+      "loss": 1.39,
+      "step": 773
+    },
+    {
+      "epoch": 1.1564644908250454,
+      "grad_norm": 0.206196591258049,
+      "learning_rate": 3.065e-05,
+      "loss": 1.4427,
+      "step": 774
+    },
+    {
+      "epoch": 1.1579586309940701,
+      "grad_norm": 0.8704677224159241,
+      "learning_rate": 3.0625000000000006e-05,
+      "loss": 1.6024,
+      "step": 775
+    },
+    {
+      "epoch": 1.1594527711630946,
+      "grad_norm": 0.36794281005859375,
+      "learning_rate": 3.06e-05,
+      "loss": 1.4583,
+      "step": 776
+    },
+    {
+      "epoch": 1.1609469113321194,
+      "grad_norm": 0.38979122042655945,
+      "learning_rate": 3.0575000000000005e-05,
+      "loss": 1.5151,
+      "step": 777
+    },
+    {
+      "epoch": 1.162441051501144,
+      "grad_norm": 0.3037826716899872,
+      "learning_rate": 3.0550000000000004e-05,
+      "loss": 1.4722,
+      "step": 778
+    },
+    {
+      "epoch": 1.1639351916701686,
+      "grad_norm": 1.2709904909133911,
+      "learning_rate": 3.0525e-05,
+      "loss": 1.3094,
+      "step": 779
+    },
+    {
+      "epoch": 1.165429331839193,
+      "grad_norm": 0.3158092498779297,
+      "learning_rate": 3.05e-05,
+      "loss": 1.2846,
+      "step": 780
+    },
+    {
+      "epoch": 1.1669234720082178,
+      "grad_norm": 0.5657904744148254,
+      "learning_rate": 3.0475000000000002e-05,
+      "loss": 1.4704,
+      "step": 781
+    },
+    {
+      "epoch": 1.1684176121772425,
+      "grad_norm": 0.2906689941883087,
+      "learning_rate": 3.045e-05,
+      "loss": 1.1798,
+      "step": 782
+    },
+    {
+      "epoch": 1.169911752346267,
+      "grad_norm": 0.5362401604652405,
+      "learning_rate": 3.0425000000000004e-05,
+      "loss": 1.2779,
+      "step": 783
+    },
+    {
+      "epoch": 1.1714058925152915,
+      "grad_norm": 0.5080185532569885,
+      "learning_rate": 3.04e-05,
+      "loss": 1.5285,
+      "step": 784
+    },
+    {
+      "epoch": 1.1729000326843162,
+      "grad_norm": 0.4617675840854645,
+      "learning_rate": 3.0375000000000003e-05,
+      "loss": 1.2205,
+      "step": 785
+    },
+    {
+      "epoch": 1.174394172853341,
+      "grad_norm": 0.38406652212142944,
+      "learning_rate": 3.035e-05,
+      "loss": 1.3588,
+      "step": 786
+    },
+    {
+      "epoch": 1.1758883130223654,
+      "grad_norm": 0.373355895280838,
+      "learning_rate": 3.0325000000000002e-05,
+      "loss": 1.1637,
+      "step": 787
+    },
+    {
+      "epoch": 1.17738245319139,
+      "grad_norm": 0.5508015155792236,
+      "learning_rate": 3.03e-05,
+      "loss": 1.4598,
+      "step": 788
+    },
+    {
+      "epoch": 1.1788765933604146,
+      "grad_norm": 0.24885353446006775,
+      "learning_rate": 3.0275000000000004e-05,
+      "loss": 1.4141,
+      "step": 789
+    },
+    {
+      "epoch": 1.1803707335294393,
+      "grad_norm": 0.33441323041915894,
+      "learning_rate": 3.025e-05,
+      "loss": 1.3876,
+      "step": 790
+    },
+    {
+      "epoch": 1.1818648736984638,
+      "grad_norm": 0.2754193842411041,
+      "learning_rate": 3.0225000000000003e-05,
+      "loss": 1.6,
+      "step": 791
+    },
+    {
+      "epoch": 1.1833590138674885,
+      "grad_norm": 0.34586068987846375,
+      "learning_rate": 3.02e-05,
+      "loss": 1.5467,
+      "step": 792
+    },
+    {
+      "epoch": 1.184853154036513,
+      "grad_norm": 0.3228543996810913,
+      "learning_rate": 3.0175e-05,
+      "loss": 1.3096,
+      "step": 793
+    },
+    {
+      "epoch": 1.1863472942055378,
+      "grad_norm": 0.3345719575881958,
+      "learning_rate": 3.015e-05,
+      "loss": 1.4846,
+      "step": 794
+    },
+    {
+      "epoch": 1.1878414343745622,
+      "grad_norm": 0.267182320356369,
+      "learning_rate": 3.0125000000000004e-05,
+      "loss": 1.2385,
+      "step": 795
+    },
+    {
+      "epoch": 1.189335574543587,
+      "grad_norm": 0.29065176844596863,
+      "learning_rate": 3.01e-05,
+      "loss": 1.3653,
+      "step": 796
+    },
+    {
+      "epoch": 1.1908297147126115,
+      "grad_norm": 0.3694697916507721,
+      "learning_rate": 3.0075000000000003e-05,
+      "loss": 1.2429,
+      "step": 797
+    },
+    {
+      "epoch": 1.1923238548816362,
+      "grad_norm": 0.3948061466217041,
+      "learning_rate": 3.0050000000000002e-05,
+      "loss": 1.3771,
+      "step": 798
+    },
+    {
+      "epoch": 1.1938179950506607,
+      "grad_norm": 0.24674028158187866,
+      "learning_rate": 3.0025000000000005e-05,
+      "loss": 1.2335,
+      "step": 799
+    },
+    {
+      "epoch": 1.1953121352196854,
+      "grad_norm": 0.32700109481811523,
+      "learning_rate": 3e-05,
+      "loss": 1.3781,
+      "step": 800
+    },
+    {
+      "epoch": 1.1968062753887099,
+      "grad_norm": 0.45519524812698364,
+      "learning_rate": 2.9975000000000004e-05,
+      "loss": 1.2227,
+      "step": 801
+    },
+    {
+      "epoch": 1.1983004155577346,
+      "grad_norm": 0.21541063487529755,
+      "learning_rate": 2.995e-05,
+      "loss": 1.2894,
+      "step": 802
+    },
+    {
+      "epoch": 1.199794555726759,
+      "grad_norm": 0.28006261587142944,
+      "learning_rate": 2.9925000000000002e-05,
+      "loss": 1.4293,
+      "step": 803
+    },
+    {
+      "epoch": 1.2012886958957838,
+      "grad_norm": 0.2731017470359802,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 1.1694,
+      "step": 804
+    },
+    {
+      "epoch": 1.2027828360648083,
+      "grad_norm": 0.22104504704475403,
+      "learning_rate": 2.9875000000000004e-05,
+      "loss": 1.282,
+      "step": 805
+    },
+    {
+      "epoch": 1.204276976233833,
+      "grad_norm": 0.2765119969844818,
+      "learning_rate": 2.985e-05,
+      "loss": 1.214,
+      "step": 806
+    },
+    {
+      "epoch": 1.2057711164028575,
+      "grad_norm": 0.24433647096157074,
+      "learning_rate": 2.9825000000000003e-05,
+      "loss": 1.3806,
+      "step": 807
+    },
+    {
+      "epoch": 1.2072652565718822,
+      "grad_norm": 0.3554052710533142,
+      "learning_rate": 2.98e-05,
+      "loss": 1.2306,
+      "step": 808
+    },
+    {
+      "epoch": 1.2087593967409067,
+      "grad_norm": 0.3093665838241577,
+      "learning_rate": 2.9775000000000002e-05,
+      "loss": 1.2714,
+      "step": 809
+    },
+    {
+      "epoch": 1.2102535369099314,
+      "grad_norm": 0.3403176963329315,
+      "learning_rate": 2.975e-05,
+      "loss": 1.2194,
+      "step": 810
+    },
+    {
+      "epoch": 1.211747677078956,
+      "grad_norm": 0.24468843638896942,
+      "learning_rate": 2.9725000000000004e-05,
+      "loss": 1.3873,
+      "step": 811
+    },
+    {
+      "epoch": 1.2132418172479806,
+      "grad_norm": 0.287300705909729,
+      "learning_rate": 2.97e-05,
+      "loss": 1.3787,
+      "step": 812
+    },
+    {
+      "epoch": 1.2147359574170051,
+      "grad_norm": 0.494862824678421,
+      "learning_rate": 2.9675000000000003e-05,
+      "loss": 1.3717,
+      "step": 813
+    },
+    {
+      "epoch": 1.2162300975860298,
+      "grad_norm": 0.32385581731796265,
+      "learning_rate": 2.965e-05,
+      "loss": 1.3565,
+      "step": 814
+    },
+    {
+      "epoch": 1.2177242377550543,
+      "grad_norm": 0.24930880963802338,
+      "learning_rate": 2.9625000000000002e-05,
+      "loss": 1.414,
+      "step": 815
+    },
+    {
+      "epoch": 1.219218377924079,
+      "grad_norm": 0.6832636594772339,
+      "learning_rate": 2.96e-05,
+      "loss": 1.4534,
+      "step": 816
+    },
+    {
+      "epoch": 1.2207125180931035,
+      "grad_norm": 0.6891420483589172,
+      "learning_rate": 2.9575000000000004e-05,
+      "loss": 1.1974,
+      "step": 817
+    },
+    {
+      "epoch": 1.2222066582621283,
+      "grad_norm": 0.26387810707092285,
+      "learning_rate": 2.955e-05,
+      "loss": 1.5896,
+      "step": 818
+    },
+    {
+      "epoch": 1.2237007984311528,
+      "grad_norm": 0.3951552212238312,
+      "learning_rate": 2.9525000000000003e-05,
+      "loss": 1.6081,
+      "step": 819
+    },
+    {
+      "epoch": 1.2251949386001775,
+      "grad_norm": 0.36233994364738464,
+      "learning_rate": 2.95e-05,
+      "loss": 1.4092,
+      "step": 820
+    },
+    {
+      "epoch": 1.226689078769202,
+      "grad_norm": 0.24323561787605286,
+      "learning_rate": 2.9475e-05,
+      "loss": 1.1833,
+      "step": 821
+    },
+    {
+      "epoch": 1.2281832189382267,
+      "grad_norm": 0.25665760040283203,
+      "learning_rate": 2.945e-05,
+      "loss": 1.3723,
+      "step": 822
+    },
+    {
+      "epoch": 1.2296773591072512,
+      "grad_norm": 0.8483716249465942,
+      "learning_rate": 2.9425000000000004e-05,
+      "loss": 1.3616,
+      "step": 823
+    },
+    {
+      "epoch": 1.231171499276276,
+      "grad_norm": 0.276362806558609,
+      "learning_rate": 2.94e-05,
+      "loss": 1.1512,
+      "step": 824
+    },
+    {
+      "epoch": 1.2326656394453004,
+      "grad_norm": 0.31287723779678345,
+      "learning_rate": 2.9375000000000003e-05,
+      "loss": 1.363,
+      "step": 825
+    },
+    {
+      "epoch": 1.234159779614325,
+      "grad_norm": 0.48258379101753235,
+      "learning_rate": 2.935e-05,
+      "loss": 1.4223,
+      "step": 826
+    },
+    {
+      "epoch": 1.2356539197833496,
+      "grad_norm": 0.3383966088294983,
+      "learning_rate": 2.9325e-05,
+      "loss": 1.3193,
+      "step": 827
+    },
+    {
+      "epoch": 1.2371480599523743,
+      "grad_norm": 0.20877547562122345,
+      "learning_rate": 2.93e-05,
+      "loss": 1.2692,
+      "step": 828
+    },
+    {
+      "epoch": 1.2386422001213988,
+      "grad_norm": 0.21563665568828583,
+      "learning_rate": 2.9275000000000003e-05,
+      "loss": 1.1647,
+      "step": 829
+    },
+    {
+      "epoch": 1.2401363402904235,
+      "grad_norm": 0.3213925361633301,
+      "learning_rate": 2.925e-05,
+      "loss": 1.4321,
+      "step": 830
+    },
+    {
+      "epoch": 1.241630480459448,
+      "grad_norm": 0.40108075737953186,
+      "learning_rate": 2.9225000000000002e-05,
+      "loss": 1.2396,
+      "step": 831
+    },
+    {
+      "epoch": 1.2431246206284727,
+      "grad_norm": 0.30319443345069885,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 1.4283,
+      "step": 832
+    },
+    {
+      "epoch": 1.2446187607974972,
+      "grad_norm": 0.2038179188966751,
+      "learning_rate": 2.9175e-05,
+      "loss": 1.3823,
+      "step": 833
+    },
+    {
+      "epoch": 1.246112900966522,
+      "grad_norm": 0.3502418100833893,
+      "learning_rate": 2.915e-05,
+      "loss": 1.2678,
+      "step": 834
+    },
+    {
+      "epoch": 1.2476070411355464,
+      "grad_norm": 0.21862909197807312,
+      "learning_rate": 2.9125000000000003e-05,
+      "loss": 1.3333,
+      "step": 835
+    },
+    {
+      "epoch": 1.2491011813045712,
+      "grad_norm": 0.3174634873867035,
+      "learning_rate": 2.91e-05,
+      "loss": 1.2475,
+      "step": 836
+    },
+    {
+      "epoch": 1.2505953214735959,
+      "grad_norm": 0.26239484548568726,
+      "learning_rate": 2.9075000000000002e-05,
+      "loss": 1.3712,
+      "step": 837
+    },
+    {
+      "epoch": 1.2520894616426204,
+      "grad_norm": 0.40120023488998413,
+      "learning_rate": 2.9049999999999998e-05,
+      "loss": 1.3346,
+      "step": 838
+    },
+    {
+      "epoch": 1.2535836018116449,
+      "grad_norm": 0.5198585391044617,
+      "learning_rate": 2.9025e-05,
+      "loss": 1.4444,
+      "step": 839
+    },
+    {
+      "epoch": 1.2550777419806696,
+      "grad_norm": 0.34028634428977966,
+      "learning_rate": 2.9e-05,
+      "loss": 1.2473,
+      "step": 840
+    },
+    {
+      "epoch": 1.2565718821496943,
+      "grad_norm": 0.35604262351989746,
+      "learning_rate": 2.8975000000000003e-05,
+      "loss": 1.3676,
+      "step": 841
+    },
+    {
+      "epoch": 1.2580660223187188,
+      "grad_norm": 0.2712811529636383,
+      "learning_rate": 2.895e-05,
+      "loss": 1.4811,
+      "step": 842
+    },
+    {
+      "epoch": 1.2595601624877433,
+      "grad_norm": 0.5156970620155334,
+      "learning_rate": 2.8925000000000002e-05,
+      "loss": 1.2853,
+      "step": 843
+    },
+    {
+      "epoch": 1.261054302656768,
+      "grad_norm": 0.560204803943634,
+      "learning_rate": 2.8899999999999998e-05,
+      "loss": 1.3404,
+      "step": 844
+    },
+    {
+      "epoch": 1.2625484428257927,
+      "grad_norm": 0.3760669529438019,
+      "learning_rate": 2.8875e-05,
+      "loss": 1.4041,
+      "step": 845
+    },
+    {
+      "epoch": 1.2640425829948172,
+      "grad_norm": 0.2032732367515564,
+      "learning_rate": 2.885e-05,
+      "loss": 1.4157,
+      "step": 846
+    },
+    {
+      "epoch": 1.2655367231638417,
+      "grad_norm": 0.5017582774162292,
+      "learning_rate": 2.8825000000000003e-05,
+      "loss": 1.5672,
+      "step": 847
+    },
+    {
+      "epoch": 1.2670308633328664,
+      "grad_norm": 0.3427116572856903,
+      "learning_rate": 2.88e-05,
+      "loss": 1.3806,
+      "step": 848
+    },
+    {
+      "epoch": 1.2685250035018911,
+      "grad_norm": 0.22686722874641418,
+      "learning_rate": 2.8775e-05,
+      "loss": 1.3692,
+      "step": 849
+    },
+    {
+      "epoch": 1.2700191436709156,
+      "grad_norm": 0.252910315990448,
+      "learning_rate": 2.8749999999999997e-05,
+      "loss": 1.3763,
+      "step": 850
+    },
+    {
+      "epoch": 1.2715132838399401,
+      "grad_norm": 0.3212681710720062,
+      "learning_rate": 2.8725e-05,
+      "loss": 1.1975,
+      "step": 851
+    },
+    {
+      "epoch": 1.2730074240089648,
+      "grad_norm": 0.42948630452156067,
+      "learning_rate": 2.87e-05,
+      "loss": 1.6315,
+      "step": 852
+    },
+    {
+      "epoch": 1.2745015641779895,
+      "grad_norm": 0.21065746247768402,
+      "learning_rate": 2.8675000000000002e-05,
+      "loss": 1.3967,
+      "step": 853
+    },
+    {
+      "epoch": 1.275995704347014,
+      "grad_norm": 0.2895759046077728,
+      "learning_rate": 2.865e-05,
+      "loss": 1.3564,
+      "step": 854
+    },
+    {
+      "epoch": 1.2774898445160385,
+      "grad_norm": 0.35186535120010376,
+      "learning_rate": 2.8625e-05,
+      "loss": 1.2381,
+      "step": 855
+    },
+    {
+      "epoch": 1.2789839846850632,
+      "grad_norm": 0.2961024045944214,
+      "learning_rate": 2.86e-05,
+      "loss": 1.4581,
+      "step": 856
+    },
+    {
+      "epoch": 1.280478124854088,
+      "grad_norm": 0.23608747124671936,
+      "learning_rate": 2.8575000000000003e-05,
+      "loss": 1.397,
+      "step": 857
+    },
+    {
+      "epoch": 1.2819722650231125,
+      "grad_norm": 0.24908743798732758,
+      "learning_rate": 2.855e-05,
+      "loss": 1.16,
+      "step": 858
+    },
+    {
+      "epoch": 1.283466405192137,
+      "grad_norm": 5.011353969573975,
+      "learning_rate": 2.8525000000000002e-05,
+      "loss": 1.3649,
+      "step": 859
+    },
+    {
+      "epoch": 1.2849605453611617,
+      "grad_norm": 0.20676688849925995,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.4075,
+      "step": 860
+    },
+    {
+      "epoch": 1.2864546855301864,
+      "grad_norm": 0.43557146191596985,
+      "learning_rate": 2.8475e-05,
+      "loss": 1.3004,
+      "step": 861
+    },
+    {
+      "epoch": 1.2879488256992109,
+      "grad_norm": 0.421140193939209,
+      "learning_rate": 2.845e-05,
+      "loss": 1.3747,
+      "step": 862
+    },
+    {
+      "epoch": 1.2894429658682356,
+      "grad_norm": 0.30302608013153076,
+      "learning_rate": 2.8425000000000003e-05,
+      "loss": 1.4815,
+      "step": 863
+    },
+    {
+      "epoch": 1.29093710603726,
+      "grad_norm": 0.2589489221572876,
+      "learning_rate": 2.84e-05,
+      "loss": 1.3638,
+      "step": 864
+    },
+    {
+      "epoch": 1.2924312462062848,
+      "grad_norm": 0.3756728768348694,
+      "learning_rate": 2.8375000000000002e-05,
+      "loss": 1.2456,
+      "step": 865
+    },
+    {
+      "epoch": 1.2939253863753093,
+      "grad_norm": 0.36108818650245667,
+      "learning_rate": 2.8349999999999998e-05,
+      "loss": 1.5138,
+      "step": 866
+    },
+    {
+      "epoch": 1.295419526544334,
+      "grad_norm": 0.4813844859600067,
+      "learning_rate": 2.8325e-05,
+      "loss": 1.4441,
+      "step": 867
+    },
+    {
+      "epoch": 1.2969136667133585,
+      "grad_norm": 0.31173527240753174,
+      "learning_rate": 2.83e-05,
+      "loss": 1.2089,
+      "step": 868
+    },
+    {
+      "epoch": 1.2984078068823832,
+      "grad_norm": 0.1975783407688141,
+      "learning_rate": 2.8275000000000003e-05,
+      "loss": 1.3155,
+      "step": 869
+    },
+    {
+      "epoch": 1.2999019470514077,
+      "grad_norm": 0.2667165696620941,
+      "learning_rate": 2.825e-05,
+      "loss": 1.2374,
+      "step": 870
+    },
+    {
+      "epoch": 1.3013960872204324,
+      "grad_norm": 0.2623427212238312,
+      "learning_rate": 2.8225e-05,
+      "loss": 1.2428,
+      "step": 871
+    },
+    {
+      "epoch": 1.302890227389457,
+      "grad_norm": 0.38565951585769653,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 1.4143,
+      "step": 872
+    },
+    {
+      "epoch": 1.3043843675584816,
+      "grad_norm": 0.2516254186630249,
+      "learning_rate": 2.8175e-05,
+      "loss": 1.3205,
+      "step": 873
+    },
+    {
+      "epoch": 1.3058785077275061,
+      "grad_norm": 0.2702304422855377,
+      "learning_rate": 2.815e-05,
+      "loss": 1.4655,
+      "step": 874
+    },
+    {
+      "epoch": 1.3073726478965308,
+      "grad_norm": 0.23490573465824127,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 1.3573,
+      "step": 875
+    },
+    {
+      "epoch": 1.3088667880655553,
+      "grad_norm": 0.2213398814201355,
+      "learning_rate": 2.8100000000000005e-05,
+      "loss": 1.4382,
+      "step": 876
+    },
+    {
+      "epoch": 1.31036092823458,
+      "grad_norm": 0.8293248414993286,
+      "learning_rate": 2.8075e-05,
+      "loss": 1.5697,
+      "step": 877
+    },
+    {
+      "epoch": 1.3118550684036046,
+      "grad_norm": 0.3292347490787506,
+      "learning_rate": 2.8050000000000004e-05,
+      "loss": 1.3114,
+      "step": 878
+    },
+    {
+      "epoch": 1.3133492085726293,
+      "grad_norm": 0.59178626537323,
+      "learning_rate": 2.8025e-05,
+      "loss": 1.5344,
+      "step": 879
+    },
+    {
+      "epoch": 1.3148433487416538,
+      "grad_norm": 0.19311776757240295,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.428,
+      "step": 880
+    },
+    {
+      "epoch": 1.3163374889106785,
+      "grad_norm": 0.2914370894432068,
+      "learning_rate": 2.7975000000000002e-05,
+      "loss": 1.2551,
+      "step": 881
+    },
+    {
+      "epoch": 1.317831629079703,
+      "grad_norm": 0.5101747512817383,
+      "learning_rate": 2.7950000000000005e-05,
+      "loss": 1.3906,
+      "step": 882
+    },
+    {
+      "epoch": 1.3193257692487277,
+      "grad_norm": 0.309827983379364,
+      "learning_rate": 2.7925e-05,
+      "loss": 1.2756,
+      "step": 883
+    },
+    {
+      "epoch": 1.3208199094177522,
+      "grad_norm": 0.17280730605125427,
+      "learning_rate": 2.7900000000000004e-05,
+      "loss": 1.2323,
+      "step": 884
+    },
+    {
+      "epoch": 1.322314049586777,
+      "grad_norm": 0.22583577036857605,
+      "learning_rate": 2.7875e-05,
+      "loss": 1.398,
+      "step": 885
+    },
+    {
+      "epoch": 1.3238081897558014,
+      "grad_norm": 0.32121244072914124,
+      "learning_rate": 2.7850000000000003e-05,
+      "loss": 1.2139,
+      "step": 886
+    },
+    {
+      "epoch": 1.325302329924826,
+      "grad_norm": 0.22227728366851807,
+      "learning_rate": 2.7825000000000002e-05,
+      "loss": 1.3756,
+      "step": 887
+    },
+    {
+      "epoch": 1.3267964700938506,
+      "grad_norm": 0.2605989873409271,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 1.3884,
+      "step": 888
+    },
+    {
+      "epoch": 1.3282906102628753,
+      "grad_norm": 0.32979118824005127,
+      "learning_rate": 2.7775e-05,
+      "loss": 1.5079,
+      "step": 889
+    },
+    {
+      "epoch": 1.3297847504318998,
+      "grad_norm": 0.2826998233795166,
+      "learning_rate": 2.7750000000000004e-05,
+      "loss": 1.2673,
+      "step": 890
+    },
+    {
+      "epoch": 1.3312788906009245,
+      "grad_norm": 0.3953245282173157,
+      "learning_rate": 2.7725e-05,
+      "loss": 1.462,
+      "step": 891
+    },
+    {
+      "epoch": 1.3327730307699492,
+      "grad_norm": 0.3536979854106903,
+      "learning_rate": 2.7700000000000002e-05,
+      "loss": 1.4166,
+      "step": 892
+    },
+    {
+      "epoch": 1.3342671709389737,
+      "grad_norm": 0.2883892059326172,
+      "learning_rate": 2.7675000000000002e-05,
+      "loss": 1.4311,
+      "step": 893
+    },
+    {
+      "epoch": 1.3357613111079982,
+      "grad_norm": 0.9811177253723145,
+      "learning_rate": 2.7650000000000005e-05,
+      "loss": 1.559,
+      "step": 894
+    },
+    {
+      "epoch": 1.337255451277023,
+      "grad_norm": 0.2742171883583069,
+      "learning_rate": 2.7625e-05,
+      "loss": 1.388,
+      "step": 895
+    },
+    {
+      "epoch": 1.3387495914460477,
+      "grad_norm": 0.421430766582489,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 1.447,
+      "step": 896
+    },
+    {
+      "epoch": 1.3402437316150722,
+      "grad_norm": 0.21545341610908508,
+      "learning_rate": 2.7575e-05,
+      "loss": 1.3656,
+      "step": 897
+    },
+    {
+      "epoch": 1.3417378717840966,
+      "grad_norm": 0.23525476455688477,
+      "learning_rate": 2.7550000000000002e-05,
+      "loss": 1.548,
+      "step": 898
+    },
+    {
+      "epoch": 1.3432320119531214,
+      "grad_norm": 0.233047753572464,
+      "learning_rate": 2.7525e-05,
+      "loss": 1.4095,
+      "step": 899
+    },
+    {
+      "epoch": 1.344726152122146,
+      "grad_norm": 0.2701248526573181,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 1.2175,
+      "step": 900
+    },
+    {
+      "epoch": 1.3462202922911706,
+      "grad_norm": 0.2926495671272278,
+      "learning_rate": 2.7475e-05,
+      "loss": 1.3209,
+      "step": 901
+    },
+    {
+      "epoch": 1.347714432460195,
+      "grad_norm": 0.24261537194252014,
+      "learning_rate": 2.7450000000000003e-05,
+      "loss": 1.2946,
+      "step": 902
+    },
+    {
+      "epoch": 1.3492085726292198,
+      "grad_norm": 0.22992177307605743,
+      "learning_rate": 2.7425e-05,
+      "loss": 1.4592,
+      "step": 903
+    },
+    {
+      "epoch": 1.3507027127982445,
+      "grad_norm": 0.33490830659866333,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 1.5484,
+      "step": 904
+    },
+    {
+      "epoch": 1.352196852967269,
+      "grad_norm": 0.4046538174152374,
+      "learning_rate": 2.7375e-05,
+      "loss": 1.5581,
+      "step": 905
+    },
+    {
+      "epoch": 1.3536909931362935,
+      "grad_norm": 0.24933142960071564,
+      "learning_rate": 2.7350000000000004e-05,
+      "loss": 1.3697,
+      "step": 906
+    },
+    {
+      "epoch": 1.3551851333053182,
+      "grad_norm": 0.2170824408531189,
+      "learning_rate": 2.7325e-05,
+      "loss": 1.27,
+      "step": 907
+    },
+    {
+      "epoch": 1.356679273474343,
+      "grad_norm": 0.19052165746688843,
+      "learning_rate": 2.7300000000000003e-05,
+      "loss": 1.5267,
+      "step": 908
+    },
+    {
+      "epoch": 1.3581734136433674,
+      "grad_norm": 0.28672656416893005,
+      "learning_rate": 2.7275e-05,
+      "loss": 1.1791,
+      "step": 909
+    },
+    {
+      "epoch": 1.359667553812392,
+      "grad_norm": 0.19672513008117676,
+      "learning_rate": 2.725e-05,
+      "loss": 1.2772,
+      "step": 910
+    },
+    {
+      "epoch": 1.3611616939814166,
+      "grad_norm": 0.24307642877101898,
+      "learning_rate": 2.7225e-05,
+      "loss": 1.5807,
+      "step": 911
+    },
+    {
+      "epoch": 1.3626558341504413,
+      "grad_norm": 0.5582717657089233,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.4225,
+      "step": 912
+    },
+    {
+      "epoch": 1.3641499743194658,
+      "grad_norm": 0.2660907804965973,
+      "learning_rate": 2.7175e-05,
+      "loss": 1.4568,
+      "step": 913
+    },
+    {
+      "epoch": 1.3656441144884903,
+      "grad_norm": 1.315430998802185,
+      "learning_rate": 2.7150000000000003e-05,
+      "loss": 1.4961,
+      "step": 914
+    },
+    {
+      "epoch": 1.367138254657515,
+      "grad_norm": 0.23424766957759857,
+      "learning_rate": 2.7125000000000002e-05,
+      "loss": 1.1405,
+      "step": 915
+    },
+    {
+      "epoch": 1.3686323948265398,
+      "grad_norm": 0.3419889509677887,
+      "learning_rate": 2.7100000000000005e-05,
+      "loss": 1.3257,
+      "step": 916
+    },
+    {
+      "epoch": 1.3701265349955642,
+      "grad_norm": 0.26014795899391174,
+      "learning_rate": 2.7075e-05,
+      "loss": 1.2588,
+      "step": 917
+    },
+    {
+      "epoch": 1.3716206751645887,
+      "grad_norm": 0.25678837299346924,
+      "learning_rate": 2.7050000000000004e-05,
+      "loss": 1.2701,
+      "step": 918
+    },
+    {
+      "epoch": 1.3731148153336135,
+      "grad_norm": 0.252119243144989,
+      "learning_rate": 2.7025e-05,
+      "loss": 1.2999,
+      "step": 919
+    },
+    {
+      "epoch": 1.3746089555026382,
+      "grad_norm": 0.21296586096286774,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.1938,
+      "step": 920
+    },
+    {
+      "epoch": 1.3761030956716627,
+      "grad_norm": 0.3696118891239166,
+      "learning_rate": 2.6975000000000002e-05,
+      "loss": 1.3354,
+      "step": 921
+    },
+    {
+      "epoch": 1.3775972358406874,
+      "grad_norm": 0.24564127624034882,
+      "learning_rate": 2.6950000000000005e-05,
+      "loss": 1.2584,
+      "step": 922
+    },
+    {
+      "epoch": 1.3790913760097119,
+      "grad_norm": 0.3945300877094269,
+      "learning_rate": 2.6925e-05,
+      "loss": 1.3515,
+      "step": 923
+    },
+    {
+      "epoch": 1.3805855161787366,
+      "grad_norm": 0.3990738093852997,
+      "learning_rate": 2.6900000000000003e-05,
+      "loss": 1.2732,
+      "step": 924
+    },
+    {
+      "epoch": 1.382079656347761,
+      "grad_norm": 0.22197787463665009,
+      "learning_rate": 2.6875e-05,
+      "loss": 1.1481,
+      "step": 925
+    },
+    {
+      "epoch": 1.3835737965167858,
+      "grad_norm": 0.24173837900161743,
+      "learning_rate": 2.6850000000000002e-05,
+      "loss": 1.3123,
+      "step": 926
+    },
+    {
+      "epoch": 1.3850679366858103,
+      "grad_norm": 0.21471425890922546,
+      "learning_rate": 2.6825e-05,
+      "loss": 1.3757,
+      "step": 927
+    },
+    {
+      "epoch": 1.386562076854835,
+      "grad_norm": 0.6325602531433105,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 1.397,
+      "step": 928
+    },
+    {
+      "epoch": 1.3880562170238595,
+      "grad_norm": 0.19349870085716248,
+      "learning_rate": 2.6775e-05,
+      "loss": 1.2721,
+      "step": 929
+    },
+    {
+      "epoch": 1.3895503571928842,
+      "grad_norm": 0.25107133388519287,
+      "learning_rate": 2.6750000000000003e-05,
+      "loss": 1.4071,
+      "step": 930
+    },
+    {
+      "epoch": 1.3910444973619087,
+      "grad_norm": 0.20738127827644348,
+      "learning_rate": 2.6725e-05,
+      "loss": 1.3528,
+      "step": 931
+    },
+    {
+      "epoch": 1.3925386375309334,
+      "grad_norm": 0.31351810693740845,
+      "learning_rate": 2.6700000000000002e-05,
+      "loss": 1.2623,
+      "step": 932
+    },
+    {
+      "epoch": 1.394032777699958,
+      "grad_norm": 0.213214710354805,
+      "learning_rate": 2.6675e-05,
+      "loss": 1.2801,
+      "step": 933
+    },
+    {
+      "epoch": 1.3955269178689826,
+      "grad_norm": 0.25812122225761414,
+      "learning_rate": 2.6650000000000004e-05,
+      "loss": 1.2651,
+      "step": 934
+    },
+    {
+      "epoch": 1.3970210580380071,
+      "grad_norm": 0.3310001492500305,
+      "learning_rate": 2.6625e-05,
+      "loss": 1.4703,
+      "step": 935
+    },
+    {
+      "epoch": 1.3985151982070319,
+      "grad_norm": 0.7875789403915405,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 1.4694,
+      "step": 936
+    },
+    {
+      "epoch": 1.4000093383760563,
+      "grad_norm": 0.25254467129707336,
+      "learning_rate": 2.6575e-05,
+      "loss": 1.4147,
+      "step": 937
+    },
+    {
+      "epoch": 1.401503478545081,
+      "grad_norm": 0.24608255922794342,
+      "learning_rate": 2.655e-05,
+      "loss": 1.4127,
+      "step": 938
+    },
+    {
+      "epoch": 1.4029976187141056,
+      "grad_norm": 0.39427649974823,
+      "learning_rate": 2.6525e-05,
+      "loss": 1.2834,
+      "step": 939
+    },
+    {
+      "epoch": 1.4044917588831303,
+      "grad_norm": 0.3120604157447815,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 1.2606,
+      "step": 940
+    },
+    {
+      "epoch": 1.4059858990521548,
+      "grad_norm": 0.5380797982215881,
+      "learning_rate": 2.6475e-05,
+      "loss": 1.2931,
+      "step": 941
+    },
+    {
+      "epoch": 1.4074800392211795,
+      "grad_norm": 0.6395367980003357,
+      "learning_rate": 2.6450000000000003e-05,
+      "loss": 1.2245,
+      "step": 942
+    },
+    {
+      "epoch": 1.408974179390204,
+      "grad_norm": 0.27643725275993347,
+      "learning_rate": 2.6425e-05,
+      "loss": 1.4102,
+      "step": 943
+    },
+    {
+      "epoch": 1.4104683195592287,
+      "grad_norm": 0.31579136848449707,
+      "learning_rate": 2.64e-05,
+      "loss": 1.4558,
+      "step": 944
+    },
+    {
+      "epoch": 1.4119624597282532,
+      "grad_norm": 0.24337553977966309,
+      "learning_rate": 2.6375e-05,
+      "loss": 1.4894,
+      "step": 945
+    },
+    {
+      "epoch": 1.413456599897278,
+      "grad_norm": 0.40405669808387756,
+      "learning_rate": 2.6350000000000004e-05,
+      "loss": 1.4032,
+      "step": 946
+    },
+    {
+      "epoch": 1.4149507400663024,
+      "grad_norm": 0.5983033776283264,
+      "learning_rate": 2.6325e-05,
+      "loss": 1.3462,
+      "step": 947
+    },
+    {
+      "epoch": 1.416444880235327,
+      "grad_norm": 0.2644513249397278,
+      "learning_rate": 2.6300000000000002e-05,
+      "loss": 1.5256,
+      "step": 948
+    },
+    {
+      "epoch": 1.4179390204043516,
+      "grad_norm": 0.37031152844429016,
+      "learning_rate": 2.6275e-05,
+      "loss": 1.4239,
+      "step": 949
+    },
+    {
+      "epoch": 1.4194331605733763,
+      "grad_norm": 0.23210497200489044,
+      "learning_rate": 2.625e-05,
+      "loss": 1.4778,
+      "step": 950
+    },
+    {
+      "epoch": 1.420927300742401,
+      "grad_norm": 0.21914657950401306,
+      "learning_rate": 2.6225e-05,
+      "loss": 1.4029,
+      "step": 951
+    },
+    {
+      "epoch": 1.4224214409114255,
+      "grad_norm": 0.26754164695739746,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 1.2504,
+      "step": 952
+    },
+    {
+      "epoch": 1.42391558108045,
+      "grad_norm": 0.22635582089424133,
+      "learning_rate": 2.6175e-05,
+      "loss": 1.2113,
+      "step": 953
+    },
+    {
+      "epoch": 1.4254097212494747,
+      "grad_norm": 0.22673365473747253,
+      "learning_rate": 2.6150000000000002e-05,
+      "loss": 1.3774,
+      "step": 954
+    },
+    {
+      "epoch": 1.4269038614184995,
+      "grad_norm": 0.39390039443969727,
+      "learning_rate": 2.6124999999999998e-05,
+      "loss": 1.5364,
+      "step": 955
+    },
+    {
+      "epoch": 1.428398001587524,
+      "grad_norm": 0.2662557065486908,
+      "learning_rate": 2.61e-05,
+      "loss": 1.4081,
+      "step": 956
+    },
+    {
+      "epoch": 1.4298921417565484,
+      "grad_norm": 1.5610660314559937,
+      "learning_rate": 2.6075e-05,
+      "loss": 1.3799,
+      "step": 957
+    },
+    {
+      "epoch": 1.4313862819255732,
+      "grad_norm": 0.33333706855773926,
+      "learning_rate": 2.6050000000000003e-05,
+      "loss": 1.3966,
+      "step": 958
+    },
+    {
+      "epoch": 1.4328804220945979,
+      "grad_norm": 0.29869765043258667,
+      "learning_rate": 2.6025e-05,
+      "loss": 1.3118,
+      "step": 959
+    },
+    {
+      "epoch": 1.4343745622636224,
+      "grad_norm": 0.26354825496673584,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.3255,
+      "step": 960
+    },
+    {
+      "epoch": 1.4358687024326469,
+      "grad_norm": 0.41859930753707886,
+      "learning_rate": 2.5974999999999998e-05,
+      "loss": 1.3504,
+      "step": 961
+    },
+    {
+      "epoch": 1.4373628426016716,
+      "grad_norm": 0.21529413759708405,
+      "learning_rate": 2.595e-05,
+      "loss": 1.2007,
+      "step": 962
+    },
+    {
+      "epoch": 1.4388569827706963,
+      "grad_norm": 0.20801830291748047,
+      "learning_rate": 2.5925e-05,
+      "loss": 1.4386,
+      "step": 963
+    },
+    {
+      "epoch": 1.4403511229397208,
+      "grad_norm": 0.3053770661354065,
+      "learning_rate": 2.5900000000000003e-05,
+      "loss": 1.254,
+      "step": 964
+    },
+    {
+      "epoch": 1.4418452631087453,
+      "grad_norm": 0.2524198889732361,
+      "learning_rate": 2.5875e-05,
+      "loss": 1.3625,
+      "step": 965
+    },
+    {
+      "epoch": 1.44333940327777,
+      "grad_norm": 0.27251192927360535,
+      "learning_rate": 2.585e-05,
+      "loss": 1.2657,
+      "step": 966
+    },
+    {
+      "epoch": 1.4448335434467947,
+      "grad_norm": 0.2282276451587677,
+      "learning_rate": 2.5824999999999998e-05,
+      "loss": 1.2054,
+      "step": 967
+    },
+    {
+      "epoch": 1.4463276836158192,
+      "grad_norm": 0.2785608768463135,
+      "learning_rate": 2.58e-05,
+      "loss": 1.4328,
+      "step": 968
+    },
+    {
+      "epoch": 1.4478218237848437,
+      "grad_norm": 0.3191685378551483,
+      "learning_rate": 2.5775e-05,
+      "loss": 1.3032,
+      "step": 969
+    },
+    {
+      "epoch": 1.4493159639538684,
+      "grad_norm": 0.26658880710601807,
+      "learning_rate": 2.5750000000000002e-05,
+      "loss": 1.436,
+      "step": 970
+    },
+    {
+      "epoch": 1.4508101041228931,
+      "grad_norm": 0.532778263092041,
+      "learning_rate": 2.5725e-05,
+      "loss": 1.4724,
+      "step": 971
+    },
+    {
+      "epoch": 1.4523042442919176,
+      "grad_norm": 0.2609519958496094,
+      "learning_rate": 2.57e-05,
+      "loss": 1.3608,
+      "step": 972
+    },
+    {
+      "epoch": 1.4537983844609421,
+      "grad_norm": 0.2050527185201645,
+      "learning_rate": 2.5675e-05,
+      "loss": 1.3276,
+      "step": 973
+    },
+    {
+      "epoch": 1.4552925246299668,
+      "grad_norm": 0.23767416179180145,
+      "learning_rate": 2.5650000000000003e-05,
+      "loss": 1.3668,
+      "step": 974
+    },
+    {
+      "epoch": 1.4567866647989915,
+      "grad_norm": 0.43234944343566895,
+      "learning_rate": 2.5625e-05,
+      "loss": 1.4892,
+      "step": 975
+    },
+    {
+      "epoch": 1.458280804968016,
+      "grad_norm": 0.2602043151855469,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.4824,
+      "step": 976
+    },
+    {
+      "epoch": 1.4597749451370405,
+      "grad_norm": 0.2256672978401184,
+      "learning_rate": 2.5574999999999998e-05,
+      "loss": 1.4242,
+      "step": 977
+    },
+    {
+      "epoch": 1.4612690853060653,
+      "grad_norm": 0.24697938561439514,
+      "learning_rate": 2.555e-05,
+      "loss": 1.2936,
+      "step": 978
+    },
+    {
+      "epoch": 1.46276322547509,
+      "grad_norm": 0.3508719205856323,
+      "learning_rate": 2.5525e-05,
+      "loss": 1.2935,
+      "step": 979
+    },
+    {
+      "epoch": 1.4642573656441145,
+      "grad_norm": 0.245631143450737,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 1.3749,
+      "step": 980
+    },
+    {
+      "epoch": 1.4657515058131392,
+      "grad_norm": 0.34013769030570984,
+      "learning_rate": 2.5475e-05,
+      "loss": 1.2343,
+      "step": 981
+    },
+    {
+      "epoch": 1.4672456459821637,
+      "grad_norm": 0.8150389790534973,
+      "learning_rate": 2.5450000000000002e-05,
+      "loss": 1.2601,
+      "step": 982
+    },
+    {
+      "epoch": 1.4687397861511884,
+      "grad_norm": 0.25743216276168823,
+      "learning_rate": 2.5424999999999998e-05,
+      "loss": 1.1799,
+      "step": 983
+    },
+    {
+      "epoch": 1.4702339263202129,
+      "grad_norm": 0.429471492767334,
+      "learning_rate": 2.54e-05,
+      "loss": 1.5372,
+      "step": 984
+    },
+    {
+      "epoch": 1.4717280664892376,
+      "grad_norm": 0.294306218624115,
+      "learning_rate": 2.5375e-05,
+      "loss": 1.2718,
+      "step": 985
+    },
+    {
+      "epoch": 1.473222206658262,
+      "grad_norm": 0.22676728665828705,
+      "learning_rate": 2.5350000000000003e-05,
+      "loss": 1.1818,
+      "step": 986
+    },
+    {
+      "epoch": 1.4747163468272868,
+      "grad_norm": 0.25699350237846375,
+      "learning_rate": 2.5325e-05,
+      "loss": 1.2167,
+      "step": 987
+    },
+    {
+      "epoch": 1.4762104869963113,
+      "grad_norm": 0.23200049996376038,
+      "learning_rate": 2.5300000000000002e-05,
+      "loss": 1.3309,
+      "step": 988
+    },
+    {
+      "epoch": 1.477704627165336,
+      "grad_norm": 0.3287791311740875,
+      "learning_rate": 2.5274999999999998e-05,
+      "loss": 1.3764,
+      "step": 989
+    },
+    {
+      "epoch": 1.4791987673343605,
+      "grad_norm": 0.5922759771347046,
+      "learning_rate": 2.525e-05,
+      "loss": 1.4138,
+      "step": 990
+    },
+    {
+      "epoch": 1.4806929075033852,
+      "grad_norm": 0.21253745257854462,
+      "learning_rate": 2.5225e-05,
+      "loss": 1.151,
+      "step": 991
+    },
+    {
+      "epoch": 1.4821870476724097,
+      "grad_norm": 0.3864874839782715,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 1.3244,
+      "step": 992
+    },
+    {
+      "epoch": 1.4836811878414344,
+      "grad_norm": 0.30087751150131226,
+      "learning_rate": 2.5175e-05,
+      "loss": 1.3481,
+      "step": 993
+    },
+    {
+      "epoch": 1.485175328010459,
+      "grad_norm": 0.2523193657398224,
+      "learning_rate": 2.515e-05,
+      "loss": 1.3509,
+      "step": 994
+    },
+    {
+      "epoch": 1.4866694681794836,
+      "grad_norm": 0.19127090275287628,
+      "learning_rate": 2.5124999999999997e-05,
+      "loss": 1.379,
+      "step": 995
+    },
+    {
+      "epoch": 1.4881636083485081,
+      "grad_norm": 0.5360708832740784,
+      "learning_rate": 2.51e-05,
+      "loss": 1.1747,
+      "step": 996
+    },
+    {
+      "epoch": 1.4896577485175329,
+      "grad_norm": 0.17975042760372162,
+      "learning_rate": 2.5075e-05,
+      "loss": 1.1739,
+      "step": 997
+    },
+    {
+      "epoch": 1.4911518886865573,
+      "grad_norm": 0.4710841476917267,
+      "learning_rate": 2.5050000000000002e-05,
+      "loss": 1.4195,
+      "step": 998
+    },
+    {
+      "epoch": 1.492646028855582,
+      "grad_norm": 0.2385946810245514,
+      "learning_rate": 2.5025e-05,
+      "loss": 1.3315,
+      "step": 999
+    },
+    {
+      "epoch": 1.4941401690246066,
+      "grad_norm": 0.424209326505661,
+      "learning_rate": 2.5e-05,
+      "loss": 1.2433,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4956343091936313,
+      "grad_norm": 0.34269607067108154,
+      "learning_rate": 2.4975e-05,
+      "loss": 1.4392,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4971284493626558,
+      "grad_norm": 1.2078485488891602,
+      "learning_rate": 2.495e-05,
+      "loss": 1.3833,
+      "step": 1002
+    },
+    {
+      "epoch": 1.4986225895316805,
+      "grad_norm": 0.3181706666946411,
+      "learning_rate": 2.4925000000000003e-05,
+      "loss": 1.3839,
+      "step": 1003
+    },
+    {
+      "epoch": 1.500116729700705,
+      "grad_norm": 0.4560883045196533,
+      "learning_rate": 2.4900000000000002e-05,
+      "loss": 1.48,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5016108698697297,
+      "grad_norm": 0.5617378950119019,
+      "learning_rate": 2.4875e-05,
+      "loss": 1.216,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5031050100387544,
+      "grad_norm": 0.28854793310165405,
+      "learning_rate": 2.485e-05,
+      "loss": 1.2834,
+      "step": 1006
+    },
+    {
+      "epoch": 1.504599150207779,
+      "grad_norm": 0.29387474060058594,
+      "learning_rate": 2.4825e-05,
+      "loss": 1.0273,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5060932903768034,
+      "grad_norm": 1.235810399055481,
+      "learning_rate": 2.48e-05,
+      "loss": 1.4177,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5075874305458281,
+      "grad_norm": 0.2525739371776581,
+      "learning_rate": 2.4775000000000003e-05,
+      "loss": 1.4798,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5090815707148528,
+      "grad_norm": 0.437465101480484,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 1.6346,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5105757108838773,
+      "grad_norm": 0.19477544724941254,
+      "learning_rate": 2.4725e-05,
+      "loss": 1.2425,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5120698510529018,
+      "grad_norm": 0.23110416531562805,
+      "learning_rate": 2.47e-05,
+      "loss": 1.2046,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5135639912219265,
+      "grad_norm": 0.2584448456764221,
+      "learning_rate": 2.4675e-05,
+      "loss": 1.2616,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5150581313909512,
+      "grad_norm": 0.24508026242256165,
+      "learning_rate": 2.465e-05,
+      "loss": 1.5643,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5165522715599757,
+      "grad_norm": 0.6831825375556946,
+      "learning_rate": 2.4625000000000002e-05,
+      "loss": 1.3956,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5180464117290002,
+      "grad_norm": 0.2081739604473114,
+      "learning_rate": 2.46e-05,
+      "loss": 1.1868,
+      "step": 1016
+    },
+    {
+      "epoch": 1.519540551898025,
+      "grad_norm": 0.36565127968788147,
+      "learning_rate": 2.4575e-05,
+      "loss": 1.3864,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5210346920670497,
+      "grad_norm": 1.2851471900939941,
+      "learning_rate": 2.455e-05,
+      "loss": 1.3694,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5225288322360742,
+      "grad_norm": 0.24699068069458008,
+      "learning_rate": 2.4525e-05,
+      "loss": 1.4973,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5240229724050987,
+      "grad_norm": 0.25761643052101135,
+      "learning_rate": 2.45e-05,
+      "loss": 1.0198,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5255171125741234,
+      "grad_norm": 0.24483232200145721,
+      "learning_rate": 2.4475000000000002e-05,
+      "loss": 1.2763,
+      "step": 1021
+    },
+    {
+      "epoch": 1.527011252743148,
+      "grad_norm": 0.23036424815654755,
+      "learning_rate": 2.445e-05,
+      "loss": 1.2409,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5285053929121726,
+      "grad_norm": 0.3976009786128998,
+      "learning_rate": 2.4425e-05,
+      "loss": 1.2585,
+      "step": 1023
+    },
+    {
+      "epoch": 1.529999533081197,
+      "grad_norm": 0.21544378995895386,
+      "learning_rate": 2.44e-05,
+      "loss": 1.5668,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5314936732502218,
+      "grad_norm": 0.19233135879039764,
+      "learning_rate": 2.4375e-05,
+      "loss": 1.3254,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5329878134192465,
+      "grad_norm": 0.22133982181549072,
+      "learning_rate": 2.435e-05,
+      "loss": 1.3993,
+      "step": 1026
+    },
+    {
+      "epoch": 1.534481953588271,
+      "grad_norm": 0.324274480342865,
+      "learning_rate": 2.4325000000000002e-05,
+      "loss": 1.4625,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5359760937572955,
+      "grad_norm": 0.21806937456130981,
+      "learning_rate": 2.43e-05,
+      "loss": 1.3523,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5374702339263202,
+      "grad_norm": 0.23030443489551544,
+      "learning_rate": 2.4275e-05,
+      "loss": 1.2272,
+      "step": 1029
+    },
+    {
+      "epoch": 1.538964374095345,
+      "grad_norm": 0.37339186668395996,
+      "learning_rate": 2.425e-05,
+      "loss": 1.3109,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5404585142643694,
+      "grad_norm": 0.26914238929748535,
+      "learning_rate": 2.4225e-05,
+      "loss": 1.2883,
+      "step": 1031
+    },
+    {
+      "epoch": 1.541952654433394,
+      "grad_norm": 0.20728585124015808,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 1.2369,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5434467946024186,
+      "grad_norm": 0.4167015254497528,
+      "learning_rate": 2.4175e-05,
+      "loss": 1.36,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5449409347714433,
+      "grad_norm": 0.255819171667099,
+      "learning_rate": 2.415e-05,
+      "loss": 1.5077,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5464350749404678,
+      "grad_norm": 0.2544974386692047,
+      "learning_rate": 2.4125e-05,
+      "loss": 1.3009,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5479292151094923,
+      "grad_norm": 0.27418991923332214,
+      "learning_rate": 2.41e-05,
+      "loss": 1.425,
+      "step": 1036
+    },
+    {
+      "epoch": 1.549423355278517,
+      "grad_norm": 0.5164713263511658,
+      "learning_rate": 2.4075e-05,
+      "loss": 1.5062,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5509174954475418,
+      "grad_norm": 0.19975262880325317,
+      "learning_rate": 2.4050000000000002e-05,
+      "loss": 1.3069,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5524116356165663,
+      "grad_norm": 0.2381182760000229,
+      "learning_rate": 2.4025e-05,
+      "loss": 1.3012,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5539057757855907,
+      "grad_norm": 0.2583548426628113,
+      "learning_rate": 2.4e-05,
+      "loss": 1.1985,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5553999159546155,
+      "grad_norm": 0.2762662470340729,
+      "learning_rate": 2.3975e-05,
+      "loss": 1.354,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5568940561236402,
+      "grad_norm": 0.2436295598745346,
+      "learning_rate": 2.395e-05,
+      "loss": 1.5246,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5583881962926647,
+      "grad_norm": 0.24542199075222015,
+      "learning_rate": 2.3925e-05,
+      "loss": 1.4547,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5598823364616892,
+      "grad_norm": 0.27998894453048706,
+      "learning_rate": 2.39e-05,
+      "loss": 1.337,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5613764766307139,
+      "grad_norm": 0.2425457239151001,
+      "learning_rate": 2.3875e-05,
+      "loss": 1.219,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5628706167997386,
+      "grad_norm": 0.26620247960090637,
+      "learning_rate": 2.385e-05,
+      "loss": 1.4698,
+      "step": 1046
+    },
+    {
+      "epoch": 1.564364756968763,
+      "grad_norm": 0.2812005579471588,
+      "learning_rate": 2.3825e-05,
+      "loss": 1.2821,
+      "step": 1047
+    },
+    {
+      "epoch": 1.5658588971377876,
+      "grad_norm": 0.24852269887924194,
+      "learning_rate": 2.38e-05,
+      "loss": 1.2905,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5673530373068123,
+      "grad_norm": 0.2867446541786194,
+      "learning_rate": 2.3775e-05,
+      "loss": 1.3753,
+      "step": 1049
+    },
+    {
+      "epoch": 1.568847177475837,
+      "grad_norm": 0.18901658058166504,
+      "learning_rate": 2.375e-05,
+      "loss": 1.2445,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5703413176448615,
+      "grad_norm": 0.2708028256893158,
+      "learning_rate": 2.3725e-05,
+      "loss": 1.3819,
+      "step": 1051
+    },
+    {
+      "epoch": 1.571835457813886,
+      "grad_norm": 0.2241097092628479,
+      "learning_rate": 2.37e-05,
+      "loss": 1.2094,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5733295979829107,
+      "grad_norm": 0.22352498769760132,
+      "learning_rate": 2.3675e-05,
+      "loss": 1.2761,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5748237381519354,
+      "grad_norm": 0.28223153948783875,
+      "learning_rate": 2.365e-05,
+      "loss": 1.2256,
+      "step": 1054
+    },
+    {
+      "epoch": 1.57631787832096,
+      "grad_norm": 0.17800620198249817,
+      "learning_rate": 2.3624999999999998e-05,
+      "loss": 1.1712,
+      "step": 1055
+    },
+    {
+      "epoch": 1.5778120184899846,
+      "grad_norm": 0.39222070574760437,
+      "learning_rate": 2.36e-05,
+      "loss": 1.5474,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5793061586590094,
+      "grad_norm": 0.24694950878620148,
+      "learning_rate": 2.3575e-05,
+      "loss": 1.4975,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5808002988280339,
+      "grad_norm": 0.2800028622150421,
+      "learning_rate": 2.355e-05,
+      "loss": 1.4577,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5822944389970584,
+      "grad_norm": 0.22472792863845825,
+      "learning_rate": 2.3525e-05,
+      "loss": 1.2913,
+      "step": 1059
+    },
+    {
+      "epoch": 1.583788579166083,
+      "grad_norm": 0.37456855177879333,
+      "learning_rate": 2.35e-05,
+      "loss": 1.5394,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5852827193351078,
+      "grad_norm": 0.25892317295074463,
+      "learning_rate": 2.3475e-05,
+      "loss": 1.3467,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5867768595041323,
+      "grad_norm": 0.2283022403717041,
+      "learning_rate": 2.345e-05,
+      "loss": 1.421,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5882709996731568,
+      "grad_norm": 0.4213232398033142,
+      "learning_rate": 2.3425000000000004e-05,
+      "loss": 1.1265,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5897651398421815,
+      "grad_norm": 0.1526349037885666,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 1.2766,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5912592800112062,
+      "grad_norm": 0.3344525098800659,
+      "learning_rate": 2.3375000000000002e-05,
+      "loss": 1.2623,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5927534201802307,
+      "grad_norm": 0.2371218055486679,
+      "learning_rate": 2.3350000000000002e-05,
+      "loss": 1.4743,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5942475603492552,
+      "grad_norm": 0.2002929300069809,
+      "learning_rate": 2.3325e-05,
+      "loss": 1.301,
+      "step": 1067
+    },
+    {
+      "epoch": 1.59574170051828,
+      "grad_norm": 0.24492183327674866,
+      "learning_rate": 2.3300000000000004e-05,
+      "loss": 1.3952,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5972358406873046,
+      "grad_norm": 0.36826092004776,
+      "learning_rate": 2.3275000000000003e-05,
+      "loss": 1.1323,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5987299808563291,
+      "grad_norm": 0.22465544939041138,
+      "learning_rate": 2.3250000000000003e-05,
+      "loss": 1.482,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6002241210253536,
+      "grad_norm": 0.40929779410362244,
+      "learning_rate": 2.3225000000000002e-05,
+      "loss": 1.3384,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6017182611943783,
+      "grad_norm": 0.26005005836486816,
+      "learning_rate": 2.32e-05,
+      "loss": 1.2387,
+      "step": 1072
+    },
+    {
+      "epoch": 1.603212401363403,
+      "grad_norm": 0.2737712264060974,
+      "learning_rate": 2.3175e-05,
+      "loss": 1.3755,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6047065415324275,
+      "grad_norm": 0.20897279679775238,
+      "learning_rate": 2.3150000000000004e-05,
+      "loss": 1.5051,
+      "step": 1074
+    },
+    {
+      "epoch": 1.606200681701452,
+      "grad_norm": 0.23773355782032013,
+      "learning_rate": 2.3125000000000003e-05,
+      "loss": 1.2317,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6076948218704767,
+      "grad_norm": 0.17421811819076538,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 1.3285,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6091889620395015,
+      "grad_norm": 0.242172509431839,
+      "learning_rate": 2.3075000000000002e-05,
+      "loss": 1.3524,
+      "step": 1077
+    },
+    {
+      "epoch": 1.610683102208526,
+      "grad_norm": 0.1881755292415619,
+      "learning_rate": 2.305e-05,
+      "loss": 1.3131,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6121772423775504,
+      "grad_norm": 0.39744487404823303,
+      "learning_rate": 2.3025e-05,
+      "loss": 1.4211,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6136713825465752,
+      "grad_norm": 0.28824132680892944,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.4354,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6151655227155999,
+      "grad_norm": 0.23316174745559692,
+      "learning_rate": 2.2975000000000003e-05,
+      "loss": 1.3879,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6166596628846244,
+      "grad_norm": 0.20622767508029938,
+      "learning_rate": 2.2950000000000002e-05,
+      "loss": 1.4856,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6181538030536489,
+      "grad_norm": 0.26608648896217346,
+      "learning_rate": 2.2925e-05,
+      "loss": 1.3223,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6196479432226736,
+      "grad_norm": 0.38514405488967896,
+      "learning_rate": 2.29e-05,
+      "loss": 1.2364,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6211420833916983,
+      "grad_norm": 0.4655008912086487,
+      "learning_rate": 2.2875e-05,
+      "loss": 1.283,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6226362235607228,
+      "grad_norm": 0.334542453289032,
+      "learning_rate": 2.2850000000000003e-05,
+      "loss": 1.4633,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6241303637297473,
+      "grad_norm": 0.2390364110469818,
+      "learning_rate": 2.2825000000000003e-05,
+      "loss": 1.4632,
+      "step": 1087
+    },
+    {
+      "epoch": 1.625624503898772,
+      "grad_norm": 0.2931368350982666,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 1.3178,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6271186440677967,
+      "grad_norm": 0.2291143387556076,
+      "learning_rate": 2.2775e-05,
+      "loss": 1.4035,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6286127842368212,
+      "grad_norm": 0.21086153388023376,
+      "learning_rate": 2.275e-05,
+      "loss": 1.4254,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6301069244058457,
+      "grad_norm": 0.20831753313541412,
+      "learning_rate": 2.2725000000000003e-05,
+      "loss": 1.332,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6316010645748704,
+      "grad_norm": 0.3618925213813782,
+      "learning_rate": 2.2700000000000003e-05,
+      "loss": 1.2239,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6330952047438951,
+      "grad_norm": 0.3585363030433655,
+      "learning_rate": 2.2675000000000002e-05,
+      "loss": 1.2772,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6345893449129196,
+      "grad_norm": 0.24144862592220306,
+      "learning_rate": 2.265e-05,
+      "loss": 1.4067,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6360834850819441,
+      "grad_norm": 0.3754459619522095,
+      "learning_rate": 2.2625e-05,
+      "loss": 1.466,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6375776252509688,
+      "grad_norm": 0.21260520815849304,
+      "learning_rate": 2.26e-05,
+      "loss": 1.1853,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6390717654199936,
+      "grad_norm": 0.21034535765647888,
+      "learning_rate": 2.2575000000000003e-05,
+      "loss": 1.1168,
+      "step": 1097
+    },
+    {
+      "epoch": 1.640565905589018,
+      "grad_norm": 0.23104006052017212,
+      "learning_rate": 2.2550000000000003e-05,
+      "loss": 1.2363,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6420600457580425,
+      "grad_norm": 0.23428553342819214,
+      "learning_rate": 2.2525000000000002e-05,
+      "loss": 1.3021,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6435541859270673,
+      "grad_norm": 0.4666197896003723,
+      "learning_rate": 2.25e-05,
+      "loss": 1.4213,
+      "step": 1100
+    },
+    {
+      "epoch": 1.645048326096092,
+      "grad_norm": 0.6136429309844971,
+      "learning_rate": 2.2475e-05,
+      "loss": 1.2789,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6465424662651165,
+      "grad_norm": 0.5505061149597168,
+      "learning_rate": 2.245e-05,
+      "loss": 1.3622,
+      "step": 1102
+    },
+    {
+      "epoch": 1.648036606434141,
+      "grad_norm": 0.20704680681228638,
+      "learning_rate": 2.2425000000000003e-05,
+      "loss": 1.3519,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6495307466031657,
+      "grad_norm": 0.24088571965694427,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.2892,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6510248867721904,
+      "grad_norm": 0.6935633420944214,
+      "learning_rate": 2.2375000000000002e-05,
+      "loss": 1.332,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6525190269412149,
+      "grad_norm": 0.2827127277851105,
+      "learning_rate": 2.235e-05,
+      "loss": 1.3024,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6540131671102394,
+      "grad_norm": 0.25590217113494873,
+      "learning_rate": 2.2325e-05,
+      "loss": 1.3248,
+      "step": 1107
+    },
+    {
+      "epoch": 1.655507307279264,
+      "grad_norm": 0.19156008958816528,
+      "learning_rate": 2.23e-05,
+      "loss": 1.4835,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6570014474482888,
+      "grad_norm": 0.18704158067703247,
+      "learning_rate": 2.2275000000000003e-05,
+      "loss": 1.2183,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6584955876173133,
+      "grad_norm": 0.21460570394992828,
+      "learning_rate": 2.2250000000000002e-05,
+      "loss": 1.2204,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6599897277863378,
+      "grad_norm": 0.4342014789581299,
+      "learning_rate": 2.2225e-05,
+      "loss": 1.5425,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6614838679553625,
+      "grad_norm": 0.25505074858665466,
+      "learning_rate": 2.22e-05,
+      "loss": 1.2719,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6629780081243872,
+      "grad_norm": 0.4050869047641754,
+      "learning_rate": 2.2175e-05,
+      "loss": 1.2727,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6644721482934117,
+      "grad_norm": 6.226749897003174,
+      "learning_rate": 2.215e-05,
+      "loss": 1.149,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6659662884624362,
+      "grad_norm": 0.17852473258972168,
+      "learning_rate": 2.2125000000000002e-05,
+      "loss": 1.3976,
+      "step": 1115
+    },
+    {
+      "epoch": 1.667460428631461,
+      "grad_norm": 0.3547992706298828,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 1.2597,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6689545688004856,
+      "grad_norm": 0.285022109746933,
+      "learning_rate": 2.2075e-05,
+      "loss": 1.4945,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6704487089695101,
+      "grad_norm": 0.2697555422782898,
+      "learning_rate": 2.205e-05,
+      "loss": 1.2034,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6719428491385349,
+      "grad_norm": 0.2486453652381897,
+      "learning_rate": 2.2025e-05,
+      "loss": 1.3333,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6734369893075596,
+      "grad_norm": 0.3298685550689697,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.2115,
+      "step": 1120
+    },
+    {
+      "epoch": 1.674931129476584,
+      "grad_norm": 0.2030079960823059,
+      "learning_rate": 2.1975000000000002e-05,
+      "loss": 1.4112,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6764252696456086,
+      "grad_norm": 0.32204288244247437,
+      "learning_rate": 2.195e-05,
+      "loss": 1.3559,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6779194098146333,
+      "grad_norm": 0.22706016898155212,
+      "learning_rate": 2.1925e-05,
+      "loss": 1.1504,
+      "step": 1123
+    },
+    {
+      "epoch": 1.679413549983658,
+      "grad_norm": 0.3011496067047119,
+      "learning_rate": 2.19e-05,
+      "loss": 1.3425,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6809076901526825,
+      "grad_norm": 0.2117116004228592,
+      "learning_rate": 2.1875e-05,
+      "loss": 1.428,
+      "step": 1125
+    },
+    {
+      "epoch": 1.682401830321707,
+      "grad_norm": 0.2260589450597763,
+      "learning_rate": 2.1850000000000003e-05,
+      "loss": 1.2989,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6838959704907317,
+      "grad_norm": 0.282379150390625,
+      "learning_rate": 2.1825000000000002e-05,
+      "loss": 1.5454,
+      "step": 1127
+    },
+    {
+      "epoch": 1.6853901106597564,
+      "grad_norm": 0.2984197735786438,
+      "learning_rate": 2.18e-05,
+      "loss": 1.4015,
+      "step": 1128
+    },
+    {
+      "epoch": 1.686884250828781,
+      "grad_norm": 0.23150713741779327,
+      "learning_rate": 2.1775e-05,
+      "loss": 1.1436,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6883783909978054,
+      "grad_norm": 0.21945536136627197,
+      "learning_rate": 2.175e-05,
+      "loss": 1.3747,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6898725311668301,
+      "grad_norm": 0.18954990804195404,
+      "learning_rate": 2.1725e-05,
+      "loss": 1.3046,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6913666713358548,
+      "grad_norm": 0.3952544927597046,
+      "learning_rate": 2.1700000000000002e-05,
+      "loss": 1.3724,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6928608115048793,
+      "grad_norm": 1.5406503677368164,
+      "learning_rate": 2.1675e-05,
+      "loss": 1.3255,
+      "step": 1133
+    },
+    {
+      "epoch": 1.6943549516739038,
+      "grad_norm": 0.20292794704437256,
+      "learning_rate": 2.165e-05,
+      "loss": 1.2522,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6958490918429285,
+      "grad_norm": 0.25634780526161194,
+      "learning_rate": 2.1625e-05,
+      "loss": 1.3941,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6973432320119533,
+      "grad_norm": 0.5052337050437927,
+      "learning_rate": 2.16e-05,
+      "loss": 1.49,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6988373721809777,
+      "grad_norm": 0.3816353678703308,
+      "learning_rate": 2.1575e-05,
+      "loss": 1.367,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7003315123500022,
+      "grad_norm": 0.19778917729854584,
+      "learning_rate": 2.1550000000000002e-05,
+      "loss": 1.3428,
+      "step": 1138
+    },
+    {
+      "epoch": 1.701825652519027,
+      "grad_norm": 0.214355006814003,
+      "learning_rate": 2.1525e-05,
+      "loss": 1.3231,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7033197926880517,
+      "grad_norm": 0.8011763691902161,
+      "learning_rate": 2.15e-05,
+      "loss": 1.2472,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7048139328570762,
+      "grad_norm": 0.21572141349315643,
+      "learning_rate": 2.1475e-05,
+      "loss": 1.3991,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7063080730261007,
+      "grad_norm": 0.17696185410022736,
+      "learning_rate": 2.145e-05,
+      "loss": 1.2861,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7078022131951254,
+      "grad_norm": 0.2655123472213745,
+      "learning_rate": 2.1425e-05,
+      "loss": 1.2419,
+      "step": 1143
+    },
+    {
+      "epoch": 1.70929635336415,
+      "grad_norm": 0.30516761541366577,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 1.2123,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7107904935331746,
+      "grad_norm": 0.33619871735572815,
+      "learning_rate": 2.1375e-05,
+      "loss": 1.3795,
+      "step": 1145
+    },
+    {
+      "epoch": 1.712284633702199,
+      "grad_norm": 0.3656558096408844,
+      "learning_rate": 2.135e-05,
+      "loss": 1.4199,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7137787738712238,
+      "grad_norm": 0.17710858583450317,
+      "learning_rate": 2.1325e-05,
+      "loss": 1.2199,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7152729140402485,
+      "grad_norm": 0.2536126375198364,
+      "learning_rate": 2.13e-05,
+      "loss": 1.3821,
+      "step": 1148
+    },
+    {
+      "epoch": 1.716767054209273,
+      "grad_norm": 0.5065903663635254,
+      "learning_rate": 2.1275000000000002e-05,
+      "loss": 1.2265,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7182611943782975,
+      "grad_norm": 0.21185562014579773,
+      "learning_rate": 2.125e-05,
+      "loss": 1.3668,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7197553345473222,
+      "grad_norm": 3.1472411155700684,
+      "learning_rate": 2.1225e-05,
+      "loss": 1.4943,
+      "step": 1151
+    },
+    {
+      "epoch": 1.721249474716347,
+      "grad_norm": 0.19197702407836914,
+      "learning_rate": 2.12e-05,
+      "loss": 1.328,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7227436148853714,
+      "grad_norm": 0.34128865599632263,
+      "learning_rate": 2.1175e-05,
+      "loss": 1.2109,
+      "step": 1153
+    },
+    {
+      "epoch": 1.724237755054396,
+      "grad_norm": 0.25855574011802673,
+      "learning_rate": 2.115e-05,
+      "loss": 1.5121,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7257318952234206,
+      "grad_norm": 0.32073017954826355,
+      "learning_rate": 2.1125000000000002e-05,
+      "loss": 1.2945,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7272260353924453,
+      "grad_norm": 0.449855774641037,
+      "learning_rate": 2.11e-05,
+      "loss": 1.4703,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7287201755614698,
+      "grad_norm": 0.6917439103126526,
+      "learning_rate": 2.1075e-05,
+      "loss": 1.4651,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7302143157304943,
+      "grad_norm": 0.5707057118415833,
+      "learning_rate": 2.105e-05,
+      "loss": 1.3591,
+      "step": 1158
+    },
+    {
+      "epoch": 1.731708455899519,
+      "grad_norm": 0.24529029428958893,
+      "learning_rate": 2.1025e-05,
+      "loss": 1.3453,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7332025960685438,
+      "grad_norm": 0.24148188531398773,
+      "learning_rate": 2.1e-05,
+      "loss": 1.4411,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7346967362375683,
+      "grad_norm": 0.28139758110046387,
+      "learning_rate": 2.0975e-05,
+      "loss": 1.263,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7361908764065928,
+      "grad_norm": 0.27365148067474365,
+      "learning_rate": 2.095e-05,
+      "loss": 1.2998,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7376850165756175,
+      "grad_norm": 0.2578750550746918,
+      "learning_rate": 2.0925e-05,
+      "loss": 1.0968,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7391791567446422,
+      "grad_norm": 0.27883380651474,
+      "learning_rate": 2.09e-05,
+      "loss": 1.2196,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7406732969136667,
+      "grad_norm": 0.22383853793144226,
+      "learning_rate": 2.0875e-05,
+      "loss": 1.3016,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7421674370826912,
+      "grad_norm": 0.2515240013599396,
+      "learning_rate": 2.085e-05,
+      "loss": 1.4733,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7436615772517159,
+      "grad_norm": 0.2814958393573761,
+      "learning_rate": 2.0825e-05,
+      "loss": 1.2962,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7451557174207406,
+      "grad_norm": 0.23939631879329681,
+      "learning_rate": 2.08e-05,
+      "loss": 1.4228,
+      "step": 1168
+    },
+    {
+      "epoch": 1.746649857589765,
+      "grad_norm": 0.25634765625,
+      "learning_rate": 2.0775e-05,
+      "loss": 1.2722,
+      "step": 1169
+    },
+    {
+      "epoch": 1.7481439977587896,
+      "grad_norm": 1.8628501892089844,
+      "learning_rate": 2.075e-05,
+      "loss": 1.2133,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7496381379278143,
+      "grad_norm": 0.2669479548931122,
+      "learning_rate": 2.0725e-05,
+      "loss": 1.2378,
+      "step": 1171
+    },
+    {
+      "epoch": 1.751132278096839,
+      "grad_norm": 0.23712307214736938,
+      "learning_rate": 2.07e-05,
+      "loss": 1.1109,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7526264182658635,
+      "grad_norm": 0.18234212696552277,
+      "learning_rate": 2.0675e-05,
+      "loss": 1.1441,
+      "step": 1173
+    },
+    {
+      "epoch": 1.754120558434888,
+      "grad_norm": 0.30427947640419006,
+      "learning_rate": 2.065e-05,
+      "loss": 1.3749,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7556146986039127,
+      "grad_norm": 0.21525146067142487,
+      "learning_rate": 2.0625e-05,
+      "loss": 1.3148,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7571088387729374,
+      "grad_norm": 0.36460453271865845,
+      "learning_rate": 2.06e-05,
+      "loss": 1.2794,
+      "step": 1176
+    },
+    {
+      "epoch": 1.758602978941962,
+      "grad_norm": 0.38430091738700867,
+      "learning_rate": 2.0575e-05,
+      "loss": 1.2026,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7600971191109867,
+      "grad_norm": 0.3811454474925995,
+      "learning_rate": 2.055e-05,
+      "loss": 1.0977,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7615912592800114,
+      "grad_norm": 0.2520235478878021,
+      "learning_rate": 2.0525e-05,
+      "loss": 1.3843,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7630853994490359,
+      "grad_norm": 0.30874738097190857,
+      "learning_rate": 2.05e-05,
+      "loss": 1.2541,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7645795396180604,
+      "grad_norm": 0.378074586391449,
+      "learning_rate": 2.0475e-05,
+      "loss": 1.1875,
+      "step": 1181
+    },
+    {
+      "epoch": 1.766073679787085,
+      "grad_norm": 0.39348742365837097,
+      "learning_rate": 2.045e-05,
+      "loss": 1.2389,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7675678199561098,
+      "grad_norm": 0.2662319540977478,
+      "learning_rate": 2.0425e-05,
+      "loss": 1.3791,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7690619601251343,
+      "grad_norm": 0.27314624190330505,
+      "learning_rate": 2.04e-05,
+      "loss": 1.5183,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7705561002941588,
+      "grad_norm": 0.2799110412597656,
+      "learning_rate": 2.0375e-05,
+      "loss": 1.3552,
+      "step": 1185
+    },
+    {
+      "epoch": 1.7720502404631835,
+      "grad_norm": 0.38711944222450256,
+      "learning_rate": 2.035e-05,
+      "loss": 1.3103,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7735443806322082,
+      "grad_norm": 0.22410157322883606,
+      "learning_rate": 2.0325e-05,
+      "loss": 1.3525,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7750385208012327,
+      "grad_norm": 0.21705371141433716,
+      "learning_rate": 2.0300000000000002e-05,
+      "loss": 1.3232,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7765326609702572,
+      "grad_norm": 0.18889586627483368,
+      "learning_rate": 2.0275e-05,
+      "loss": 1.0816,
+      "step": 1189
+    },
+    {
+      "epoch": 1.778026801139282,
+      "grad_norm": 0.32244136929512024,
+      "learning_rate": 2.025e-05,
+      "loss": 1.3168,
+      "step": 1190
+    },
+    {
+      "epoch": 1.7795209413083066,
+      "grad_norm": 0.17929215729236603,
+      "learning_rate": 2.0225000000000004e-05,
+      "loss": 1.286,
+      "step": 1191
+    },
+    {
+      "epoch": 1.7810150814773311,
+      "grad_norm": 0.21204537153244019,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 1.1635,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7825092216463556,
+      "grad_norm": 0.22701428830623627,
+      "learning_rate": 2.0175000000000003e-05,
+      "loss": 1.4269,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7840033618153803,
+      "grad_norm": 0.37344813346862793,
+      "learning_rate": 2.0150000000000002e-05,
+      "loss": 1.4175,
+      "step": 1194
+    },
+    {
+      "epoch": 1.785497501984405,
+      "grad_norm": 0.2566252052783966,
+      "learning_rate": 2.0125e-05,
+      "loss": 1.332,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7869916421534295,
+      "grad_norm": 0.3862023949623108,
+      "learning_rate": 2.01e-05,
+      "loss": 1.5337,
+      "step": 1196
+    },
+    {
+      "epoch": 1.788485782322454,
+      "grad_norm": 0.20119082927703857,
+      "learning_rate": 2.0075000000000003e-05,
+      "loss": 1.2904,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7899799224914787,
+      "grad_norm": 0.3420519530773163,
+      "learning_rate": 2.0050000000000003e-05,
+      "loss": 1.1776,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7914740626605035,
+      "grad_norm": 0.22510448098182678,
+      "learning_rate": 2.0025000000000002e-05,
+      "loss": 1.2816,
+      "step": 1199
+    },
+    {
+      "epoch": 1.792968202829528,
+      "grad_norm": 0.5224809050559998,
+      "learning_rate": 2e-05,
+      "loss": 1.357,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7944623429985525,
+      "grad_norm": 0.23496760427951813,
+      "learning_rate": 1.9975e-05,
+      "loss": 1.3329,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7959564831675772,
+      "grad_norm": 0.2106732875108719,
+      "learning_rate": 1.995e-05,
+      "loss": 1.4448,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7974506233366019,
+      "grad_norm": 0.22301816940307617,
+      "learning_rate": 1.9925000000000003e-05,
+      "loss": 1.4227,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7989447635056264,
+      "grad_norm": 0.20743943750858307,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 1.4388,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8004389036746509,
+      "grad_norm": 0.24213577806949615,
+      "learning_rate": 1.9875000000000002e-05,
+      "loss": 1.3773,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8019330438436756,
+      "grad_norm": 0.2809053063392639,
+      "learning_rate": 1.985e-05,
+      "loss": 1.3239,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8034271840127003,
+      "grad_norm": 0.17947784066200256,
+      "learning_rate": 1.9825e-05,
+      "loss": 1.3838,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8049213241817248,
+      "grad_norm": 0.249902606010437,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 1.3094,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8064154643507493,
+      "grad_norm": 0.3139473795890808,
+      "learning_rate": 1.9775000000000003e-05,
+      "loss": 1.292,
+      "step": 1209
+    },
+    {
+      "epoch": 1.807909604519774,
+      "grad_norm": 0.3059343695640564,
+      "learning_rate": 1.9750000000000002e-05,
+      "loss": 1.4008,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8094037446887987,
+      "grad_norm": 0.22615757584571838,
+      "learning_rate": 1.9725000000000002e-05,
+      "loss": 1.4586,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8108978848578232,
+      "grad_norm": 0.395865261554718,
+      "learning_rate": 1.97e-05,
+      "loss": 1.2452,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8123920250268477,
+      "grad_norm": 0.29632511734962463,
+      "learning_rate": 1.9675e-05,
+      "loss": 1.3564,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8138861651958724,
+      "grad_norm": 0.3064177632331848,
+      "learning_rate": 1.9650000000000003e-05,
+      "loss": 1.2446,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8153803053648971,
+      "grad_norm": 0.30816328525543213,
+      "learning_rate": 1.9625000000000003e-05,
+      "loss": 1.3273,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8168744455339216,
+      "grad_norm": 0.18040741980075836,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.2015,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8183685857029461,
+      "grad_norm": 0.24273858964443207,
+      "learning_rate": 1.9575e-05,
+      "loss": 1.2347,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8198627258719708,
+      "grad_norm": 0.2441396564245224,
+      "learning_rate": 1.955e-05,
+      "loss": 1.3922,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8213568660409956,
+      "grad_norm": 0.2110196053981781,
+      "learning_rate": 1.9525e-05,
+      "loss": 1.2187,
+      "step": 1219
+    },
+    {
+      "epoch": 1.82285100621002,
+      "grad_norm": 0.2482791244983673,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 1.4491,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8243451463790445,
+      "grad_norm": 1.2229297161102295,
+      "learning_rate": 1.9475000000000002e-05,
+      "loss": 1.3337,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8258392865480693,
+      "grad_norm": 0.2569945156574249,
+      "learning_rate": 1.9450000000000002e-05,
+      "loss": 1.4649,
+      "step": 1222
+    },
+    {
+      "epoch": 1.827333426717094,
+      "grad_norm": 0.23530051112174988,
+      "learning_rate": 1.9425e-05,
+      "loss": 1.3162,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8288275668861185,
+      "grad_norm": 0.23422133922576904,
+      "learning_rate": 1.94e-05,
+      "loss": 1.2149,
+      "step": 1224
+    },
+    {
+      "epoch": 1.830321707055143,
+      "grad_norm": 0.8085640072822571,
+      "learning_rate": 1.9375e-05,
+      "loss": 1.3529,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8318158472241677,
+      "grad_norm": 0.23965558409690857,
+      "learning_rate": 1.9350000000000003e-05,
+      "loss": 1.2814,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8333099873931924,
+      "grad_norm": 0.33934149146080017,
+      "learning_rate": 1.9325000000000002e-05,
+      "loss": 1.3801,
+      "step": 1227
+    },
+    {
+      "epoch": 1.834804127562217,
+      "grad_norm": 0.20922242105007172,
+      "learning_rate": 1.93e-05,
+      "loss": 1.2043,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8362982677312414,
+      "grad_norm": 0.23952169716358185,
+      "learning_rate": 1.9275e-05,
+      "loss": 1.2686,
+      "step": 1229
+    },
+    {
+      "epoch": 1.837792407900266,
+      "grad_norm": 0.35494279861450195,
+      "learning_rate": 1.925e-05,
+      "loss": 1.42,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8392865480692908,
+      "grad_norm": 0.2617303729057312,
+      "learning_rate": 1.9225e-05,
+      "loss": 1.3625,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8407806882383153,
+      "grad_norm": 0.26504993438720703,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.2937,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8422748284073398,
+      "grad_norm": 0.3170221149921417,
+      "learning_rate": 1.9175000000000002e-05,
+      "loss": 1.5529,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8437689685763645,
+      "grad_norm": 0.24354490637779236,
+      "learning_rate": 1.915e-05,
+      "loss": 1.1315,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8452631087453892,
+      "grad_norm": 0.2081626057624817,
+      "learning_rate": 1.9125e-05,
+      "loss": 1.4484,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8467572489144137,
+      "grad_norm": 0.1551181972026825,
+      "learning_rate": 1.91e-05,
+      "loss": 1.3253,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8482513890834384,
+      "grad_norm": 0.17210716009140015,
+      "learning_rate": 1.9075000000000003e-05,
+      "loss": 1.3218,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8497455292524632,
+      "grad_norm": 0.2032608836889267,
+      "learning_rate": 1.9050000000000002e-05,
+      "loss": 1.2805,
+      "step": 1238
+    },
+    {
+      "epoch": 1.8512396694214877,
+      "grad_norm": 0.45856183767318726,
+      "learning_rate": 1.9025e-05,
+      "loss": 1.1087,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8527338095905121,
+      "grad_norm": 0.18352068960666656,
+      "learning_rate": 1.9e-05,
+      "loss": 1.3764,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8542279497595369,
+      "grad_norm": 0.2714996039867401,
+      "learning_rate": 1.8975e-05,
+      "loss": 1.5606,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8557220899285616,
+      "grad_norm": 0.19584618508815765,
+      "learning_rate": 1.895e-05,
+      "loss": 1.3594,
+      "step": 1242
+    },
+    {
+      "epoch": 1.857216230097586,
+      "grad_norm": 0.2596234083175659,
+      "learning_rate": 1.8925000000000003e-05,
+      "loss": 1.3978,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8587103702666106,
+      "grad_norm": 0.19500860571861267,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 1.2443,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8602045104356353,
+      "grad_norm": 0.6344171166419983,
+      "learning_rate": 1.8875e-05,
+      "loss": 1.591,
+      "step": 1245
+    },
+    {
+      "epoch": 1.86169865060466,
+      "grad_norm": 0.24306315183639526,
+      "learning_rate": 1.885e-05,
+      "loss": 1.3199,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8631927907736845,
+      "grad_norm": 0.20791853964328766,
+      "learning_rate": 1.8825e-05,
+      "loss": 1.2792,
+      "step": 1247
+    },
+    {
+      "epoch": 1.864686930942709,
+      "grad_norm": 0.3387373685836792,
+      "learning_rate": 1.88e-05,
+      "loss": 1.3661,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8661810711117337,
+      "grad_norm": 0.17687146365642548,
+      "learning_rate": 1.8775000000000002e-05,
+      "loss": 1.2467,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8676752112807584,
+      "grad_norm": 0.2075313925743103,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.3595,
+      "step": 1250
+    },
+    {
+      "epoch": 1.869169351449783,
+      "grad_norm": 0.4056931734085083,
+      "learning_rate": 1.8725e-05,
+      "loss": 1.293,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8706634916188074,
+      "grad_norm": 0.21190667152404785,
+      "learning_rate": 1.87e-05,
+      "loss": 1.2857,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8721576317878321,
+      "grad_norm": 0.8928343653678894,
+      "learning_rate": 1.8675e-05,
+      "loss": 1.3515,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8736517719568568,
+      "grad_norm": 0.20301412045955658,
+      "learning_rate": 1.865e-05,
+      "loss": 1.3271,
+      "step": 1254
+    },
+    {
+      "epoch": 1.8751459121258813,
+      "grad_norm": 0.21219702064990997,
+      "learning_rate": 1.8625000000000002e-05,
+      "loss": 1.3911,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8766400522949058,
+      "grad_norm": 0.20927630364894867,
+      "learning_rate": 1.86e-05,
+      "loss": 1.1812,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8781341924639305,
+      "grad_norm": 0.19185447692871094,
+      "learning_rate": 1.8575e-05,
+      "loss": 1.3061,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8796283326329553,
+      "grad_norm": 0.3142222464084625,
+      "learning_rate": 1.855e-05,
+      "loss": 1.4002,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8811224728019798,
+      "grad_norm": 0.2145937979221344,
+      "learning_rate": 1.8525e-05,
+      "loss": 1.192,
+      "step": 1259
+    },
+    {
+      "epoch": 1.8826166129710042,
+      "grad_norm": 0.15852953493595123,
+      "learning_rate": 1.85e-05,
+      "loss": 1.097,
+      "step": 1260
+    },
+    {
+      "epoch": 1.884110753140029,
+      "grad_norm": 0.3558714985847473,
+      "learning_rate": 1.8475000000000002e-05,
+      "loss": 1.1759,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8856048933090537,
+      "grad_norm": 0.2350102961063385,
+      "learning_rate": 1.845e-05,
+      "loss": 1.2792,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8870990334780782,
+      "grad_norm": 0.23744919896125793,
+      "learning_rate": 1.8425e-05,
+      "loss": 1.3294,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8885931736471027,
+      "grad_norm": 0.4808562994003296,
+      "learning_rate": 1.84e-05,
+      "loss": 1.3644,
+      "step": 1264
+    },
+    {
+      "epoch": 1.8900873138161274,
+      "grad_norm": 0.18481269478797913,
+      "learning_rate": 1.8375e-05,
+      "loss": 1.4389,
+      "step": 1265
+    },
+    {
+      "epoch": 1.891581453985152,
+      "grad_norm": 0.1974407136440277,
+      "learning_rate": 1.8350000000000002e-05,
+      "loss": 1.3803,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8930755941541766,
+      "grad_norm": 0.18426170945167542,
+      "learning_rate": 1.8325e-05,
+      "loss": 1.4062,
+      "step": 1267
+    },
+    {
+      "epoch": 1.894569734323201,
+      "grad_norm": 0.15900960564613342,
+      "learning_rate": 1.83e-05,
+      "loss": 1.2374,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8960638744922258,
+      "grad_norm": 0.21791104972362518,
+      "learning_rate": 1.8275e-05,
+      "loss": 1.3535,
+      "step": 1269
+    },
+    {
+      "epoch": 1.8975580146612505,
+      "grad_norm": 0.2844353914260864,
+      "learning_rate": 1.825e-05,
+      "loss": 1.4481,
+      "step": 1270
+    },
+    {
+      "epoch": 1.899052154830275,
+      "grad_norm": 0.2063741832971573,
+      "learning_rate": 1.8225e-05,
+      "loss": 1.2813,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9005462949992995,
+      "grad_norm": 0.28945934772491455,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 1.3534,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9020404351683242,
+      "grad_norm": 0.3286309540271759,
+      "learning_rate": 1.8175e-05,
+      "loss": 1.335,
+      "step": 1273
+    },
+    {
+      "epoch": 1.903534575337349,
+      "grad_norm": 0.33386310935020447,
+      "learning_rate": 1.815e-05,
+      "loss": 1.1667,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9050287155063734,
+      "grad_norm": 0.4026789665222168,
+      "learning_rate": 1.8125e-05,
+      "loss": 1.296,
+      "step": 1275
+    },
+    {
+      "epoch": 1.906522855675398,
+      "grad_norm": 0.1674850583076477,
+      "learning_rate": 1.81e-05,
+      "loss": 1.2641,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9080169958444226,
+      "grad_norm": 0.6385492086410522,
+      "learning_rate": 1.8075e-05,
+      "loss": 1.2057,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9095111360134474,
+      "grad_norm": 0.7805976271629333,
+      "learning_rate": 1.805e-05,
+      "loss": 1.3314,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9110052761824718,
+      "grad_norm": 0.17184190452098846,
+      "learning_rate": 1.8025e-05,
+      "loss": 1.3208,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9124994163514963,
+      "grad_norm": 0.21604672074317932,
+      "learning_rate": 1.8e-05,
+      "loss": 1.3815,
+      "step": 1280
+    },
+    {
+      "epoch": 1.913993556520521,
+      "grad_norm": 0.3038337826728821,
+      "learning_rate": 1.7975e-05,
+      "loss": 1.3725,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9154876966895458,
+      "grad_norm": 0.2851409614086151,
+      "learning_rate": 1.795e-05,
+      "loss": 1.3311,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9169818368585703,
+      "grad_norm": 0.22219157218933105,
+      "learning_rate": 1.7925e-05,
+      "loss": 1.3016,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9184759770275948,
+      "grad_norm": 0.29200297594070435,
+      "learning_rate": 1.79e-05,
+      "loss": 1.2594,
+      "step": 1284
+    },
+    {
+      "epoch": 1.9199701171966195,
+      "grad_norm": 0.4923015832901001,
+      "learning_rate": 1.7875e-05,
+      "loss": 1.2475,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9214642573656442,
+      "grad_norm": 0.22583426535129547,
+      "learning_rate": 1.785e-05,
+      "loss": 1.1543,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9229583975346687,
+      "grad_norm": 0.21450522541999817,
+      "learning_rate": 1.7825e-05,
+      "loss": 1.3766,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9244525377036932,
+      "grad_norm": 0.22731810808181763,
+      "learning_rate": 1.78e-05,
+      "loss": 1.3878,
+      "step": 1288
+    },
+    {
+      "epoch": 1.925946677872718,
+      "grad_norm": 0.17790710926055908,
+      "learning_rate": 1.7775e-05,
+      "loss": 1.2292,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9274408180417426,
+      "grad_norm": 2.0318102836608887,
+      "learning_rate": 1.775e-05,
+      "loss": 1.3844,
+      "step": 1290
+    },
+    {
+      "epoch": 1.928934958210767,
+      "grad_norm": 0.21264667809009552,
+      "learning_rate": 1.7725e-05,
+      "loss": 1.3728,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9304290983797916,
+      "grad_norm": 0.23059220612049103,
+      "learning_rate": 1.77e-05,
+      "loss": 1.3793,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9319232385488163,
+      "grad_norm": 0.2054581195116043,
+      "learning_rate": 1.7675e-05,
+      "loss": 1.2632,
+      "step": 1293
+    },
+    {
+      "epoch": 1.933417378717841,
+      "grad_norm": 0.9071913361549377,
+      "learning_rate": 1.765e-05,
+      "loss": 1.4137,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9349115188868655,
+      "grad_norm": 0.2947272062301636,
+      "learning_rate": 1.7625e-05,
+      "loss": 1.3661,
+      "step": 1295
+    },
+    {
+      "epoch": 1.93640565905589,
+      "grad_norm": 0.2021481841802597,
+      "learning_rate": 1.76e-05,
+      "loss": 1.2498,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9378997992249147,
+      "grad_norm": 0.22752541303634644,
+      "learning_rate": 1.7575e-05,
+      "loss": 1.3359,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9393939393939394,
+      "grad_norm": 0.21617968380451202,
+      "learning_rate": 1.755e-05,
+      "loss": 1.1765,
+      "step": 1298
+    },
+    {
+      "epoch": 1.940888079562964,
+      "grad_norm": 0.23794393241405487,
+      "learning_rate": 1.7525e-05,
+      "loss": 1.2394,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9423822197319887,
+      "grad_norm": 0.22893854975700378,
+      "learning_rate": 1.75e-05,
+      "loss": 1.3831,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9438763599010134,
+      "grad_norm": 0.2074202597141266,
+      "learning_rate": 1.7475e-05,
+      "loss": 1.3276,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9453705000700379,
+      "grad_norm": 0.5460404753684998,
+      "learning_rate": 1.745e-05,
+      "loss": 1.2246,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9468646402390624,
+      "grad_norm": 0.551843523979187,
+      "learning_rate": 1.7425e-05,
+      "loss": 1.2742,
+      "step": 1303
+    },
+    {
+      "epoch": 1.948358780408087,
+      "grad_norm": 0.2572923004627228,
+      "learning_rate": 1.74e-05,
+      "loss": 1.3383,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9498529205771118,
+      "grad_norm": 0.20585736632347107,
+      "learning_rate": 1.7375e-05,
+      "loss": 1.4466,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9513470607461363,
+      "grad_norm": 0.2732091546058655,
+      "learning_rate": 1.7349999999999998e-05,
+      "loss": 1.0664,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9528412009151608,
+      "grad_norm": 0.2675059735774994,
+      "learning_rate": 1.7325e-05,
+      "loss": 1.2259,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9543353410841855,
+      "grad_norm": 0.2213507890701294,
+      "learning_rate": 1.73e-05,
+      "loss": 1.3838,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9558294812532102,
+      "grad_norm": 0.18887294828891754,
+      "learning_rate": 1.7275e-05,
+      "loss": 1.226,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9573236214222347,
+      "grad_norm": 0.21098558604717255,
+      "learning_rate": 1.725e-05,
+      "loss": 1.4265,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9588177615912592,
+      "grad_norm": 0.2039249986410141,
+      "learning_rate": 1.7225e-05,
+      "loss": 1.2338,
+      "step": 1311
+    },
+    {
+      "epoch": 1.960311901760284,
+      "grad_norm": 0.2207634598016739,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 1.2166,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9618060419293086,
+      "grad_norm": 0.2204556167125702,
+      "learning_rate": 1.7175e-05,
+      "loss": 1.2149,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9633001820983331,
+      "grad_norm": 0.19095070660114288,
+      "learning_rate": 1.7150000000000004e-05,
+      "loss": 1.3607,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9647943222673576,
+      "grad_norm": 0.1517525315284729,
+      "learning_rate": 1.7125000000000003e-05,
+      "loss": 1.172,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9662884624363823,
+      "grad_norm": 0.21833693981170654,
+      "learning_rate": 1.7100000000000002e-05,
+      "loss": 1.4572,
+      "step": 1316
+    },
+    {
+      "epoch": 1.967782602605407,
+      "grad_norm": 0.2464878410100937,
+      "learning_rate": 1.7075e-05,
+      "loss": 1.4209,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9692767427744315,
+      "grad_norm": 0.16277910768985748,
+      "learning_rate": 1.705e-05,
+      "loss": 1.4236,
+      "step": 1318
+    },
+    {
+      "epoch": 1.970770882943456,
+      "grad_norm": 0.23348970711231232,
+      "learning_rate": 1.7025e-05,
+      "loss": 1.0046,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9722650231124808,
+      "grad_norm": 0.3407599627971649,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.3508,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9737591632815055,
+      "grad_norm": 0.8567706346511841,
+      "learning_rate": 1.6975000000000003e-05,
+      "loss": 1.453,
+      "step": 1321
+    },
+    {
+      "epoch": 1.97525330345053,
+      "grad_norm": 0.18511173129081726,
+      "learning_rate": 1.6950000000000002e-05,
+      "loss": 1.3107,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9767474436195545,
+      "grad_norm": 0.21256686747074127,
+      "learning_rate": 1.6925e-05,
+      "loss": 1.3865,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9782415837885792,
+      "grad_norm": 0.25371041893959045,
+      "learning_rate": 1.69e-05,
+      "loss": 1.3664,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9797357239576039,
+      "grad_norm": 0.15344837307929993,
+      "learning_rate": 1.6875000000000004e-05,
+      "loss": 1.2623,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9812298641266284,
+      "grad_norm": 0.27859944105148315,
+      "learning_rate": 1.6850000000000003e-05,
+      "loss": 1.3419,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9827240042956529,
+      "grad_norm": 0.30782634019851685,
+      "learning_rate": 1.6825000000000002e-05,
+      "loss": 1.1088,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9842181444646776,
+      "grad_norm": 0.26927539706230164,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.3839,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9857122846337023,
+      "grad_norm": 0.5329779982566833,
+      "learning_rate": 1.6775e-05,
+      "loss": 1.3443,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9872064248027268,
+      "grad_norm": 0.2376929074525833,
+      "learning_rate": 1.675e-05,
+      "loss": 1.2627,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9887005649717513,
+      "grad_norm": 0.21043722331523895,
+      "learning_rate": 1.6725000000000003e-05,
+      "loss": 1.4348,
+      "step": 1331
+    },
+    {
+      "epoch": 1.990194705140776,
+      "grad_norm": 0.2253684550523758,
+      "learning_rate": 1.6700000000000003e-05,
+      "loss": 1.1814,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9916888453098007,
+      "grad_norm": 0.23785364627838135,
+      "learning_rate": 1.6675000000000002e-05,
+      "loss": 1.0943,
+      "step": 1333
+    },
+    {
+      "epoch": 1.9931829854788252,
+      "grad_norm": 0.29076915979385376,
+      "learning_rate": 1.665e-05,
+      "loss": 1.381,
+      "step": 1334
+    },
+    {
+      "epoch": 1.9946771256478497,
+      "grad_norm": 0.21580135822296143,
+      "learning_rate": 1.6625e-05,
+      "loss": 1.2626,
+      "step": 1335
+    },
+    {
+      "epoch": 1.9961712658168744,
+      "grad_norm": 0.30649375915527344,
+      "learning_rate": 1.66e-05,
+      "loss": 1.5823,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9976654059858991,
+      "grad_norm": 0.1969205141067505,
+      "learning_rate": 1.6575000000000003e-05,
+      "loss": 1.1942,
+      "step": 1337
+    },
+    {
+      "epoch": 1.9991595461549236,
+      "grad_norm": 0.3132281005382538,
+      "learning_rate": 1.6550000000000002e-05,
+      "loss": 1.4082,
+      "step": 1338
+    },
+    {
+      "epoch": 2.000653686323948,
+      "grad_norm": 0.1845480054616928,
+      "learning_rate": 1.6525000000000002e-05,
+      "loss": 1.3161,
+      "step": 1339
+    },
+    {
+      "epoch": 2.002147826492973,
+      "grad_norm": 0.3126375079154968,
+      "learning_rate": 1.65e-05,
+      "loss": 1.3993,
+      "step": 1340
+    },
+    {
+      "epoch": 2.0036419666619976,
+      "grad_norm": 0.42225363850593567,
+      "learning_rate": 1.6475e-05,
+      "loss": 1.2508,
+      "step": 1341
+    },
+    {
+      "epoch": 2.005136106831022,
+      "grad_norm": 0.16167034208774567,
+      "learning_rate": 1.645e-05,
+      "loss": 1.1553,
+      "step": 1342
+    },
+    {
+      "epoch": 2.0066302470000466,
+      "grad_norm": 0.19103498756885529,
+      "learning_rate": 1.6425000000000003e-05,
+      "loss": 1.2684,
+      "step": 1343
+    },
+    {
+      "epoch": 2.0081243871690715,
+      "grad_norm": 0.23756802082061768,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 1.2853,
+      "step": 1344
+    },
+    {
+      "epoch": 2.009618527338096,
+      "grad_norm": 0.23275527358055115,
+      "learning_rate": 1.6375e-05,
+      "loss": 1.254,
+      "step": 1345
+    },
+    {
+      "epoch": 2.0111126675071205,
+      "grad_norm": 0.17952296137809753,
+      "learning_rate": 1.635e-05,
+      "loss": 1.2585,
+      "step": 1346
+    },
+    {
+      "epoch": 2.012606807676145,
+      "grad_norm": 0.9764080047607422,
+      "learning_rate": 1.6325e-05,
+      "loss": 1.2869,
+      "step": 1347
+    },
+    {
+      "epoch": 2.01410094784517,
+      "grad_norm": 0.22916385531425476,
+      "learning_rate": 1.63e-05,
+      "loss": 1.3214,
+      "step": 1348
+    },
+    {
+      "epoch": 2.0155950880141944,
+      "grad_norm": 0.18610818684101105,
+      "learning_rate": 1.6275000000000003e-05,
+      "loss": 1.2997,
+      "step": 1349
+    },
+    {
+      "epoch": 2.017089228183219,
+      "grad_norm": 0.19289493560791016,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 1.2952,
+      "step": 1350
+    },
+    {
+      "epoch": 2.0185833683522434,
+      "grad_norm": 0.19546003639698029,
+      "learning_rate": 1.6225e-05,
+      "loss": 1.3537,
+      "step": 1351
+    },
+    {
+      "epoch": 2.0200775085212683,
+      "grad_norm": 0.23147369921207428,
+      "learning_rate": 1.62e-05,
+      "loss": 1.4243,
+      "step": 1352
+    },
+    {
+      "epoch": 2.021571648690293,
+      "grad_norm": 0.18282002210617065,
+      "learning_rate": 1.6175e-05,
+      "loss": 1.0762,
+      "step": 1353
+    },
+    {
+      "epoch": 2.0230657888593173,
+      "grad_norm": 0.18470488488674164,
+      "learning_rate": 1.6150000000000003e-05,
+      "loss": 1.4082,
+      "step": 1354
+    },
+    {
+      "epoch": 2.024559929028342,
+      "grad_norm": 0.2198750376701355,
+      "learning_rate": 1.6125000000000002e-05,
+      "loss": 1.4813,
+      "step": 1355
+    },
+    {
+      "epoch": 2.0260540691973667,
+      "grad_norm": 0.2406291961669922,
+      "learning_rate": 1.6100000000000002e-05,
+      "loss": 1.4522,
+      "step": 1356
+    },
+    {
+      "epoch": 2.0275482093663912,
+      "grad_norm": 0.3409310281276703,
+      "learning_rate": 1.6075e-05,
+      "loss": 1.501,
+      "step": 1357
+    },
+    {
+      "epoch": 2.0290423495354157,
+      "grad_norm": 0.2944510281085968,
+      "learning_rate": 1.605e-05,
+      "loss": 1.2892,
+      "step": 1358
+    },
+    {
+      "epoch": 2.0305364897044402,
+      "grad_norm": 0.2915669083595276,
+      "learning_rate": 1.6025e-05,
+      "loss": 1.2348,
+      "step": 1359
+    },
+    {
+      "epoch": 2.032030629873465,
+      "grad_norm": 0.5818542242050171,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.2945,
+      "step": 1360
+    },
+    {
+      "epoch": 2.0335247700424897,
+      "grad_norm": 0.6113788485527039,
+      "learning_rate": 1.5975000000000002e-05,
+      "loss": 1.4605,
+      "step": 1361
+    },
+    {
+      "epoch": 2.035018910211514,
+      "grad_norm": 0.20517785847187042,
+      "learning_rate": 1.595e-05,
+      "loss": 1.3224,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0365130503805386,
+      "grad_norm": 0.16138556599617004,
+      "learning_rate": 1.5925e-05,
+      "loss": 1.1755,
+      "step": 1363
+    },
+    {
+      "epoch": 2.0380071905495636,
+      "grad_norm": 0.25767239928245544,
+      "learning_rate": 1.59e-05,
+      "loss": 1.1872,
+      "step": 1364
+    },
+    {
+      "epoch": 2.039501330718588,
+      "grad_norm": 0.2035055011510849,
+      "learning_rate": 1.5875e-05,
+      "loss": 1.5063,
+      "step": 1365
+    },
+    {
+      "epoch": 2.0409954708876126,
+      "grad_norm": 0.25032323598861694,
+      "learning_rate": 1.5850000000000002e-05,
+      "loss": 1.2495,
+      "step": 1366
+    },
+    {
+      "epoch": 2.042489611056637,
+      "grad_norm": 0.2417837679386139,
+      "learning_rate": 1.5825000000000002e-05,
+      "loss": 1.1995,
+      "step": 1367
+    },
+    {
+      "epoch": 2.043983751225662,
+      "grad_norm": 0.21311062574386597,
+      "learning_rate": 1.58e-05,
+      "loss": 1.2697,
+      "step": 1368
+    },
+    {
+      "epoch": 2.0454778913946865,
+      "grad_norm": 0.19063788652420044,
+      "learning_rate": 1.5775e-05,
+      "loss": 1.2617,
+      "step": 1369
+    },
+    {
+      "epoch": 2.046972031563711,
+      "grad_norm": 0.23214475810527802,
+      "learning_rate": 1.575e-05,
+      "loss": 1.2846,
+      "step": 1370
+    },
+    {
+      "epoch": 2.0484661717327355,
+      "grad_norm": 0.22164221107959747,
+      "learning_rate": 1.5725e-05,
+      "loss": 1.1985,
+      "step": 1371
+    },
+    {
+      "epoch": 2.0499603119017604,
+      "grad_norm": 0.42086952924728394,
+      "learning_rate": 1.5700000000000002e-05,
+      "loss": 1.2233,
+      "step": 1372
+    },
+    {
+      "epoch": 2.051454452070785,
+      "grad_norm": 0.19688980281352997,
+      "learning_rate": 1.5675e-05,
+      "loss": 1.4493,
+      "step": 1373
+    },
+    {
+      "epoch": 2.0529485922398094,
+      "grad_norm": 0.19368377327919006,
+      "learning_rate": 1.565e-05,
+      "loss": 1.3613,
+      "step": 1374
+    },
+    {
+      "epoch": 2.054442732408834,
+      "grad_norm": 0.17706935107707977,
+      "learning_rate": 1.5625e-05,
+      "loss": 1.3222,
+      "step": 1375
+    },
+    {
+      "epoch": 2.055936872577859,
+      "grad_norm": 0.37456485629081726,
+      "learning_rate": 1.56e-05,
+      "loss": 1.211,
+      "step": 1376
+    },
+    {
+      "epoch": 2.0574310127468833,
+      "grad_norm": 0.41341719031333923,
+      "learning_rate": 1.5575e-05,
+      "loss": 1.3862,
+      "step": 1377
+    },
+    {
+      "epoch": 2.058925152915908,
+      "grad_norm": 0.21140192449092865,
+      "learning_rate": 1.5550000000000002e-05,
+      "loss": 1.2258,
+      "step": 1378
+    },
+    {
+      "epoch": 2.0604192930849323,
+      "grad_norm": 0.22072890400886536,
+      "learning_rate": 1.5525e-05,
+      "loss": 1.1975,
+      "step": 1379
+    },
+    {
+      "epoch": 2.0619134332539573,
+      "grad_norm": 0.2429424673318863,
+      "learning_rate": 1.55e-05,
+      "loss": 1.1305,
+      "step": 1380
+    },
+    {
+      "epoch": 2.0634075734229818,
+      "grad_norm": 0.17470882833003998,
+      "learning_rate": 1.5475e-05,
+      "loss": 1.2887,
+      "step": 1381
+    },
+    {
+      "epoch": 2.0649017135920062,
+      "grad_norm": 0.2611446678638458,
+      "learning_rate": 1.545e-05,
+      "loss": 1.3548,
+      "step": 1382
+    },
+    {
+      "epoch": 2.0663958537610307,
+      "grad_norm": 0.18730558454990387,
+      "learning_rate": 1.5425000000000002e-05,
+      "loss": 1.3207,
+      "step": 1383
+    },
+    {
+      "epoch": 2.0678899939300557,
+      "grad_norm": 0.22779415547847748,
+      "learning_rate": 1.54e-05,
+      "loss": 1.3039,
+      "step": 1384
+    },
+    {
+      "epoch": 2.06938413409908,
+      "grad_norm": 0.22781801223754883,
+      "learning_rate": 1.5375e-05,
+      "loss": 1.2513,
+      "step": 1385
+    },
+    {
+      "epoch": 2.0708782742681047,
+      "grad_norm": 0.6857816576957703,
+      "learning_rate": 1.535e-05,
+      "loss": 1.3592,
+      "step": 1386
+    },
+    {
+      "epoch": 2.0723724144371296,
+      "grad_norm": 0.22949860990047455,
+      "learning_rate": 1.5325e-05,
+      "loss": 1.3609,
+      "step": 1387
+    },
+    {
+      "epoch": 2.073866554606154,
+      "grad_norm": 0.29369252920150757,
+      "learning_rate": 1.53e-05,
+      "loss": 1.3531,
+      "step": 1388
+    },
+    {
+      "epoch": 2.0753606947751786,
+      "grad_norm": 0.17681251466274261,
+      "learning_rate": 1.5275000000000002e-05,
+      "loss": 1.3514,
+      "step": 1389
+    },
+    {
+      "epoch": 2.076854834944203,
+      "grad_norm": 0.19840802252292633,
+      "learning_rate": 1.525e-05,
+      "loss": 1.3562,
+      "step": 1390
+    },
+    {
+      "epoch": 2.0783489751132276,
+      "grad_norm": 0.38067206740379333,
+      "learning_rate": 1.5225e-05,
+      "loss": 1.3926,
+      "step": 1391
+    },
+    {
+      "epoch": 2.0798431152822525,
+      "grad_norm": 0.23130175471305847,
+      "learning_rate": 1.52e-05,
+      "loss": 1.2658,
+      "step": 1392
+    },
+    {
+      "epoch": 2.081337255451277,
+      "grad_norm": 0.24068257212638855,
+      "learning_rate": 1.5175e-05,
+      "loss": 1.2055,
+      "step": 1393
+    },
+    {
+      "epoch": 2.0828313956203015,
+      "grad_norm": 0.2135651409626007,
+      "learning_rate": 1.515e-05,
+      "loss": 1.3947,
+      "step": 1394
+    },
+    {
+      "epoch": 2.0843255357893264,
+      "grad_norm": 0.25119951367378235,
+      "learning_rate": 1.5125e-05,
+      "loss": 1.4246,
+      "step": 1395
+    },
+    {
+      "epoch": 2.085819675958351,
+      "grad_norm": 0.2441864162683487,
+      "learning_rate": 1.51e-05,
+      "loss": 1.4057,
+      "step": 1396
+    },
+    {
+      "epoch": 2.0873138161273754,
+      "grad_norm": 0.3345242440700531,
+      "learning_rate": 1.5075e-05,
+      "loss": 1.4348,
+      "step": 1397
+    },
+    {
+      "epoch": 2.0888079562964,
+      "grad_norm": 0.2515486180782318,
+      "learning_rate": 1.505e-05,
+      "loss": 1.2924,
+      "step": 1398
+    },
+    {
+      "epoch": 2.090302096465425,
+      "grad_norm": 0.1982625424861908,
+      "learning_rate": 1.5025000000000001e-05,
+      "loss": 1.3543,
+      "step": 1399
+    },
+    {
+      "epoch": 2.0917962366344494,
+      "grad_norm": 0.2248583287000656,
+      "learning_rate": 1.5e-05,
+      "loss": 1.3344,
+      "step": 1400
+    },
+    {
+      "epoch": 2.093290376803474,
+      "grad_norm": 0.22528663277626038,
+      "learning_rate": 1.4975e-05,
+      "loss": 1.3416,
+      "step": 1401
+    },
+    {
+      "epoch": 2.0947845169724983,
+      "grad_norm": 0.19980430603027344,
+      "learning_rate": 1.4950000000000001e-05,
+      "loss": 1.3569,
+      "step": 1402
+    },
+    {
+      "epoch": 2.0962786571415233,
+      "grad_norm": 0.1910398304462433,
+      "learning_rate": 1.4925e-05,
+      "loss": 1.402,
+      "step": 1403
+    },
+    {
+      "epoch": 2.0977727973105478,
+      "grad_norm": 0.2712978720664978,
+      "learning_rate": 1.49e-05,
+      "loss": 1.3505,
+      "step": 1404
+    },
+    {
+      "epoch": 2.0992669374795723,
+      "grad_norm": 0.3270023763179779,
+      "learning_rate": 1.4875e-05,
+      "loss": 1.2635,
+      "step": 1405
+    },
+    {
+      "epoch": 2.1007610776485968,
+      "grad_norm": 0.2294662743806839,
+      "learning_rate": 1.485e-05,
+      "loss": 1.151,
+      "step": 1406
+    },
+    {
+      "epoch": 2.1022552178176217,
+      "grad_norm": 4.064767360687256,
+      "learning_rate": 1.4825e-05,
+      "loss": 1.3138,
+      "step": 1407
+    },
+    {
+      "epoch": 2.103749357986646,
+      "grad_norm": 0.44238874316215515,
+      "learning_rate": 1.48e-05,
+      "loss": 1.2766,
+      "step": 1408
+    },
+    {
+      "epoch": 2.1052434981556707,
+      "grad_norm": 0.42204442620277405,
+      "learning_rate": 1.4775e-05,
+      "loss": 1.4241,
+      "step": 1409
+    },
+    {
+      "epoch": 2.106737638324695,
+      "grad_norm": 1.8588513135910034,
+      "learning_rate": 1.475e-05,
+      "loss": 1.297,
+      "step": 1410
+    },
+    {
+      "epoch": 2.10823177849372,
+      "grad_norm": 0.21745501458644867,
+      "learning_rate": 1.4725e-05,
+      "loss": 1.2065,
+      "step": 1411
+    },
+    {
+      "epoch": 2.1097259186627446,
+      "grad_norm": 0.22868776321411133,
+      "learning_rate": 1.47e-05,
+      "loss": 1.3406,
+      "step": 1412
+    },
+    {
+      "epoch": 2.111220058831769,
+      "grad_norm": 0.23604220151901245,
+      "learning_rate": 1.4675e-05,
+      "loss": 1.3829,
+      "step": 1413
+    },
+    {
+      "epoch": 2.1127141990007936,
+      "grad_norm": 0.19632968306541443,
+      "learning_rate": 1.465e-05,
+      "loss": 1.2526,
+      "step": 1414
+    },
+    {
+      "epoch": 2.1142083391698185,
+      "grad_norm": 0.20098775625228882,
+      "learning_rate": 1.4625e-05,
+      "loss": 1.1966,
+      "step": 1415
+    },
+    {
+      "epoch": 2.115702479338843,
+      "grad_norm": 0.181330606341362,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 1.4068,
+      "step": 1416
+    },
+    {
+      "epoch": 2.1171966195078675,
+      "grad_norm": 0.21224381029605865,
+      "learning_rate": 1.4575e-05,
+      "loss": 1.2888,
+      "step": 1417
+    },
+    {
+      "epoch": 2.118690759676892,
+      "grad_norm": 0.2324310541152954,
+      "learning_rate": 1.455e-05,
+      "loss": 1.2753,
+      "step": 1418
+    },
+    {
+      "epoch": 2.120184899845917,
+      "grad_norm": 0.18913941085338593,
+      "learning_rate": 1.4524999999999999e-05,
+      "loss": 1.2022,
+      "step": 1419
+    },
+    {
+      "epoch": 2.1216790400149415,
+      "grad_norm": 0.21584191918373108,
+      "learning_rate": 1.45e-05,
+      "loss": 1.3163,
+      "step": 1420
+    },
+    {
+      "epoch": 2.123173180183966,
+      "grad_norm": 0.17268019914627075,
+      "learning_rate": 1.4475e-05,
+      "loss": 1.1978,
+      "step": 1421
+    },
+    {
+      "epoch": 2.1246673203529904,
+      "grad_norm": 0.16772915422916412,
+      "learning_rate": 1.4449999999999999e-05,
+      "loss": 1.4418,
+      "step": 1422
+    },
+    {
+      "epoch": 2.1261614605220154,
+      "grad_norm": 0.2106754332780838,
+      "learning_rate": 1.4425e-05,
+      "loss": 1.4449,
+      "step": 1423
+    },
+    {
+      "epoch": 2.12765560069104,
+      "grad_norm": 0.3900804817676544,
+      "learning_rate": 1.44e-05,
+      "loss": 1.2949,
+      "step": 1424
+    },
+    {
+      "epoch": 2.1291497408600644,
+      "grad_norm": 0.21320056915283203,
+      "learning_rate": 1.4374999999999999e-05,
+      "loss": 1.1929,
+      "step": 1425
+    },
+    {
+      "epoch": 2.130643881029089,
+      "grad_norm": 0.8794408440589905,
+      "learning_rate": 1.435e-05,
+      "loss": 1.3492,
+      "step": 1426
+    },
+    {
+      "epoch": 2.132138021198114,
+      "grad_norm": 0.20707076787948608,
+      "learning_rate": 1.4325e-05,
+      "loss": 1.3354,
+      "step": 1427
+    },
+    {
+      "epoch": 2.1336321613671383,
+      "grad_norm": 0.19381563365459442,
+      "learning_rate": 1.43e-05,
+      "loss": 1.1734,
+      "step": 1428
+    },
+    {
+      "epoch": 2.135126301536163,
+      "grad_norm": 0.6543886065483093,
+      "learning_rate": 1.4275e-05,
+      "loss": 1.2938,
+      "step": 1429
+    },
+    {
+      "epoch": 2.1366204417051873,
+      "grad_norm": 0.2099548727273941,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 1.4015,
+      "step": 1430
+    },
+    {
+      "epoch": 2.138114581874212,
+      "grad_norm": 0.1843765527009964,
+      "learning_rate": 1.4225e-05,
+      "loss": 1.2783,
+      "step": 1431
+    },
+    {
+      "epoch": 2.1396087220432367,
+      "grad_norm": 0.17574813961982727,
+      "learning_rate": 1.42e-05,
+      "loss": 1.1302,
+      "step": 1432
+    },
+    {
+      "epoch": 2.141102862212261,
+      "grad_norm": 0.18504847586154938,
+      "learning_rate": 1.4174999999999999e-05,
+      "loss": 1.3835,
+      "step": 1433
+    },
+    {
+      "epoch": 2.1425970023812857,
+      "grad_norm": 0.5265726447105408,
+      "learning_rate": 1.415e-05,
+      "loss": 1.2911,
+      "step": 1434
+    },
+    {
+      "epoch": 2.1440911425503106,
+      "grad_norm": 0.2602658271789551,
+      "learning_rate": 1.4125e-05,
+      "loss": 1.1983,
+      "step": 1435
+    },
+    {
+      "epoch": 2.145585282719335,
+      "grad_norm": 0.3915673792362213,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 1.4647,
+      "step": 1436
+    },
+    {
+      "epoch": 2.1470794228883596,
+      "grad_norm": 0.22867648303508759,
+      "learning_rate": 1.4075e-05,
+      "loss": 1.2761,
+      "step": 1437
+    },
+    {
+      "epoch": 2.148573563057384,
+      "grad_norm": 0.3059532940387726,
+      "learning_rate": 1.4050000000000003e-05,
+      "loss": 1.2868,
+      "step": 1438
+    },
+    {
+      "epoch": 2.150067703226409,
+      "grad_norm": 0.17894840240478516,
+      "learning_rate": 1.4025000000000002e-05,
+      "loss": 1.2723,
+      "step": 1439
+    },
+    {
+      "epoch": 2.1515618433954335,
+      "grad_norm": 0.3326786458492279,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.2988,
+      "step": 1440
+    },
+    {
+      "epoch": 2.153055983564458,
+      "grad_norm": 0.24176494777202606,
+      "learning_rate": 1.3975000000000003e-05,
+      "loss": 1.5227,
+      "step": 1441
+    },
+    {
+      "epoch": 2.1545501237334825,
+      "grad_norm": 0.45723408460617065,
+      "learning_rate": 1.3950000000000002e-05,
+      "loss": 1.355,
+      "step": 1442
+    },
+    {
+      "epoch": 2.1560442639025075,
+      "grad_norm": 0.2570329010486603,
+      "learning_rate": 1.3925000000000001e-05,
+      "loss": 1.1168,
+      "step": 1443
+    },
+    {
+      "epoch": 2.157538404071532,
+      "grad_norm": 0.24646492302417755,
+      "learning_rate": 1.3900000000000002e-05,
+      "loss": 1.2781,
+      "step": 1444
+    },
+    {
+      "epoch": 2.1590325442405565,
+      "grad_norm": 0.21383704245090485,
+      "learning_rate": 1.3875000000000002e-05,
+      "loss": 1.4397,
+      "step": 1445
+    },
+    {
+      "epoch": 2.1605266844095814,
+      "grad_norm": 0.17543964087963104,
+      "learning_rate": 1.3850000000000001e-05,
+      "loss": 1.2952,
+      "step": 1446
+    },
+    {
+      "epoch": 2.162020824578606,
+      "grad_norm": 0.48758426308631897,
+      "learning_rate": 1.3825000000000002e-05,
+      "loss": 1.3334,
+      "step": 1447
+    },
+    {
+      "epoch": 2.1635149647476304,
+      "grad_norm": 0.25018617510795593,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 1.3461,
+      "step": 1448
+    },
+    {
+      "epoch": 2.165009104916655,
+      "grad_norm": 0.2622385621070862,
+      "learning_rate": 1.3775000000000001e-05,
+      "loss": 1.2779,
+      "step": 1449
+    },
+    {
+      "epoch": 2.1665032450856794,
+      "grad_norm": 0.22317945957183838,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 1.3226,
+      "step": 1450
+    },
+    {
+      "epoch": 2.1679973852547043,
+      "grad_norm": 0.25490185618400574,
+      "learning_rate": 1.3725000000000002e-05,
+      "loss": 1.2453,
+      "step": 1451
+    },
+    {
+      "epoch": 2.169491525423729,
+      "grad_norm": 0.28227943181991577,
+      "learning_rate": 1.3700000000000001e-05,
+      "loss": 1.2265,
+      "step": 1452
+    },
+    {
+      "epoch": 2.1709856655927533,
+      "grad_norm": 0.24756629765033722,
+      "learning_rate": 1.3675000000000002e-05,
+      "loss": 1.4324,
+      "step": 1453
+    },
+    {
+      "epoch": 2.1724798057617782,
+      "grad_norm": 0.2755368649959564,
+      "learning_rate": 1.3650000000000001e-05,
+      "loss": 1.3576,
+      "step": 1454
+    },
+    {
+      "epoch": 2.1739739459308027,
+      "grad_norm": 0.2696566879749298,
+      "learning_rate": 1.3625e-05,
+      "loss": 1.39,
+      "step": 1455
+    },
+    {
+      "epoch": 2.1754680860998272,
+      "grad_norm": 0.20134660601615906,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.2761,
+      "step": 1456
+    },
+    {
+      "epoch": 2.1769622262688517,
+      "grad_norm": 0.1893608272075653,
+      "learning_rate": 1.3575000000000001e-05,
+      "loss": 1.1057,
+      "step": 1457
+    },
+    {
+      "epoch": 2.1784563664378767,
+      "grad_norm": 0.26885709166526794,
+      "learning_rate": 1.3550000000000002e-05,
+      "loss": 1.4219,
+      "step": 1458
+    },
+    {
+      "epoch": 2.179950506606901,
+      "grad_norm": 0.3592870831489563,
+      "learning_rate": 1.3525000000000002e-05,
+      "loss": 1.2893,
+      "step": 1459
+    },
+    {
+      "epoch": 2.1814446467759256,
+      "grad_norm": 0.2307906299829483,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.295,
+      "step": 1460
+    },
+    {
+      "epoch": 2.18293878694495,
+      "grad_norm": 0.23158007860183716,
+      "learning_rate": 1.3475000000000002e-05,
+      "loss": 1.2771,
+      "step": 1461
+    },
+    {
+      "epoch": 2.184432927113975,
+      "grad_norm": 0.3418627977371216,
+      "learning_rate": 1.3450000000000002e-05,
+      "loss": 1.2596,
+      "step": 1462
+    },
+    {
+      "epoch": 2.1859270672829996,
+      "grad_norm": 0.18299835920333862,
+      "learning_rate": 1.3425000000000001e-05,
+      "loss": 1.3383,
+      "step": 1463
+    },
+    {
+      "epoch": 2.187421207452024,
+      "grad_norm": 0.28012511134147644,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 1.4001,
+      "step": 1464
+    },
+    {
+      "epoch": 2.1889153476210486,
+      "grad_norm": 0.26353952288627625,
+      "learning_rate": 1.3375000000000002e-05,
+      "loss": 1.3973,
+      "step": 1465
+    },
+    {
+      "epoch": 2.1904094877900735,
+      "grad_norm": 0.2554903030395508,
+      "learning_rate": 1.3350000000000001e-05,
+      "loss": 1.2561,
+      "step": 1466
+    },
+    {
+      "epoch": 2.191903627959098,
+      "grad_norm": 0.21797333657741547,
+      "learning_rate": 1.3325000000000002e-05,
+      "loss": 1.2779,
+      "step": 1467
+    },
+    {
+      "epoch": 2.1933977681281225,
+      "grad_norm": 0.18936985731124878,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 1.3408,
+      "step": 1468
+    },
+    {
+      "epoch": 2.194891908297147,
+      "grad_norm": 0.16586045920848846,
+      "learning_rate": 1.3275e-05,
+      "loss": 1.2249,
+      "step": 1469
+    },
+    {
+      "epoch": 2.196386048466172,
+      "grad_norm": 0.3664158582687378,
+      "learning_rate": 1.3250000000000002e-05,
+      "loss": 1.2892,
+      "step": 1470
+    },
+    {
+      "epoch": 2.1978801886351964,
+      "grad_norm": 0.16312362253665924,
+      "learning_rate": 1.3225000000000001e-05,
+      "loss": 1.2441,
+      "step": 1471
+    },
+    {
+      "epoch": 2.199374328804221,
+      "grad_norm": 0.5265755653381348,
+      "learning_rate": 1.32e-05,
+      "loss": 1.3866,
+      "step": 1472
+    },
+    {
+      "epoch": 2.2008684689732454,
+      "grad_norm": 0.2296518087387085,
+      "learning_rate": 1.3175000000000002e-05,
+      "loss": 1.3673,
+      "step": 1473
+    },
+    {
+      "epoch": 2.2023626091422703,
+      "grad_norm": 0.3048577308654785,
+      "learning_rate": 1.3150000000000001e-05,
+      "loss": 1.356,
+      "step": 1474
+    },
+    {
+      "epoch": 2.203856749311295,
+      "grad_norm": 0.2861064076423645,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.2939,
+      "step": 1475
+    },
+    {
+      "epoch": 2.2053508894803193,
+      "grad_norm": 0.24364018440246582,
+      "learning_rate": 1.3100000000000002e-05,
+      "loss": 1.4165,
+      "step": 1476
+    },
+    {
+      "epoch": 2.206845029649344,
+      "grad_norm": 0.21503858268260956,
+      "learning_rate": 1.3075000000000001e-05,
+      "loss": 1.2548,
+      "step": 1477
+    },
+    {
+      "epoch": 2.2083391698183688,
+      "grad_norm": 0.2088351547718048,
+      "learning_rate": 1.305e-05,
+      "loss": 1.2361,
+      "step": 1478
+    },
+    {
+      "epoch": 2.2098333099873932,
+      "grad_norm": 0.21245965361595154,
+      "learning_rate": 1.3025000000000002e-05,
+      "loss": 1.2242,
+      "step": 1479
+    },
+    {
+      "epoch": 2.2113274501564177,
+      "grad_norm": 0.2069002240896225,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.3018,
+      "step": 1480
+    },
+    {
+      "epoch": 2.2128215903254422,
+      "grad_norm": 0.182997465133667,
+      "learning_rate": 1.2975e-05,
+      "loss": 1.3056,
+      "step": 1481
+    },
+    {
+      "epoch": 2.214315730494467,
+      "grad_norm": 0.30062931776046753,
+      "learning_rate": 1.2950000000000001e-05,
+      "loss": 1.3128,
+      "step": 1482
+    },
+    {
+      "epoch": 2.2158098706634917,
+      "grad_norm": 0.16962677240371704,
+      "learning_rate": 1.2925e-05,
+      "loss": 1.2552,
+      "step": 1483
+    },
+    {
+      "epoch": 2.217304010832516,
+      "grad_norm": 0.22674013674259186,
+      "learning_rate": 1.29e-05,
+      "loss": 1.2199,
+      "step": 1484
+    },
+    {
+      "epoch": 2.2187981510015407,
+      "grad_norm": 0.30475592613220215,
+      "learning_rate": 1.2875000000000001e-05,
+      "loss": 1.3368,
+      "step": 1485
+    },
+    {
+      "epoch": 2.2202922911705656,
+      "grad_norm": 0.18455040454864502,
+      "learning_rate": 1.285e-05,
+      "loss": 1.2818,
+      "step": 1486
+    },
+    {
+      "epoch": 2.22178643133959,
+      "grad_norm": 0.1701553463935852,
+      "learning_rate": 1.2825000000000002e-05,
+      "loss": 1.2886,
+      "step": 1487
+    },
+    {
+      "epoch": 2.2232805715086146,
+      "grad_norm": 0.3963810205459595,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.3285,
+      "step": 1488
+    },
+    {
+      "epoch": 2.224774711677639,
+      "grad_norm": 0.3438395857810974,
+      "learning_rate": 1.2775e-05,
+      "loss": 1.2927,
+      "step": 1489
+    },
+    {
+      "epoch": 2.226268851846664,
+      "grad_norm": 0.21394677460193634,
+      "learning_rate": 1.2750000000000002e-05,
+      "loss": 1.3431,
+      "step": 1490
+    },
+    {
+      "epoch": 2.2277629920156885,
+      "grad_norm": 2.9313278198242188,
+      "learning_rate": 1.2725000000000001e-05,
+      "loss": 1.4353,
+      "step": 1491
+    },
+    {
+      "epoch": 2.229257132184713,
+      "grad_norm": 0.5035104751586914,
+      "learning_rate": 1.27e-05,
+      "loss": 1.4292,
+      "step": 1492
+    },
+    {
+      "epoch": 2.2307512723537375,
+      "grad_norm": 0.34209391474723816,
+      "learning_rate": 1.2675000000000001e-05,
+      "loss": 1.2921,
+      "step": 1493
+    },
+    {
+      "epoch": 2.2322454125227624,
+      "grad_norm": 0.26486945152282715,
+      "learning_rate": 1.2650000000000001e-05,
+      "loss": 1.2368,
+      "step": 1494
+    },
+    {
+      "epoch": 2.233739552691787,
+      "grad_norm": 0.21212176978588104,
+      "learning_rate": 1.2625e-05,
+      "loss": 1.5764,
+      "step": 1495
+    },
+    {
+      "epoch": 2.2352336928608114,
+      "grad_norm": 0.24310234189033508,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.4328,
+      "step": 1496
+    },
+    {
+      "epoch": 2.236727833029836,
+      "grad_norm": 0.20805980265140533,
+      "learning_rate": 1.2575e-05,
+      "loss": 1.149,
+      "step": 1497
+    },
+    {
+      "epoch": 2.238221973198861,
+      "grad_norm": 0.3014054596424103,
+      "learning_rate": 1.255e-05,
+      "loss": 1.2276,
+      "step": 1498
+    },
+    {
+      "epoch": 2.2397161133678853,
+      "grad_norm": 0.18532797694206238,
+      "learning_rate": 1.2525000000000001e-05,
+      "loss": 1.3269,
+      "step": 1499
+    },
+    {
+      "epoch": 2.24121025353691,
+      "grad_norm": 0.19393125176429749,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1997,
+      "step": 1500
+    },
+    {
+      "epoch": 2.2427043937059343,
+      "grad_norm": 0.3987423777580261,
+      "learning_rate": 1.2475e-05,
+      "loss": 1.4332,
+      "step": 1501
+    },
+    {
+      "epoch": 2.2441985338749593,
+      "grad_norm": 0.17318999767303467,
+      "learning_rate": 1.2450000000000001e-05,
+      "loss": 1.1957,
+      "step": 1502
+    },
+    {
+      "epoch": 2.2456926740439838,
+      "grad_norm": 1.6794114112854004,
+      "learning_rate": 1.2425e-05,
+      "loss": 1.3814,
+      "step": 1503
+    },
+    {
+      "epoch": 2.2471868142130083,
+      "grad_norm": 0.2993067502975464,
+      "learning_rate": 1.24e-05,
+      "loss": 1.2473,
+      "step": 1504
+    },
+    {
+      "epoch": 2.248680954382033,
+      "grad_norm": 0.35345566272735596,
+      "learning_rate": 1.2375000000000001e-05,
+      "loss": 1.315,
+      "step": 1505
+    },
+    {
+      "epoch": 2.2501750945510577,
+      "grad_norm": 0.24247638881206512,
+      "learning_rate": 1.235e-05,
+      "loss": 1.1762,
+      "step": 1506
+    },
+    {
+      "epoch": 2.251669234720082,
+      "grad_norm": 0.29046157002449036,
+      "learning_rate": 1.2325e-05,
+      "loss": 1.2739,
+      "step": 1507
+    },
+    {
+      "epoch": 2.2531633748891067,
+      "grad_norm": 0.22655948996543884,
+      "learning_rate": 1.23e-05,
+      "loss": 1.2653,
+      "step": 1508
+    },
+    {
+      "epoch": 2.254657515058131,
+      "grad_norm": 0.2086847871541977,
+      "learning_rate": 1.2275e-05,
+      "loss": 1.2846,
+      "step": 1509
+    },
+    {
+      "epoch": 2.256151655227156,
+      "grad_norm": 0.2213176190853119,
+      "learning_rate": 1.225e-05,
+      "loss": 1.2555,
+      "step": 1510
+    },
+    {
+      "epoch": 2.2576457953961806,
+      "grad_norm": 0.22715193033218384,
+      "learning_rate": 1.2225e-05,
+      "loss": 1.1853,
+      "step": 1511
+    },
+    {
+      "epoch": 2.259139935565205,
+      "grad_norm": 0.6379641890525818,
+      "learning_rate": 1.22e-05,
+      "loss": 1.3702,
+      "step": 1512
+    },
+    {
+      "epoch": 2.26063407573423,
+      "grad_norm": 0.37107372283935547,
+      "learning_rate": 1.2175e-05,
+      "loss": 1.2811,
+      "step": 1513
+    },
+    {
+      "epoch": 2.2621282159032545,
+      "grad_norm": 0.21601450443267822,
+      "learning_rate": 1.215e-05,
+      "loss": 1.5273,
+      "step": 1514
+    },
+    {
+      "epoch": 2.263622356072279,
+      "grad_norm": 0.2155398428440094,
+      "learning_rate": 1.2125e-05,
+      "loss": 1.3442,
+      "step": 1515
+    },
+    {
+      "epoch": 2.2651164962413035,
+      "grad_norm": 0.2486492097377777,
+      "learning_rate": 1.2100000000000001e-05,
+      "loss": 1.4766,
+      "step": 1516
+    },
+    {
+      "epoch": 2.266610636410328,
+      "grad_norm": 0.2049209475517273,
+      "learning_rate": 1.2075e-05,
+      "loss": 1.1512,
+      "step": 1517
+    },
+    {
+      "epoch": 2.268104776579353,
+      "grad_norm": 0.2839110493659973,
+      "learning_rate": 1.205e-05,
+      "loss": 1.3691,
+      "step": 1518
+    },
+    {
+      "epoch": 2.2695989167483774,
+      "grad_norm": 0.25949689745903015,
+      "learning_rate": 1.2025000000000001e-05,
+      "loss": 1.4586,
+      "step": 1519
+    },
+    {
+      "epoch": 2.271093056917402,
+      "grad_norm": 0.2694650888442993,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2326,
+      "step": 1520
+    },
+    {
+      "epoch": 2.272587197086427,
+      "grad_norm": 0.30505526065826416,
+      "learning_rate": 1.1975e-05,
+      "loss": 1.4554,
+      "step": 1521
+    },
+    {
+      "epoch": 2.2740813372554514,
+      "grad_norm": 0.20058128237724304,
+      "learning_rate": 1.195e-05,
+      "loss": 1.28,
+      "step": 1522
+    },
+    {
+      "epoch": 2.275575477424476,
+      "grad_norm": 0.2585764527320862,
+      "learning_rate": 1.1925e-05,
+      "loss": 1.2229,
+      "step": 1523
+    },
+    {
+      "epoch": 2.2770696175935004,
+      "grad_norm": 0.6231411695480347,
+      "learning_rate": 1.19e-05,
+      "loss": 1.3299,
+      "step": 1524
+    },
+    {
+      "epoch": 2.2785637577625253,
+      "grad_norm": 0.14931946992874146,
+      "learning_rate": 1.1875e-05,
+      "loss": 1.2723,
+      "step": 1525
+    },
+    {
+      "epoch": 2.28005789793155,
+      "grad_norm": 0.1917908936738968,
+      "learning_rate": 1.185e-05,
+      "loss": 1.2713,
+      "step": 1526
+    },
+    {
+      "epoch": 2.2815520381005743,
+      "grad_norm": 0.17372849583625793,
+      "learning_rate": 1.1825e-05,
+      "loss": 1.2496,
+      "step": 1527
+    },
+    {
+      "epoch": 2.2830461782695988,
+      "grad_norm": 0.18889087438583374,
+      "learning_rate": 1.18e-05,
+      "loss": 1.2819,
+      "step": 1528
+    },
+    {
+      "epoch": 2.2845403184386237,
+      "grad_norm": 0.23077258467674255,
+      "learning_rate": 1.1775e-05,
+      "loss": 1.1978,
+      "step": 1529
+    },
+    {
+      "epoch": 2.286034458607648,
+      "grad_norm": 0.24370890855789185,
+      "learning_rate": 1.175e-05,
+      "loss": 1.3296,
+      "step": 1530
+    },
+    {
+      "epoch": 2.2875285987766727,
+      "grad_norm": 0.5306088924407959,
+      "learning_rate": 1.1725e-05,
+      "loss": 1.2872,
+      "step": 1531
+    },
+    {
+      "epoch": 2.289022738945697,
+      "grad_norm": 0.15988537669181824,
+      "learning_rate": 1.1700000000000001e-05,
+      "loss": 1.2157,
+      "step": 1532
+    },
+    {
+      "epoch": 2.290516879114722,
+      "grad_norm": 0.2093038707971573,
+      "learning_rate": 1.1675000000000001e-05,
+      "loss": 1.1948,
+      "step": 1533
+    },
+    {
+      "epoch": 2.2920110192837466,
+      "grad_norm": 0.24095523357391357,
+      "learning_rate": 1.1650000000000002e-05,
+      "loss": 1.1789,
+      "step": 1534
+    },
+    {
+      "epoch": 2.293505159452771,
+      "grad_norm": 0.3092842102050781,
+      "learning_rate": 1.1625000000000001e-05,
+      "loss": 1.3552,
+      "step": 1535
+    },
+    {
+      "epoch": 2.2949992996217956,
+      "grad_norm": 0.32683873176574707,
+      "learning_rate": 1.16e-05,
+      "loss": 1.3141,
+      "step": 1536
+    },
+    {
+      "epoch": 2.2964934397908205,
+      "grad_norm": 0.24342307448387146,
+      "learning_rate": 1.1575000000000002e-05,
+      "loss": 1.2125,
+      "step": 1537
+    },
+    {
+      "epoch": 2.297987579959845,
+      "grad_norm": 0.21990644931793213,
+      "learning_rate": 1.1550000000000001e-05,
+      "loss": 1.1445,
+      "step": 1538
+    },
+    {
+      "epoch": 2.2994817201288695,
+      "grad_norm": 0.19987203180789948,
+      "learning_rate": 1.1525e-05,
+      "loss": 1.2947,
+      "step": 1539
+    },
+    {
+      "epoch": 2.300975860297894,
+      "grad_norm": 0.21032652258872986,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 1.333,
+      "step": 1540
+    },
+    {
+      "epoch": 2.302470000466919,
+      "grad_norm": 0.5228366255760193,
+      "learning_rate": 1.1475000000000001e-05,
+      "loss": 1.3451,
+      "step": 1541
+    },
+    {
+      "epoch": 2.3039641406359435,
+      "grad_norm": 0.2448897361755371,
+      "learning_rate": 1.145e-05,
+      "loss": 1.3963,
+      "step": 1542
+    },
+    {
+      "epoch": 2.305458280804968,
+      "grad_norm": 0.27660074830055237,
+      "learning_rate": 1.1425000000000002e-05,
+      "loss": 1.2531,
+      "step": 1543
+    },
+    {
+      "epoch": 2.3069524209739924,
+      "grad_norm": 0.24510708451271057,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 1.2857,
+      "step": 1544
+    },
+    {
+      "epoch": 2.3084465611430174,
+      "grad_norm": 0.19473227858543396,
+      "learning_rate": 1.1375e-05,
+      "loss": 1.237,
+      "step": 1545
+    },
+    {
+      "epoch": 2.309940701312042,
+      "grad_norm": 0.1733362078666687,
+      "learning_rate": 1.1350000000000001e-05,
+      "loss": 1.3782,
+      "step": 1546
+    },
+    {
+      "epoch": 2.3114348414810664,
+      "grad_norm": 0.21493330597877502,
+      "learning_rate": 1.1325e-05,
+      "loss": 1.2649,
+      "step": 1547
+    },
+    {
+      "epoch": 2.312928981650091,
+      "grad_norm": 0.3164669871330261,
+      "learning_rate": 1.13e-05,
+      "loss": 1.4352,
+      "step": 1548
+    },
+    {
+      "epoch": 2.314423121819116,
+      "grad_norm": 0.26332464814186096,
+      "learning_rate": 1.1275000000000001e-05,
+      "loss": 1.4147,
+      "step": 1549
+    },
+    {
+      "epoch": 2.3159172619881403,
+      "grad_norm": 0.21300271153450012,
+      "learning_rate": 1.125e-05,
+      "loss": 1.3022,
+      "step": 1550
+    },
+    {
+      "epoch": 2.317411402157165,
+      "grad_norm": 0.20926804840564728,
+      "learning_rate": 1.1225e-05,
+      "loss": 1.4585,
+      "step": 1551
+    },
+    {
+      "epoch": 2.3189055423261893,
+      "grad_norm": 0.19575107097625732,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.2977,
+      "step": 1552
+    },
+    {
+      "epoch": 2.320399682495214,
+      "grad_norm": 0.369392067193985,
+      "learning_rate": 1.1175e-05,
+      "loss": 1.2781,
+      "step": 1553
+    },
+    {
+      "epoch": 2.3218938226642387,
+      "grad_norm": 0.21531105041503906,
+      "learning_rate": 1.115e-05,
+      "loss": 1.1297,
+      "step": 1554
+    },
+    {
+      "epoch": 2.323387962833263,
+      "grad_norm": 0.2804863750934601,
+      "learning_rate": 1.1125000000000001e-05,
+      "loss": 1.3836,
+      "step": 1555
+    },
+    {
+      "epoch": 2.324882103002288,
+      "grad_norm": 0.23862357437610626,
+      "learning_rate": 1.11e-05,
+      "loss": 1.3508,
+      "step": 1556
+    },
+    {
+      "epoch": 2.3263762431713126,
+      "grad_norm": 0.2789822816848755,
+      "learning_rate": 1.1075e-05,
+      "loss": 1.3357,
+      "step": 1557
+    },
+    {
+      "epoch": 2.327870383340337,
+      "grad_norm": 0.4456028640270233,
+      "learning_rate": 1.1050000000000001e-05,
+      "loss": 1.259,
+      "step": 1558
+    },
+    {
+      "epoch": 2.3293645235093616,
+      "grad_norm": 0.25457635521888733,
+      "learning_rate": 1.1025e-05,
+      "loss": 1.2132,
+      "step": 1559
+    },
+    {
+      "epoch": 2.330858663678386,
+      "grad_norm": 0.4204428791999817,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.3101,
+      "step": 1560
+    },
+    {
+      "epoch": 2.332352803847411,
+      "grad_norm": 0.3540639877319336,
+      "learning_rate": 1.0975e-05,
+      "loss": 1.3815,
+      "step": 1561
+    },
+    {
+      "epoch": 2.3338469440164356,
+      "grad_norm": 0.2761100232601166,
+      "learning_rate": 1.095e-05,
+      "loss": 1.3601,
+      "step": 1562
+    },
+    {
+      "epoch": 2.33534108418546,
+      "grad_norm": 0.29980364441871643,
+      "learning_rate": 1.0925000000000001e-05,
+      "loss": 1.2059,
+      "step": 1563
+    },
+    {
+      "epoch": 2.336835224354485,
+      "grad_norm": 0.24911551177501678,
+      "learning_rate": 1.09e-05,
+      "loss": 1.2788,
+      "step": 1564
+    },
+    {
+      "epoch": 2.3383293645235095,
+      "grad_norm": 0.37243562936782837,
+      "learning_rate": 1.0875e-05,
+      "loss": 1.1434,
+      "step": 1565
+    },
+    {
+      "epoch": 2.339823504692534,
+      "grad_norm": 0.2589317560195923,
+      "learning_rate": 1.0850000000000001e-05,
+      "loss": 1.397,
+      "step": 1566
+    },
+    {
+      "epoch": 2.3413176448615585,
+      "grad_norm": 0.21287329494953156,
+      "learning_rate": 1.0825e-05,
+      "loss": 1.076,
+      "step": 1567
+    },
+    {
+      "epoch": 2.342811785030583,
+      "grad_norm": 0.17493118345737457,
+      "learning_rate": 1.08e-05,
+      "loss": 1.3775,
+      "step": 1568
+    },
+    {
+      "epoch": 2.344305925199608,
+      "grad_norm": 0.24420218169689178,
+      "learning_rate": 1.0775000000000001e-05,
+      "loss": 1.4579,
+      "step": 1569
+    },
+    {
+      "epoch": 2.3458000653686324,
+      "grad_norm": 0.18936094641685486,
+      "learning_rate": 1.075e-05,
+      "loss": 1.3076,
+      "step": 1570
+    },
+    {
+      "epoch": 2.347294205537657,
+      "grad_norm": 0.24192200601100922,
+      "learning_rate": 1.0725e-05,
+      "loss": 1.3495,
+      "step": 1571
+    },
+    {
+      "epoch": 2.348788345706682,
+      "grad_norm": 0.32218971848487854,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 1.4041,
+      "step": 1572
+    },
+    {
+      "epoch": 2.3502824858757063,
+      "grad_norm": 0.2487693578004837,
+      "learning_rate": 1.0675e-05,
+      "loss": 1.2895,
+      "step": 1573
+    },
+    {
+      "epoch": 2.351776626044731,
+      "grad_norm": 0.14848390221595764,
+      "learning_rate": 1.065e-05,
+      "loss": 1.2928,
+      "step": 1574
+    },
+    {
+      "epoch": 2.3532707662137553,
+      "grad_norm": 0.19325977563858032,
+      "learning_rate": 1.0625e-05,
+      "loss": 1.2783,
+      "step": 1575
+    },
+    {
+      "epoch": 2.35476490638278,
+      "grad_norm": 0.293878436088562,
+      "learning_rate": 1.06e-05,
+      "loss": 1.3369,
+      "step": 1576
+    },
+    {
+      "epoch": 2.3562590465518047,
+      "grad_norm": 0.2537538707256317,
+      "learning_rate": 1.0575e-05,
+      "loss": 1.3704,
+      "step": 1577
+    },
+    {
+      "epoch": 2.3577531867208292,
+      "grad_norm": 0.41805949807167053,
+      "learning_rate": 1.055e-05,
+      "loss": 1.3075,
+      "step": 1578
+    },
+    {
+      "epoch": 2.3592473268898537,
+      "grad_norm": 0.20566672086715698,
+      "learning_rate": 1.0525e-05,
+      "loss": 1.1988,
+      "step": 1579
+    },
+    {
+      "epoch": 2.3607414670588787,
+      "grad_norm": 0.21227644383907318,
+      "learning_rate": 1.05e-05,
+      "loss": 1.3543,
+      "step": 1580
+    },
+    {
+      "epoch": 2.362235607227903,
+      "grad_norm": 0.30649107694625854,
+      "learning_rate": 1.0475e-05,
+      "loss": 1.3404,
+      "step": 1581
+    },
+    {
+      "epoch": 2.3637297473969276,
+      "grad_norm": 0.2410416305065155,
+      "learning_rate": 1.045e-05,
+      "loss": 1.4158,
+      "step": 1582
+    },
+    {
+      "epoch": 2.365223887565952,
+      "grad_norm": 0.1789264678955078,
+      "learning_rate": 1.0425e-05,
+      "loss": 1.4128,
+      "step": 1583
+    },
+    {
+      "epoch": 2.366718027734977,
+      "grad_norm": 0.280666708946228,
+      "learning_rate": 1.04e-05,
+      "loss": 1.3142,
+      "step": 1584
+    },
+    {
+      "epoch": 2.3682121679040016,
+      "grad_norm": 0.2530580461025238,
+      "learning_rate": 1.0375e-05,
+      "loss": 1.2719,
+      "step": 1585
+    },
+    {
+      "epoch": 2.369706308073026,
+      "grad_norm": 0.3545680642127991,
+      "learning_rate": 1.035e-05,
+      "loss": 1.354,
+      "step": 1586
+    },
+    {
+      "epoch": 2.3712004482420506,
+      "grad_norm": 0.4236186742782593,
+      "learning_rate": 1.0325e-05,
+      "loss": 1.0796,
+      "step": 1587
+    },
+    {
+      "epoch": 2.3726945884110755,
+      "grad_norm": 0.22073593735694885,
+      "learning_rate": 1.03e-05,
+      "loss": 1.3344,
+      "step": 1588
+    },
+    {
+      "epoch": 2.3741887285801,
+      "grad_norm": 0.2675930857658386,
+      "learning_rate": 1.0275e-05,
+      "loss": 1.3888,
+      "step": 1589
+    },
+    {
+      "epoch": 2.3756828687491245,
+      "grad_norm": 0.24186880886554718,
+      "learning_rate": 1.025e-05,
+      "loss": 1.158,
+      "step": 1590
+    },
+    {
+      "epoch": 2.377177008918149,
+      "grad_norm": 0.36522793769836426,
+      "learning_rate": 1.0225e-05,
+      "loss": 1.3172,
+      "step": 1591
+    },
+    {
+      "epoch": 2.378671149087174,
+      "grad_norm": 0.33718815445899963,
+      "learning_rate": 1.02e-05,
+      "loss": 1.3242,
+      "step": 1592
+    },
+    {
+      "epoch": 2.3801652892561984,
+      "grad_norm": 0.19472959637641907,
+      "learning_rate": 1.0175e-05,
+      "loss": 1.0996,
+      "step": 1593
+    },
+    {
+      "epoch": 2.381659429425223,
+      "grad_norm": 0.2957690954208374,
+      "learning_rate": 1.0150000000000001e-05,
+      "loss": 1.4084,
+      "step": 1594
+    },
+    {
+      "epoch": 2.3831535695942474,
+      "grad_norm": 0.27677658200263977,
+      "learning_rate": 1.0125e-05,
+      "loss": 1.3074,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3846477097632723,
+      "grad_norm": 0.2773936092853546,
+      "learning_rate": 1.0100000000000002e-05,
+      "loss": 1.2948,
+      "step": 1596
+    },
+    {
+      "epoch": 2.386141849932297,
+      "grad_norm": 0.3917483687400818,
+      "learning_rate": 1.0075000000000001e-05,
+      "loss": 1.325,
+      "step": 1597
+    },
+    {
+      "epoch": 2.3876359901013213,
+      "grad_norm": 0.20603208243846893,
+      "learning_rate": 1.005e-05,
+      "loss": 1.1257,
+      "step": 1598
+    },
+    {
+      "epoch": 2.389130130270346,
+      "grad_norm": 0.2912125885486603,
+      "learning_rate": 1.0025000000000001e-05,
+      "loss": 1.2564,
+      "step": 1599
+    },
+    {
+      "epoch": 2.3906242704393708,
+      "grad_norm": 0.2543962001800537,
+      "learning_rate": 1e-05,
+      "loss": 1.241,
+      "step": 1600
+    },
+    {
+      "epoch": 2.3921184106083953,
+      "grad_norm": 0.27241256833076477,
+      "learning_rate": 9.975e-06,
+      "loss": 1.3456,
+      "step": 1601
+    },
+    {
+      "epoch": 2.3936125507774197,
+      "grad_norm": 0.3361535668373108,
+      "learning_rate": 9.950000000000001e-06,
+      "loss": 1.2202,
+      "step": 1602
+    },
+    {
+      "epoch": 2.3951066909464442,
+      "grad_norm": 0.3284830152988434,
+      "learning_rate": 9.925e-06,
+      "loss": 1.2916,
+      "step": 1603
+    },
+    {
+      "epoch": 2.396600831115469,
+      "grad_norm": 0.2478102296590805,
+      "learning_rate": 9.900000000000002e-06,
+      "loss": 1.3076,
+      "step": 1604
+    },
+    {
+      "epoch": 2.3980949712844937,
+      "grad_norm": 0.19693656265735626,
+      "learning_rate": 9.875000000000001e-06,
+      "loss": 1.3351,
+      "step": 1605
+    },
+    {
+      "epoch": 2.399589111453518,
+      "grad_norm": 0.2959229350090027,
+      "learning_rate": 9.85e-06,
+      "loss": 1.2544,
+      "step": 1606
+    },
+    {
+      "epoch": 2.4010832516225427,
+      "grad_norm": 0.18445654213428497,
+      "learning_rate": 9.825000000000002e-06,
+      "loss": 1.2567,
+      "step": 1607
+    },
+    {
+      "epoch": 2.4025773917915676,
+      "grad_norm": 0.2136373221874237,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.5701,
+      "step": 1608
+    },
+    {
+      "epoch": 2.404071531960592,
+      "grad_norm": 0.23023132979869843,
+      "learning_rate": 9.775e-06,
+      "loss": 1.2132,
+      "step": 1609
+    },
+    {
+      "epoch": 2.4055656721296166,
+      "grad_norm": 0.23368653655052185,
+      "learning_rate": 9.750000000000002e-06,
+      "loss": 1.482,
+      "step": 1610
+    },
+    {
+      "epoch": 2.407059812298641,
+      "grad_norm": 0.2301778644323349,
+      "learning_rate": 9.725000000000001e-06,
+      "loss": 1.4484,
+      "step": 1611
+    },
+    {
+      "epoch": 2.408553952467666,
+      "grad_norm": 0.20297956466674805,
+      "learning_rate": 9.7e-06,
+      "loss": 1.304,
+      "step": 1612
+    },
+    {
+      "epoch": 2.4100480926366905,
+      "grad_norm": 0.207230806350708,
+      "learning_rate": 9.675000000000001e-06,
+      "loss": 1.2294,
+      "step": 1613
+    },
+    {
+      "epoch": 2.411542232805715,
+      "grad_norm": 0.2300531417131424,
+      "learning_rate": 9.65e-06,
+      "loss": 1.4689,
+      "step": 1614
+    },
+    {
+      "epoch": 2.41303637297474,
+      "grad_norm": 0.26880812644958496,
+      "learning_rate": 9.625e-06,
+      "loss": 1.3848,
+      "step": 1615
+    },
+    {
+      "epoch": 2.4145305131437644,
+      "grad_norm": 0.23205740749835968,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.3925,
+      "step": 1616
+    },
+    {
+      "epoch": 2.416024653312789,
+      "grad_norm": 0.21103398501873016,
+      "learning_rate": 9.575e-06,
+      "loss": 1.1464,
+      "step": 1617
+    },
+    {
+      "epoch": 2.4175187934818134,
+      "grad_norm": 0.2809285819530487,
+      "learning_rate": 9.55e-06,
+      "loss": 1.3438,
+      "step": 1618
+    },
+    {
+      "epoch": 2.419012933650838,
+      "grad_norm": 0.19656510651111603,
+      "learning_rate": 9.525000000000001e-06,
+      "loss": 1.2569,
+      "step": 1619
+    },
+    {
+      "epoch": 2.420507073819863,
+      "grad_norm": 0.30520132184028625,
+      "learning_rate": 9.5e-06,
+      "loss": 1.395,
+      "step": 1620
+    },
+    {
+      "epoch": 2.4220012139888873,
+      "grad_norm": 0.18955890834331512,
+      "learning_rate": 9.475e-06,
+      "loss": 1.2901,
+      "step": 1621
+    },
+    {
+      "epoch": 2.423495354157912,
+      "grad_norm": 0.18221187591552734,
+      "learning_rate": 9.450000000000001e-06,
+      "loss": 1.2954,
+      "step": 1622
+    },
+    {
+      "epoch": 2.424989494326937,
+      "grad_norm": 0.19730985164642334,
+      "learning_rate": 9.425e-06,
+      "loss": 1.3947,
+      "step": 1623
+    },
+    {
+      "epoch": 2.4264836344959613,
+      "grad_norm": 0.2914806604385376,
+      "learning_rate": 9.4e-06,
+      "loss": 1.3879,
+      "step": 1624
+    },
+    {
+      "epoch": 2.4279777746649858,
+      "grad_norm": 0.20165985822677612,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.2932,
+      "step": 1625
+    },
+    {
+      "epoch": 2.4294719148340103,
+      "grad_norm": 0.39926818013191223,
+      "learning_rate": 9.35e-06,
+      "loss": 1.2433,
+      "step": 1626
+    },
+    {
+      "epoch": 2.4309660550030348,
+      "grad_norm": 0.32672664523124695,
+      "learning_rate": 9.325e-06,
+      "loss": 1.2686,
+      "step": 1627
+    },
+    {
+      "epoch": 2.4324601951720597,
+      "grad_norm": 0.2178495228290558,
+      "learning_rate": 9.3e-06,
+      "loss": 1.4247,
+      "step": 1628
+    },
+    {
+      "epoch": 2.433954335341084,
+      "grad_norm": 0.24548844993114471,
+      "learning_rate": 9.275e-06,
+      "loss": 1.3101,
+      "step": 1629
+    },
+    {
+      "epoch": 2.4354484755101087,
+      "grad_norm": 0.39940354228019714,
+      "learning_rate": 9.25e-06,
+      "loss": 1.4471,
+      "step": 1630
+    },
+    {
+      "epoch": 2.4369426156791336,
+      "grad_norm": 0.2500586211681366,
+      "learning_rate": 9.225e-06,
+      "loss": 1.316,
+      "step": 1631
+    },
+    {
+      "epoch": 2.438436755848158,
+      "grad_norm": 0.18969309329986572,
+      "learning_rate": 9.2e-06,
+      "loss": 1.1965,
+      "step": 1632
+    },
+    {
+      "epoch": 2.4399308960171826,
+      "grad_norm": 0.2904510498046875,
+      "learning_rate": 9.175000000000001e-06,
+      "loss": 1.1452,
+      "step": 1633
+    },
+    {
+      "epoch": 2.441425036186207,
+      "grad_norm": 0.18344439566135406,
+      "learning_rate": 9.15e-06,
+      "loss": 1.4705,
+      "step": 1634
+    },
+    {
+      "epoch": 2.4429191763552316,
+      "grad_norm": 0.2132856845855713,
+      "learning_rate": 9.125e-06,
+      "loss": 1.2666,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4444133165242565,
+      "grad_norm": 0.26673436164855957,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 1.3129,
+      "step": 1636
+    },
+    {
+      "epoch": 2.445907456693281,
+      "grad_norm": 0.29528361558914185,
+      "learning_rate": 9.075e-06,
+      "loss": 1.2356,
+      "step": 1637
+    },
+    {
+      "epoch": 2.4474015968623055,
+      "grad_norm": 0.23240934312343597,
+      "learning_rate": 9.05e-06,
+      "loss": 1.4241,
+      "step": 1638
+    },
+    {
+      "epoch": 2.4488957370313305,
+      "grad_norm": 0.19893576204776764,
+      "learning_rate": 9.025e-06,
+      "loss": 1.1293,
+      "step": 1639
+    },
+    {
+      "epoch": 2.450389877200355,
+      "grad_norm": 0.16400805115699768,
+      "learning_rate": 9e-06,
+      "loss": 1.232,
+      "step": 1640
+    },
+    {
+      "epoch": 2.4518840173693794,
+      "grad_norm": 0.1535254716873169,
+      "learning_rate": 8.975e-06,
+      "loss": 1.242,
+      "step": 1641
+    },
+    {
+      "epoch": 2.453378157538404,
+      "grad_norm": 0.35040244460105896,
+      "learning_rate": 8.95e-06,
+      "loss": 1.1444,
+      "step": 1642
+    },
+    {
+      "epoch": 2.4548722977074284,
+      "grad_norm": 0.20848290622234344,
+      "learning_rate": 8.925e-06,
+      "loss": 1.3863,
+      "step": 1643
+    },
+    {
+      "epoch": 2.4563664378764534,
+      "grad_norm": 0.22934338450431824,
+      "learning_rate": 8.9e-06,
+      "loss": 1.199,
+      "step": 1644
+    },
+    {
+      "epoch": 2.457860578045478,
+      "grad_norm": 0.19408972561359406,
+      "learning_rate": 8.875e-06,
+      "loss": 1.2936,
+      "step": 1645
+    },
+    {
+      "epoch": 2.4593547182145024,
+      "grad_norm": 0.21626761555671692,
+      "learning_rate": 8.85e-06,
+      "loss": 1.2033,
+      "step": 1646
+    },
+    {
+      "epoch": 2.4608488583835273,
+      "grad_norm": 0.19572362303733826,
+      "learning_rate": 8.825e-06,
+      "loss": 1.4063,
+      "step": 1647
+    },
+    {
+      "epoch": 2.462342998552552,
+      "grad_norm": 0.24215538799762726,
+      "learning_rate": 8.8e-06,
+      "loss": 1.2936,
+      "step": 1648
+    },
+    {
+      "epoch": 2.4638371387215763,
+      "grad_norm": 0.19395551085472107,
+      "learning_rate": 8.775e-06,
+      "loss": 1.3743,
+      "step": 1649
+    },
+    {
+      "epoch": 2.4653312788906008,
+      "grad_norm": 0.2217566967010498,
+      "learning_rate": 8.75e-06,
+      "loss": 1.2934,
+      "step": 1650
+    },
+    {
+      "epoch": 2.4668254190596257,
+      "grad_norm": 0.20734982192516327,
+      "learning_rate": 8.725e-06,
+      "loss": 1.3652,
+      "step": 1651
+    },
+    {
+      "epoch": 2.46831955922865,
+      "grad_norm": 0.21142613887786865,
+      "learning_rate": 8.7e-06,
+      "loss": 1.3618,
+      "step": 1652
+    },
+    {
+      "epoch": 2.4698136993976747,
+      "grad_norm": 0.3824837803840637,
+      "learning_rate": 8.674999999999999e-06,
+      "loss": 1.442,
+      "step": 1653
+    },
+    {
+      "epoch": 2.471307839566699,
+      "grad_norm": 0.3306943476200104,
+      "learning_rate": 8.65e-06,
+      "loss": 1.3585,
+      "step": 1654
+    },
+    {
+      "epoch": 2.472801979735724,
+      "grad_norm": 0.6740633845329285,
+      "learning_rate": 8.625e-06,
+      "loss": 1.4169,
+      "step": 1655
+    },
+    {
+      "epoch": 2.4742961199047486,
+      "grad_norm": 0.19902844727039337,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 1.3756,
+      "step": 1656
+    },
+    {
+      "epoch": 2.475790260073773,
+      "grad_norm": 0.3373403251171112,
+      "learning_rate": 8.575000000000002e-06,
+      "loss": 1.2222,
+      "step": 1657
+    },
+    {
+      "epoch": 2.4772844002427976,
+      "grad_norm": 0.21016831696033478,
+      "learning_rate": 8.550000000000001e-06,
+      "loss": 1.2962,
+      "step": 1658
+    },
+    {
+      "epoch": 2.4787785404118226,
+      "grad_norm": 0.241668701171875,
+      "learning_rate": 8.525e-06,
+      "loss": 1.2163,
+      "step": 1659
+    },
+    {
+      "epoch": 2.480272680580847,
+      "grad_norm": 0.2818886637687683,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 1.3332,
+      "step": 1660
+    },
+    {
+      "epoch": 2.4817668207498715,
+      "grad_norm": 0.23386937379837036,
+      "learning_rate": 8.475000000000001e-06,
+      "loss": 1.4589,
+      "step": 1661
+    },
+    {
+      "epoch": 2.483260960918896,
+      "grad_norm": 0.31478041410446167,
+      "learning_rate": 8.45e-06,
+      "loss": 1.3337,
+      "step": 1662
+    },
+    {
+      "epoch": 2.484755101087921,
+      "grad_norm": 0.1997414529323578,
+      "learning_rate": 8.425000000000001e-06,
+      "loss": 1.1732,
+      "step": 1663
+    },
+    {
+      "epoch": 2.4862492412569455,
+      "grad_norm": 0.27734795212745667,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.2748,
+      "step": 1664
+    },
+    {
+      "epoch": 2.48774338142597,
+      "grad_norm": 0.23954452574253082,
+      "learning_rate": 8.375e-06,
+      "loss": 1.2237,
+      "step": 1665
+    },
+    {
+      "epoch": 2.4892375215949945,
+      "grad_norm": 0.27997615933418274,
+      "learning_rate": 8.350000000000001e-06,
+      "loss": 1.3232,
+      "step": 1666
+    },
+    {
+      "epoch": 2.4907316617640194,
+      "grad_norm": 0.1582835614681244,
+      "learning_rate": 8.325e-06,
+      "loss": 1.2447,
+      "step": 1667
+    },
+    {
+      "epoch": 2.492225801933044,
+      "grad_norm": 0.9380664229393005,
+      "learning_rate": 8.3e-06,
+      "loss": 1.363,
+      "step": 1668
+    },
+    {
+      "epoch": 2.4937199421020684,
+      "grad_norm": 0.2942129969596863,
+      "learning_rate": 8.275000000000001e-06,
+      "loss": 1.3214,
+      "step": 1669
+    },
+    {
+      "epoch": 2.495214082271093,
+      "grad_norm": 0.6295128464698792,
+      "learning_rate": 8.25e-06,
+      "loss": 1.3361,
+      "step": 1670
+    },
+    {
+      "epoch": 2.496708222440118,
+      "grad_norm": 0.19411110877990723,
+      "learning_rate": 8.225e-06,
+      "loss": 1.2091,
+      "step": 1671
+    },
+    {
+      "epoch": 2.4982023626091423,
+      "grad_norm": 0.2738019824028015,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 1.5115,
+      "step": 1672
+    },
+    {
+      "epoch": 2.499696502778167,
+      "grad_norm": 0.21079149842262268,
+      "learning_rate": 8.175e-06,
+      "loss": 1.2621,
+      "step": 1673
+    },
+    {
+      "epoch": 2.5011906429471917,
+      "grad_norm": 0.34388652443885803,
+      "learning_rate": 8.15e-06,
+      "loss": 1.2276,
+      "step": 1674
+    },
+    {
+      "epoch": 2.5026847831162162,
+      "grad_norm": 0.25901058316230774,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.2417,
+      "step": 1675
+    },
+    {
+      "epoch": 2.5041789232852407,
+      "grad_norm": 0.39429500699043274,
+      "learning_rate": 8.1e-06,
+      "loss": 1.3081,
+      "step": 1676
+    },
+    {
+      "epoch": 2.505673063454265,
+      "grad_norm": 0.3238086700439453,
+      "learning_rate": 8.075000000000001e-06,
+      "loss": 1.4371,
+      "step": 1677
+    },
+    {
+      "epoch": 2.5071672036232897,
+      "grad_norm": 0.2773047387599945,
+      "learning_rate": 8.050000000000001e-06,
+      "loss": 1.287,
+      "step": 1678
+    },
+    {
+      "epoch": 2.5086613437923146,
+      "grad_norm": 0.20827920734882355,
+      "learning_rate": 8.025e-06,
+      "loss": 1.2057,
+      "step": 1679
+    },
+    {
+      "epoch": 2.510155483961339,
+      "grad_norm": 0.18946285545825958,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.3255,
+      "step": 1680
+    },
+    {
+      "epoch": 2.5116496241303636,
+      "grad_norm": 0.2746080160140991,
+      "learning_rate": 7.975e-06,
+      "loss": 1.3344,
+      "step": 1681
+    },
+    {
+      "epoch": 2.5131437642993886,
+      "grad_norm": 0.21937142312526703,
+      "learning_rate": 7.95e-06,
+      "loss": 1.3531,
+      "step": 1682
+    },
+    {
+      "epoch": 2.514637904468413,
+      "grad_norm": 0.25020983815193176,
+      "learning_rate": 7.925000000000001e-06,
+      "loss": 1.3749,
+      "step": 1683
+    },
+    {
+      "epoch": 2.5161320446374376,
+      "grad_norm": 0.19704294204711914,
+      "learning_rate": 7.9e-06,
+      "loss": 1.3779,
+      "step": 1684
+    },
+    {
+      "epoch": 2.517626184806462,
+      "grad_norm": 0.14473195374011993,
+      "learning_rate": 7.875e-06,
+      "loss": 1.2173,
+      "step": 1685
+    },
+    {
+      "epoch": 2.5191203249754865,
+      "grad_norm": 0.17848993837833405,
+      "learning_rate": 7.850000000000001e-06,
+      "loss": 1.3379,
+      "step": 1686
+    },
+    {
+      "epoch": 2.5206144651445115,
+      "grad_norm": 0.23943091928958893,
+      "learning_rate": 7.825e-06,
+      "loss": 1.2539,
+      "step": 1687
+    },
+    {
+      "epoch": 2.522108605313536,
+      "grad_norm": 0.17616920173168182,
+      "learning_rate": 7.8e-06,
+      "loss": 1.1976,
+      "step": 1688
+    },
+    {
+      "epoch": 2.5236027454825605,
+      "grad_norm": 0.3900967240333557,
+      "learning_rate": 7.775000000000001e-06,
+      "loss": 1.1891,
+      "step": 1689
+    },
+    {
+      "epoch": 2.5250968856515854,
+      "grad_norm": 0.22222371399402618,
+      "learning_rate": 7.75e-06,
+      "loss": 1.3877,
+      "step": 1690
+    },
+    {
+      "epoch": 2.52659102582061,
+      "grad_norm": 0.2556477189064026,
+      "learning_rate": 7.725e-06,
+      "loss": 1.4977,
+      "step": 1691
+    },
+    {
+      "epoch": 2.5280851659896344,
+      "grad_norm": 0.1834731101989746,
+      "learning_rate": 7.7e-06,
+      "loss": 1.2776,
+      "step": 1692
+    },
+    {
+      "epoch": 2.529579306158659,
+      "grad_norm": 0.2504706084728241,
+      "learning_rate": 7.675e-06,
+      "loss": 1.2564,
+      "step": 1693
+    },
+    {
+      "epoch": 2.5310734463276834,
+      "grad_norm": 0.2430802583694458,
+      "learning_rate": 7.65e-06,
+      "loss": 1.2452,
+      "step": 1694
+    },
+    {
+      "epoch": 2.5325675864967083,
+      "grad_norm": 0.1919189840555191,
+      "learning_rate": 7.625e-06,
+      "loss": 1.3058,
+      "step": 1695
+    },
+    {
+      "epoch": 2.534061726665733,
+      "grad_norm": 0.1974821835756302,
+      "learning_rate": 7.6e-06,
+      "loss": 1.3763,
+      "step": 1696
+    },
+    {
+      "epoch": 2.5355558668347573,
+      "grad_norm": 0.1689937561750412,
+      "learning_rate": 7.575e-06,
+      "loss": 1.2053,
+      "step": 1697
+    },
+    {
+      "epoch": 2.5370500070037822,
+      "grad_norm": 0.308765172958374,
+      "learning_rate": 7.55e-06,
+      "loss": 1.3868,
+      "step": 1698
+    },
+    {
+      "epoch": 2.5385441471728067,
+      "grad_norm": 0.19889512658119202,
+      "learning_rate": 7.525e-06,
+      "loss": 1.3385,
+      "step": 1699
+    },
+    {
+      "epoch": 2.5400382873418312,
+      "grad_norm": 0.2441468983888626,
+      "learning_rate": 7.5e-06,
+      "loss": 1.3314,
+      "step": 1700
+    },
+    {
+      "epoch": 2.5415324275108557,
+      "grad_norm": 0.2642837464809418,
+      "learning_rate": 7.4750000000000004e-06,
+      "loss": 1.3409,
+      "step": 1701
+    },
+    {
+      "epoch": 2.5430265676798802,
+      "grad_norm": 0.17711983621120453,
+      "learning_rate": 7.45e-06,
+      "loss": 1.2269,
+      "step": 1702
+    },
+    {
+      "epoch": 2.544520707848905,
+      "grad_norm": 0.21140949428081512,
+      "learning_rate": 7.425e-06,
+      "loss": 1.2436,
+      "step": 1703
+    },
+    {
+      "epoch": 2.5460148480179297,
+      "grad_norm": 0.17483963072299957,
+      "learning_rate": 7.4e-06,
+      "loss": 1.306,
+      "step": 1704
+    },
+    {
+      "epoch": 2.547508988186954,
+      "grad_norm": 0.2772678732872009,
+      "learning_rate": 7.375e-06,
+      "loss": 1.3276,
+      "step": 1705
+    },
+    {
+      "epoch": 2.549003128355979,
+      "grad_norm": 0.18797051906585693,
+      "learning_rate": 7.35e-06,
+      "loss": 1.0811,
+      "step": 1706
+    },
+    {
+      "epoch": 2.5504972685250036,
+      "grad_norm": 0.4826465845108032,
+      "learning_rate": 7.325e-06,
+      "loss": 1.3009,
+      "step": 1707
+    },
+    {
+      "epoch": 2.551991408694028,
+      "grad_norm": 0.16395214200019836,
+      "learning_rate": 7.2999999999999996e-06,
+      "loss": 1.1172,
+      "step": 1708
+    },
+    {
+      "epoch": 2.5534855488630526,
+      "grad_norm": 0.20583327114582062,
+      "learning_rate": 7.275e-06,
+      "loss": 1.2985,
+      "step": 1709
+    },
+    {
+      "epoch": 2.554979689032077,
+      "grad_norm": 0.24862010776996613,
+      "learning_rate": 7.25e-06,
+      "loss": 1.4367,
+      "step": 1710
+    },
+    {
+      "epoch": 2.556473829201102,
+      "grad_norm": 0.2522416412830353,
+      "learning_rate": 7.2249999999999994e-06,
+      "loss": 1.0932,
+      "step": 1711
+    },
+    {
+      "epoch": 2.5579679693701265,
+      "grad_norm": 0.2193172425031662,
+      "learning_rate": 7.2e-06,
+      "loss": 1.4272,
+      "step": 1712
+    },
+    {
+      "epoch": 2.559462109539151,
+      "grad_norm": 0.23507429659366608,
+      "learning_rate": 7.175e-06,
+      "loss": 1.2178,
+      "step": 1713
+    },
+    {
+      "epoch": 2.560956249708176,
+      "grad_norm": 0.2719856798648834,
+      "learning_rate": 7.15e-06,
+      "loss": 1.4399,
+      "step": 1714
+    },
+    {
+      "epoch": 2.5624503898772004,
+      "grad_norm": 0.2828925848007202,
+      "learning_rate": 7.1249999999999995e-06,
+      "loss": 1.4121,
+      "step": 1715
+    },
+    {
+      "epoch": 2.563944530046225,
+      "grad_norm": 0.2640264332294464,
+      "learning_rate": 7.1e-06,
+      "loss": 1.3727,
+      "step": 1716
+    },
+    {
+      "epoch": 2.5654386702152494,
+      "grad_norm": 0.20125895738601685,
+      "learning_rate": 7.075e-06,
+      "loss": 1.3127,
+      "step": 1717
+    },
+    {
+      "epoch": 2.566932810384274,
+      "grad_norm": 0.2524392008781433,
+      "learning_rate": 7.049999999999999e-06,
+      "loss": 1.3005,
+      "step": 1718
+    },
+    {
+      "epoch": 2.568426950553299,
+      "grad_norm": 0.24239836633205414,
+      "learning_rate": 7.025000000000001e-06,
+      "loss": 1.4189,
+      "step": 1719
+    },
+    {
+      "epoch": 2.5699210907223233,
+      "grad_norm": 0.17018896341323853,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.2798,
+      "step": 1720
+    },
+    {
+      "epoch": 2.571415230891348,
+      "grad_norm": 0.1992681324481964,
+      "learning_rate": 6.975000000000001e-06,
+      "loss": 1.3166,
+      "step": 1721
+    },
+    {
+      "epoch": 2.5729093710603728,
+      "grad_norm": 0.19908703863620758,
+      "learning_rate": 6.950000000000001e-06,
+      "loss": 1.4153,
+      "step": 1722
+    },
+    {
+      "epoch": 2.5744035112293973,
+      "grad_norm": 0.26776745915412903,
+      "learning_rate": 6.925000000000001e-06,
+      "loss": 1.2286,
+      "step": 1723
+    },
+    {
+      "epoch": 2.5758976513984218,
+      "grad_norm": 0.2907460927963257,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 1.559,
+      "step": 1724
+    },
+    {
+      "epoch": 2.5773917915674467,
+      "grad_norm": 0.30556491017341614,
+      "learning_rate": 6.875000000000001e-06,
+      "loss": 1.3937,
+      "step": 1725
+    },
+    {
+      "epoch": 2.578885931736471,
+      "grad_norm": 0.2142094224691391,
+      "learning_rate": 6.8500000000000005e-06,
+      "loss": 1.3817,
+      "step": 1726
+    },
+    {
+      "epoch": 2.5803800719054957,
+      "grad_norm": 0.230880469083786,
+      "learning_rate": 6.825000000000001e-06,
+      "loss": 1.2456,
+      "step": 1727
+    },
+    {
+      "epoch": 2.58187421207452,
+      "grad_norm": 0.1611093133687973,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.2031,
+      "step": 1728
+    },
+    {
+      "epoch": 2.5833683522435447,
+      "grad_norm": 0.2961817979812622,
+      "learning_rate": 6.775000000000001e-06,
+      "loss": 1.2962,
+      "step": 1729
+    },
+    {
+      "epoch": 2.5848624924125696,
+      "grad_norm": 0.25123047828674316,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 1.237,
+      "step": 1730
+    },
+    {
+      "epoch": 2.586356632581594,
+      "grad_norm": 0.19058406352996826,
+      "learning_rate": 6.725000000000001e-06,
+      "loss": 1.3968,
+      "step": 1731
+    },
+    {
+      "epoch": 2.5878507727506186,
+      "grad_norm": 0.28494787216186523,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 1.519,
+      "step": 1732
+    },
+    {
+      "epoch": 2.5893449129196435,
+      "grad_norm": 0.26594290137290955,
+      "learning_rate": 6.6750000000000005e-06,
+      "loss": 1.2813,
+      "step": 1733
+    },
+    {
+      "epoch": 2.590839053088668,
+      "grad_norm": 0.362155556678772,
+      "learning_rate": 6.650000000000001e-06,
+      "loss": 1.3084,
+      "step": 1734
+    },
+    {
+      "epoch": 2.5923331932576925,
+      "grad_norm": 0.2774062156677246,
+      "learning_rate": 6.625000000000001e-06,
+      "loss": 1.3572,
+      "step": 1735
+    },
+    {
+      "epoch": 2.593827333426717,
+      "grad_norm": 0.26589712500572205,
+      "learning_rate": 6.6e-06,
+      "loss": 1.3727,
+      "step": 1736
+    },
+    {
+      "epoch": 2.5953214735957415,
+      "grad_norm": 0.1808343082666397,
+      "learning_rate": 6.5750000000000006e-06,
+      "loss": 1.2308,
+      "step": 1737
+    },
+    {
+      "epoch": 2.5968156137647664,
+      "grad_norm": 0.2863466739654541,
+      "learning_rate": 6.550000000000001e-06,
+      "loss": 1.2566,
+      "step": 1738
+    },
+    {
+      "epoch": 2.598309753933791,
+      "grad_norm": 0.39508727192878723,
+      "learning_rate": 6.525e-06,
+      "loss": 1.2495,
+      "step": 1739
+    },
+    {
+      "epoch": 2.5998038941028154,
+      "grad_norm": 0.20653335750102997,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.3729,
+      "step": 1740
+    },
+    {
+      "epoch": 2.6012980342718404,
+      "grad_norm": 0.2799397110939026,
+      "learning_rate": 6.475000000000001e-06,
+      "loss": 1.1533,
+      "step": 1741
+    },
+    {
+      "epoch": 2.602792174440865,
+      "grad_norm": 0.2511701285839081,
+      "learning_rate": 6.45e-06,
+      "loss": 1.3166,
+      "step": 1742
+    },
+    {
+      "epoch": 2.6042863146098894,
+      "grad_norm": 0.42425861954689026,
+      "learning_rate": 6.425e-06,
+      "loss": 1.2655,
+      "step": 1743
+    },
+    {
+      "epoch": 2.605780454778914,
+      "grad_norm": 0.16690030694007874,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.1529,
+      "step": 1744
+    },
+    {
+      "epoch": 2.6072745949479383,
+      "grad_norm": 0.29977497458457947,
+      "learning_rate": 6.375000000000001e-06,
+      "loss": 1.2709,
+      "step": 1745
+    },
+    {
+      "epoch": 2.6087687351169633,
+      "grad_norm": 0.17054329812526703,
+      "learning_rate": 6.35e-06,
+      "loss": 1.2346,
+      "step": 1746
+    },
+    {
+      "epoch": 2.6102628752859878,
+      "grad_norm": 0.18217666447162628,
+      "learning_rate": 6.3250000000000004e-06,
+      "loss": 1.363,
+      "step": 1747
+    },
+    {
+      "epoch": 2.6117570154550123,
+      "grad_norm": 0.18411266803741455,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 1.2045,
+      "step": 1748
+    },
+    {
+      "epoch": 2.613251155624037,
+      "grad_norm": 0.22760531306266785,
+      "learning_rate": 6.275e-06,
+      "loss": 1.3307,
+      "step": 1749
+    },
+    {
+      "epoch": 2.6147452957930617,
+      "grad_norm": 0.3128303587436676,
+      "learning_rate": 6.25e-06,
+      "loss": 1.4054,
+      "step": 1750
+    },
+    {
+      "epoch": 2.616239435962086,
+      "grad_norm": 0.1927785873413086,
+      "learning_rate": 6.2250000000000005e-06,
+      "loss": 1.2833,
+      "step": 1751
+    },
+    {
+      "epoch": 2.6177335761311107,
+      "grad_norm": 0.19453319907188416,
+      "learning_rate": 6.2e-06,
+      "loss": 1.1372,
+      "step": 1752
+    },
+    {
+      "epoch": 2.619227716300135,
+      "grad_norm": 0.3225673735141754,
+      "learning_rate": 6.175e-06,
+      "loss": 1.2579,
+      "step": 1753
+    },
+    {
+      "epoch": 2.62072185646916,
+      "grad_norm": 0.18228133022785187,
+      "learning_rate": 6.15e-06,
+      "loss": 1.4589,
+      "step": 1754
+    },
+    {
+      "epoch": 2.6222159966381846,
+      "grad_norm": 0.23270145058631897,
+      "learning_rate": 6.125e-06,
+      "loss": 1.403,
+      "step": 1755
+    },
+    {
+      "epoch": 2.623710136807209,
+      "grad_norm": 0.21594181656837463,
+      "learning_rate": 6.1e-06,
+      "loss": 1.4261,
+      "step": 1756
+    },
+    {
+      "epoch": 2.625204276976234,
+      "grad_norm": 0.3312102258205414,
+      "learning_rate": 6.075e-06,
+      "loss": 1.2938,
+      "step": 1757
+    },
+    {
+      "epoch": 2.6266984171452585,
+      "grad_norm": 0.2661645710468292,
+      "learning_rate": 6.0500000000000005e-06,
+      "loss": 1.411,
+      "step": 1758
+    },
+    {
+      "epoch": 2.628192557314283,
+      "grad_norm": 0.1631266474723816,
+      "learning_rate": 6.025e-06,
+      "loss": 1.3814,
+      "step": 1759
+    },
+    {
+      "epoch": 2.6296866974833075,
+      "grad_norm": 0.2169022560119629,
+      "learning_rate": 6e-06,
+      "loss": 1.4121,
+      "step": 1760
+    },
+    {
+      "epoch": 2.631180837652332,
+      "grad_norm": 0.23801642656326294,
+      "learning_rate": 5.975e-06,
+      "loss": 1.1812,
+      "step": 1761
+    },
+    {
+      "epoch": 2.632674977821357,
+      "grad_norm": 0.19962309300899506,
+      "learning_rate": 5.95e-06,
+      "loss": 1.2209,
+      "step": 1762
+    },
+    {
+      "epoch": 2.6341691179903814,
+      "grad_norm": 0.2108369767665863,
+      "learning_rate": 5.925e-06,
+      "loss": 1.315,
+      "step": 1763
+    },
+    {
+      "epoch": 2.635663258159406,
+      "grad_norm": 0.22944194078445435,
+      "learning_rate": 5.9e-06,
+      "loss": 1.2293,
+      "step": 1764
+    },
+    {
+      "epoch": 2.637157398328431,
+      "grad_norm": 0.23019960522651672,
+      "learning_rate": 5.875e-06,
+      "loss": 1.3415,
+      "step": 1765
+    },
+    {
+      "epoch": 2.6386515384974554,
+      "grad_norm": 0.18209655582904816,
+      "learning_rate": 5.850000000000001e-06,
+      "loss": 1.2341,
+      "step": 1766
+    },
+    {
+      "epoch": 2.64014567866648,
+      "grad_norm": 0.1567586064338684,
+      "learning_rate": 5.825000000000001e-06,
+      "loss": 1.121,
+      "step": 1767
+    },
+    {
+      "epoch": 2.6416398188355044,
+      "grad_norm": 0.87409907579422,
+      "learning_rate": 5.8e-06,
+      "loss": 1.2113,
+      "step": 1768
+    },
+    {
+      "epoch": 2.643133959004529,
+      "grad_norm": 0.2868337035179138,
+      "learning_rate": 5.775000000000001e-06,
+      "loss": 1.1993,
+      "step": 1769
+    },
+    {
+      "epoch": 2.644628099173554,
+      "grad_norm": 0.18613767623901367,
+      "learning_rate": 5.750000000000001e-06,
+      "loss": 1.1514,
+      "step": 1770
+    },
+    {
+      "epoch": 2.6461222393425783,
+      "grad_norm": 0.18503230810165405,
+      "learning_rate": 5.725e-06,
+      "loss": 1.1599,
+      "step": 1771
+    },
+    {
+      "epoch": 2.647616379511603,
+      "grad_norm": 0.30136963725090027,
+      "learning_rate": 5.7000000000000005e-06,
+      "loss": 1.2663,
+      "step": 1772
+    },
+    {
+      "epoch": 2.6491105196806277,
+      "grad_norm": 0.42282551527023315,
+      "learning_rate": 5.675000000000001e-06,
+      "loss": 1.3327,
+      "step": 1773
+    },
+    {
+      "epoch": 2.650604659849652,
+      "grad_norm": 0.7006960511207581,
+      "learning_rate": 5.65e-06,
+      "loss": 1.2708,
+      "step": 1774
+    },
+    {
+      "epoch": 2.6520988000186767,
+      "grad_norm": 0.15408888459205627,
+      "learning_rate": 5.625e-06,
+      "loss": 1.2682,
+      "step": 1775
+    },
+    {
+      "epoch": 2.653592940187701,
+      "grad_norm": 0.25027719140052795,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.3571,
+      "step": 1776
+    },
+    {
+      "epoch": 2.6550870803567257,
+      "grad_norm": 0.19256514310836792,
+      "learning_rate": 5.575e-06,
+      "loss": 1.2386,
+      "step": 1777
+    },
+    {
+      "epoch": 2.6565812205257506,
+      "grad_norm": 0.18717578053474426,
+      "learning_rate": 5.55e-06,
+      "loss": 1.1667,
+      "step": 1778
+    },
+    {
+      "epoch": 2.658075360694775,
+      "grad_norm": 0.16257885098457336,
+      "learning_rate": 5.5250000000000005e-06,
+      "loss": 1.2965,
+      "step": 1779
+    },
+    {
+      "epoch": 2.6595695008637996,
+      "grad_norm": 0.2796028256416321,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.3928,
+      "step": 1780
+    },
+    {
+      "epoch": 2.6610636410328246,
+      "grad_norm": 0.2988707423210144,
+      "learning_rate": 5.475e-06,
+      "loss": 1.4746,
+      "step": 1781
+    },
+    {
+      "epoch": 2.662557781201849,
+      "grad_norm": 0.2644762396812439,
+      "learning_rate": 5.45e-06,
+      "loss": 1.3648,
+      "step": 1782
+    },
+    {
+      "epoch": 2.6640519213708735,
+      "grad_norm": 0.23171842098236084,
+      "learning_rate": 5.4250000000000006e-06,
+      "loss": 1.266,
+      "step": 1783
+    },
+    {
+      "epoch": 2.6655460615398985,
+      "grad_norm": 0.21569930016994476,
+      "learning_rate": 5.4e-06,
+      "loss": 1.344,
+      "step": 1784
+    },
+    {
+      "epoch": 2.667040201708923,
+      "grad_norm": 0.31907474994659424,
+      "learning_rate": 5.375e-06,
+      "loss": 1.5272,
+      "step": 1785
+    },
+    {
+      "epoch": 2.6685343418779475,
+      "grad_norm": 0.18725049495697021,
+      "learning_rate": 5.3500000000000004e-06,
+      "loss": 1.4265,
+      "step": 1786
+    },
+    {
+      "epoch": 2.670028482046972,
+      "grad_norm": 0.22397437691688538,
+      "learning_rate": 5.325e-06,
+      "loss": 1.1694,
+      "step": 1787
+    },
+    {
+      "epoch": 2.6715226222159965,
+      "grad_norm": 0.2414587140083313,
+      "learning_rate": 5.3e-06,
+      "loss": 1.1475,
+      "step": 1788
+    },
+    {
+      "epoch": 2.6730167623850214,
+      "grad_norm": 0.307595819234848,
+      "learning_rate": 5.275e-06,
+      "loss": 1.36,
+      "step": 1789
+    },
+    {
+      "epoch": 2.674510902554046,
+      "grad_norm": 0.20186826586723328,
+      "learning_rate": 5.25e-06,
+      "loss": 1.35,
+      "step": 1790
+    },
+    {
+      "epoch": 2.6760050427230704,
+      "grad_norm": 0.20282383263111115,
+      "learning_rate": 5.225e-06,
+      "loss": 1.2654,
+      "step": 1791
+    },
+    {
+      "epoch": 2.6774991828920953,
+      "grad_norm": 0.21675416827201843,
+      "learning_rate": 5.2e-06,
+      "loss": 1.2554,
+      "step": 1792
+    },
+    {
+      "epoch": 2.67899332306112,
+      "grad_norm": 0.47374227643013,
+      "learning_rate": 5.175e-06,
+      "loss": 1.4186,
+      "step": 1793
+    },
+    {
+      "epoch": 2.6804874632301443,
+      "grad_norm": 0.17790532112121582,
+      "learning_rate": 5.15e-06,
+      "loss": 1.277,
+      "step": 1794
+    },
+    {
+      "epoch": 2.681981603399169,
+      "grad_norm": 0.19837185740470886,
+      "learning_rate": 5.125e-06,
+      "loss": 1.3053,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6834757435681933,
+      "grad_norm": 0.174654021859169,
+      "learning_rate": 5.1e-06,
+      "loss": 1.2047,
+      "step": 1796
+    },
+    {
+      "epoch": 2.6849698837372182,
+      "grad_norm": 0.6168277263641357,
+      "learning_rate": 5.0750000000000005e-06,
+      "loss": 1.3662,
+      "step": 1797
+    },
+    {
+      "epoch": 2.6864640239062427,
+      "grad_norm": 0.33519673347473145,
+      "learning_rate": 5.050000000000001e-06,
+      "loss": 1.515,
+      "step": 1798
+    },
+    {
+      "epoch": 2.687958164075267,
+      "grad_norm": 0.2172117829322815,
+      "learning_rate": 5.025e-06,
+      "loss": 1.2729,
+      "step": 1799
+    },
+    {
+      "epoch": 2.689452304244292,
+      "grad_norm": 0.2625578045845032,
+      "learning_rate": 5e-06,
+      "loss": 1.3624,
+      "step": 1800
+    },
+    {
+      "epoch": 2.6909464444133167,
+      "grad_norm": 0.26138144731521606,
+      "learning_rate": 4.975000000000001e-06,
+      "loss": 1.2029,
+      "step": 1801
+    },
+    {
+      "epoch": 2.692440584582341,
+      "grad_norm": 0.25910723209381104,
+      "learning_rate": 4.950000000000001e-06,
+      "loss": 1.4087,
+      "step": 1802
+    },
+    {
+      "epoch": 2.6939347247513656,
+      "grad_norm": 0.3034740090370178,
+      "learning_rate": 4.925e-06,
+      "loss": 1.162,
+      "step": 1803
+    },
+    {
+      "epoch": 2.69542886492039,
+      "grad_norm": 0.21220143139362335,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.2157,
+      "step": 1804
+    },
+    {
+      "epoch": 2.696923005089415,
+      "grad_norm": 0.27650538086891174,
+      "learning_rate": 4.875000000000001e-06,
+      "loss": 1.3509,
+      "step": 1805
+    },
+    {
+      "epoch": 2.6984171452584396,
+      "grad_norm": 0.22808341681957245,
+      "learning_rate": 4.85e-06,
+      "loss": 1.2005,
+      "step": 1806
+    },
+    {
+      "epoch": 2.699911285427464,
+      "grad_norm": 0.18333181738853455,
+      "learning_rate": 4.825e-06,
+      "loss": 1.2351,
+      "step": 1807
+    },
+    {
+      "epoch": 2.701405425596489,
+      "grad_norm": 0.5237784385681152,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.3003,
+      "step": 1808
+    },
+    {
+      "epoch": 2.7028995657655135,
+      "grad_norm": 0.4465050995349884,
+      "learning_rate": 4.775e-06,
+      "loss": 1.474,
+      "step": 1809
+    },
+    {
+      "epoch": 2.704393705934538,
+      "grad_norm": 0.4441722333431244,
+      "learning_rate": 4.75e-06,
+      "loss": 1.4661,
+      "step": 1810
+    },
+    {
+      "epoch": 2.7058878461035625,
+      "grad_norm": 0.3523397743701935,
+      "learning_rate": 4.7250000000000005e-06,
+      "loss": 1.2954,
+      "step": 1811
+    },
+    {
+      "epoch": 2.707381986272587,
+      "grad_norm": 0.2340545803308487,
+      "learning_rate": 4.7e-06,
+      "loss": 1.1822,
+      "step": 1812
+    },
+    {
+      "epoch": 2.708876126441612,
+      "grad_norm": 0.34939736127853394,
+      "learning_rate": 4.675e-06,
+      "loss": 1.2706,
+      "step": 1813
+    },
+    {
+      "epoch": 2.7103702666106364,
+      "grad_norm": 0.24331608414649963,
+      "learning_rate": 4.65e-06,
+      "loss": 1.3334,
+      "step": 1814
+    },
+    {
+      "epoch": 2.711864406779661,
+      "grad_norm": 0.1805030107498169,
+      "learning_rate": 4.625e-06,
+      "loss": 1.3894,
+      "step": 1815
+    },
+    {
+      "epoch": 2.713358546948686,
+      "grad_norm": 0.6400303244590759,
+      "learning_rate": 4.6e-06,
+      "loss": 1.4153,
+      "step": 1816
+    },
+    {
+      "epoch": 2.7148526871177103,
+      "grad_norm": 0.2347690910100937,
+      "learning_rate": 4.575e-06,
+      "loss": 1.4021,
+      "step": 1817
+    },
+    {
+      "epoch": 2.716346827286735,
+      "grad_norm": 2.5086920261383057,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 1.4322,
+      "step": 1818
+    },
+    {
+      "epoch": 2.7178409674557593,
+      "grad_norm": 0.18511001765727997,
+      "learning_rate": 4.525e-06,
+      "loss": 1.2838,
+      "step": 1819
+    },
+    {
+      "epoch": 2.719335107624784,
+      "grad_norm": 0.2383240908384323,
+      "learning_rate": 4.5e-06,
+      "loss": 1.0991,
+      "step": 1820
+    },
+    {
+      "epoch": 2.7208292477938087,
+      "grad_norm": 0.23585478961467743,
+      "learning_rate": 4.475e-06,
+      "loss": 1.2393,
+      "step": 1821
+    },
+    {
+      "epoch": 2.7223233879628332,
+      "grad_norm": 0.2568042278289795,
+      "learning_rate": 4.45e-06,
+      "loss": 1.1486,
+      "step": 1822
+    },
+    {
+      "epoch": 2.7238175281318577,
+      "grad_norm": 0.20862841606140137,
+      "learning_rate": 4.425e-06,
+      "loss": 1.1885,
+      "step": 1823
+    },
+    {
+      "epoch": 2.7253116683008827,
+      "grad_norm": 0.19219589233398438,
+      "learning_rate": 4.4e-06,
+      "loss": 1.4042,
+      "step": 1824
+    },
+    {
+      "epoch": 2.726805808469907,
+      "grad_norm": 0.3805523216724396,
+      "learning_rate": 4.375e-06,
+      "loss": 1.2164,
+      "step": 1825
+    },
+    {
+      "epoch": 2.7282999486389317,
+      "grad_norm": 0.49763739109039307,
+      "learning_rate": 4.35e-06,
+      "loss": 1.3467,
+      "step": 1826
+    },
+    {
+      "epoch": 2.729794088807956,
+      "grad_norm": 0.17709727585315704,
+      "learning_rate": 4.325e-06,
+      "loss": 1.1515,
+      "step": 1827
+    },
+    {
+      "epoch": 2.7312882289769806,
+      "grad_norm": 0.20847739279270172,
+      "learning_rate": 4.2999999999999995e-06,
+      "loss": 1.1841,
+      "step": 1828
+    },
+    {
+      "epoch": 2.7327823691460056,
+      "grad_norm": 0.21395619213581085,
+      "learning_rate": 4.2750000000000006e-06,
+      "loss": 1.3783,
+      "step": 1829
+    },
+    {
+      "epoch": 2.73427650931503,
+      "grad_norm": 0.4332375228404999,
+      "learning_rate": 4.250000000000001e-06,
+      "loss": 1.1825,
+      "step": 1830
+    },
+    {
+      "epoch": 2.7357706494840546,
+      "grad_norm": 0.8193243145942688,
+      "learning_rate": 4.225e-06,
+      "loss": 1.5271,
+      "step": 1831
+    },
+    {
+      "epoch": 2.7372647896530795,
+      "grad_norm": 0.19088123738765717,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.3111,
+      "step": 1832
+    },
+    {
+      "epoch": 2.738758929822104,
+      "grad_norm": 0.4072718620300293,
+      "learning_rate": 4.175000000000001e-06,
+      "loss": 1.2883,
+      "step": 1833
+    },
+    {
+      "epoch": 2.7402530699911285,
+      "grad_norm": 0.2014656513929367,
+      "learning_rate": 4.15e-06,
+      "loss": 1.2469,
+      "step": 1834
+    },
+    {
+      "epoch": 2.741747210160153,
+      "grad_norm": 0.2613576054573059,
+      "learning_rate": 4.125e-06,
+      "loss": 1.2822,
+      "step": 1835
+    },
+    {
+      "epoch": 2.7432413503291775,
+      "grad_norm": 0.21430745720863342,
+      "learning_rate": 4.1000000000000006e-06,
+      "loss": 1.2495,
+      "step": 1836
+    },
+    {
+      "epoch": 2.7447354904982024,
+      "grad_norm": 0.18431149423122406,
+      "learning_rate": 4.075e-06,
+      "loss": 1.3818,
+      "step": 1837
+    },
+    {
+      "epoch": 2.746229630667227,
+      "grad_norm": 0.2120877504348755,
+      "learning_rate": 4.05e-06,
+      "loss": 1.2272,
+      "step": 1838
+    },
+    {
+      "epoch": 2.7477237708362514,
+      "grad_norm": 2.143777370452881,
+      "learning_rate": 4.0250000000000004e-06,
+      "loss": 1.309,
+      "step": 1839
+    },
+    {
+      "epoch": 2.7492179110052763,
+      "grad_norm": 0.2145431488752365,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3135,
+      "step": 1840
+    },
+    {
+      "epoch": 2.750712051174301,
+      "grad_norm": 0.2832334339618683,
+      "learning_rate": 3.975e-06,
+      "loss": 1.3395,
+      "step": 1841
+    },
+    {
+      "epoch": 2.7522061913433253,
+      "grad_norm": 0.1784759759902954,
+      "learning_rate": 3.95e-06,
+      "loss": 1.3056,
+      "step": 1842
+    },
+    {
+      "epoch": 2.7537003315123503,
+      "grad_norm": 0.21450982987880707,
+      "learning_rate": 3.9250000000000005e-06,
+      "loss": 1.1001,
+      "step": 1843
+    },
+    {
+      "epoch": 2.7551944716813748,
+      "grad_norm": 0.2597975730895996,
+      "learning_rate": 3.9e-06,
+      "loss": 1.3779,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7566886118503993,
+      "grad_norm": 0.2315620481967926,
+      "learning_rate": 3.875e-06,
+      "loss": 1.3914,
+      "step": 1845
+    },
+    {
+      "epoch": 2.7581827520194238,
+      "grad_norm": 0.25470679998397827,
+      "learning_rate": 3.85e-06,
+      "loss": 1.2633,
+      "step": 1846
+    },
+    {
+      "epoch": 2.7596768921884482,
+      "grad_norm": 0.2289915233850479,
+      "learning_rate": 3.825e-06,
+      "loss": 1.2158,
+      "step": 1847
+    },
+    {
+      "epoch": 2.761171032357473,
+      "grad_norm": 0.6363893151283264,
+      "learning_rate": 3.8e-06,
+      "loss": 1.4446,
+      "step": 1848
+    },
+    {
+      "epoch": 2.7626651725264977,
+      "grad_norm": 0.19268900156021118,
+      "learning_rate": 3.775e-06,
+      "loss": 1.2836,
+      "step": 1849
+    },
+    {
+      "epoch": 2.764159312695522,
+      "grad_norm": 0.515834629535675,
+      "learning_rate": 3.75e-06,
+      "loss": 1.3646,
+      "step": 1850
+    },
+    {
+      "epoch": 2.765653452864547,
+      "grad_norm": 0.2783832848072052,
+      "learning_rate": 3.725e-06,
+      "loss": 1.3876,
+      "step": 1851
+    },
+    {
+      "epoch": 2.7671475930335716,
+      "grad_norm": 0.21633508801460266,
+      "learning_rate": 3.7e-06,
+      "loss": 1.1456,
+      "step": 1852
+    },
+    {
+      "epoch": 2.768641733202596,
+      "grad_norm": 0.22185078263282776,
+      "learning_rate": 3.675e-06,
+      "loss": 1.5259,
+      "step": 1853
+    },
+    {
+      "epoch": 2.7701358733716206,
+      "grad_norm": 0.29004672169685364,
+      "learning_rate": 3.6499999999999998e-06,
+      "loss": 1.3148,
+      "step": 1854
+    },
+    {
+      "epoch": 2.771630013540645,
+      "grad_norm": 0.2006434053182602,
+      "learning_rate": 3.625e-06,
+      "loss": 1.1994,
+      "step": 1855
+    },
+    {
+      "epoch": 2.77312415370967,
+      "grad_norm": 0.21816261112689972,
+      "learning_rate": 3.6e-06,
+      "loss": 1.3848,
+      "step": 1856
+    },
+    {
+      "epoch": 2.7746182938786945,
+      "grad_norm": 0.21714898943901062,
+      "learning_rate": 3.575e-06,
+      "loss": 1.2319,
+      "step": 1857
+    },
+    {
+      "epoch": 2.776112434047719,
+      "grad_norm": 0.21400469541549683,
+      "learning_rate": 3.55e-06,
+      "loss": 1.4311,
+      "step": 1858
+    },
+    {
+      "epoch": 2.777606574216744,
+      "grad_norm": 0.20019961893558502,
+      "learning_rate": 3.5249999999999997e-06,
+      "loss": 1.3558,
+      "step": 1859
+    },
+    {
+      "epoch": 2.7791007143857684,
+      "grad_norm": 0.24723859131336212,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 1.1142,
+      "step": 1860
+    },
+    {
+      "epoch": 2.780594854554793,
+      "grad_norm": 0.22617855668067932,
+      "learning_rate": 3.4750000000000006e-06,
+      "loss": 1.3948,
+      "step": 1861
+    },
+    {
+      "epoch": 2.7820889947238174,
+      "grad_norm": 0.324815958738327,
+      "learning_rate": 3.4500000000000004e-06,
+      "loss": 1.2817,
+      "step": 1862
+    },
+    {
+      "epoch": 2.783583134892842,
+      "grad_norm": 0.19511297345161438,
+      "learning_rate": 3.4250000000000002e-06,
+      "loss": 1.5176,
+      "step": 1863
+    },
+    {
+      "epoch": 2.785077275061867,
+      "grad_norm": 0.8152601718902588,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.1762,
+      "step": 1864
+    },
+    {
+      "epoch": 2.7865714152308914,
+      "grad_norm": 0.23388169705867767,
+      "learning_rate": 3.3750000000000003e-06,
+      "loss": 1.1857,
+      "step": 1865
+    },
+    {
+      "epoch": 2.788065555399916,
+      "grad_norm": 0.28998321294784546,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 1.0647,
+      "step": 1866
+    },
+    {
+      "epoch": 2.789559695568941,
+      "grad_norm": 0.2541603446006775,
+      "learning_rate": 3.3250000000000004e-06,
+      "loss": 1.3563,
+      "step": 1867
+    },
+    {
+      "epoch": 2.7910538357379653,
+      "grad_norm": 0.25344282388687134,
+      "learning_rate": 3.3e-06,
+      "loss": 1.4685,
+      "step": 1868
+    },
+    {
+      "epoch": 2.7925479759069898,
+      "grad_norm": 0.22146444022655487,
+      "learning_rate": 3.2750000000000004e-06,
+      "loss": 1.3198,
+      "step": 1869
+    },
+    {
+      "epoch": 2.7940421160760143,
+      "grad_norm": 0.21067120134830475,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 1.3077,
+      "step": 1870
+    },
+    {
+      "epoch": 2.7955362562450388,
+      "grad_norm": 0.45738738775253296,
+      "learning_rate": 3.225e-06,
+      "loss": 1.3165,
+      "step": 1871
+    },
+    {
+      "epoch": 2.7970303964140637,
+      "grad_norm": 0.1831294596195221,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.208,
+      "step": 1872
+    },
+    {
+      "epoch": 2.798524536583088,
+      "grad_norm": 0.2563336491584778,
+      "learning_rate": 3.175e-06,
+      "loss": 1.4487,
+      "step": 1873
+    },
+    {
+      "epoch": 2.8000186767521127,
+      "grad_norm": 0.15740127861499786,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 1.2381,
+      "step": 1874
+    },
+    {
+      "epoch": 2.8015128169211376,
+      "grad_norm": 0.2575199007987976,
+      "learning_rate": 3.125e-06,
+      "loss": 1.2935,
+      "step": 1875
+    },
+    {
+      "epoch": 2.803006957090162,
+      "grad_norm": 0.20497269928455353,
+      "learning_rate": 3.1e-06,
+      "loss": 1.5041,
+      "step": 1876
+    },
+    {
+      "epoch": 2.8045010972591866,
+      "grad_norm": 0.4272780418395996,
+      "learning_rate": 3.075e-06,
+      "loss": 1.3039,
+      "step": 1877
+    },
+    {
+      "epoch": 2.805995237428211,
+      "grad_norm": 0.26943546533584595,
+      "learning_rate": 3.05e-06,
+      "loss": 1.1747,
+      "step": 1878
+    },
+    {
+      "epoch": 2.8074893775972356,
+      "grad_norm": 0.3755415678024292,
+      "learning_rate": 3.0250000000000003e-06,
+      "loss": 1.394,
+      "step": 1879
+    },
+    {
+      "epoch": 2.8089835177662605,
+      "grad_norm": 0.2485462874174118,
+      "learning_rate": 3e-06,
+      "loss": 1.317,
+      "step": 1880
+    },
+    {
+      "epoch": 2.810477657935285,
+      "grad_norm": 0.1786644607782364,
+      "learning_rate": 2.975e-06,
+      "loss": 1.2546,
+      "step": 1881
+    },
+    {
+      "epoch": 2.8119717981043095,
+      "grad_norm": 1.068226933479309,
+      "learning_rate": 2.95e-06,
+      "loss": 1.3132,
+      "step": 1882
+    },
+    {
+      "epoch": 2.8134659382733345,
+      "grad_norm": 0.19659435749053955,
+      "learning_rate": 2.9250000000000004e-06,
+      "loss": 1.2877,
+      "step": 1883
+    },
+    {
+      "epoch": 2.814960078442359,
+      "grad_norm": 0.2480842024087906,
+      "learning_rate": 2.9e-06,
+      "loss": 1.337,
+      "step": 1884
+    },
+    {
+      "epoch": 2.8164542186113835,
+      "grad_norm": 0.3635180592536926,
+      "learning_rate": 2.8750000000000004e-06,
+      "loss": 1.2722,
+      "step": 1885
+    },
+    {
+      "epoch": 2.817948358780408,
+      "grad_norm": 0.238739475607872,
+      "learning_rate": 2.8500000000000002e-06,
+      "loss": 1.4595,
+      "step": 1886
+    },
+    {
+      "epoch": 2.8194424989494324,
+      "grad_norm": 0.2953251600265503,
+      "learning_rate": 2.825e-06,
+      "loss": 1.4326,
+      "step": 1887
+    },
+    {
+      "epoch": 2.8209366391184574,
+      "grad_norm": 0.17587164044380188,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.3387,
+      "step": 1888
+    },
+    {
+      "epoch": 2.822430779287482,
+      "grad_norm": 0.2200200855731964,
+      "learning_rate": 2.775e-06,
+      "loss": 1.2648,
+      "step": 1889
+    },
+    {
+      "epoch": 2.8239249194565064,
+      "grad_norm": 0.2534538507461548,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.3252,
+      "step": 1890
+    },
+    {
+      "epoch": 2.8254190596255313,
+      "grad_norm": 0.23841866850852966,
+      "learning_rate": 2.725e-06,
+      "loss": 1.2432,
+      "step": 1891
+    },
+    {
+      "epoch": 2.826913199794556,
+      "grad_norm": 0.23968037962913513,
+      "learning_rate": 2.7e-06,
+      "loss": 1.1769,
+      "step": 1892
+    },
+    {
+      "epoch": 2.8284073399635803,
+      "grad_norm": 0.3135208189487457,
+      "learning_rate": 2.6750000000000002e-06,
+      "loss": 1.1362,
+      "step": 1893
+    },
+    {
+      "epoch": 2.829901480132605,
+      "grad_norm": 0.1951269954442978,
+      "learning_rate": 2.65e-06,
+      "loss": 1.3334,
+      "step": 1894
+    },
+    {
+      "epoch": 2.8313956203016293,
+      "grad_norm": 0.21999233961105347,
+      "learning_rate": 2.625e-06,
+      "loss": 1.348,
+      "step": 1895
+    },
+    {
+      "epoch": 2.832889760470654,
+      "grad_norm": 0.3923155963420868,
+      "learning_rate": 2.6e-06,
+      "loss": 1.2762,
+      "step": 1896
+    },
+    {
+      "epoch": 2.8343839006396787,
+      "grad_norm": 0.25877097249031067,
+      "learning_rate": 2.575e-06,
+      "loss": 1.2909,
+      "step": 1897
+    },
+    {
+      "epoch": 2.835878040808703,
+      "grad_norm": 0.2079850733280182,
+      "learning_rate": 2.55e-06,
+      "loss": 1.2758,
+      "step": 1898
+    },
+    {
+      "epoch": 2.837372180977728,
+      "grad_norm": 0.1782015711069107,
+      "learning_rate": 2.5250000000000004e-06,
+      "loss": 1.4591,
+      "step": 1899
+    },
+    {
+      "epoch": 2.8388663211467526,
+      "grad_norm": 0.21035560965538025,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0855,
+      "step": 1900
+    },
+    {
+      "epoch": 2.840360461315777,
+      "grad_norm": 0.22053273022174835,
+      "learning_rate": 2.4750000000000004e-06,
+      "loss": 1.2851,
+      "step": 1901
+    },
+    {
+      "epoch": 2.841854601484802,
+      "grad_norm": 0.5123848915100098,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 1.3428,
+      "step": 1902
+    },
+    {
+      "epoch": 2.8433487416538266,
+      "grad_norm": 0.2593129873275757,
+      "learning_rate": 2.425e-06,
+      "loss": 1.4084,
+      "step": 1903
+    },
+    {
+      "epoch": 2.844842881822851,
+      "grad_norm": 0.25032708048820496,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.3046,
+      "step": 1904
+    },
+    {
+      "epoch": 2.8463370219918755,
+      "grad_norm": 0.19986706972122192,
+      "learning_rate": 2.375e-06,
+      "loss": 1.3027,
+      "step": 1905
+    },
+    {
+      "epoch": 2.8478311621609,
+      "grad_norm": 0.2648144066333771,
+      "learning_rate": 2.35e-06,
+      "loss": 1.2129,
+      "step": 1906
+    },
+    {
+      "epoch": 2.849325302329925,
+      "grad_norm": 0.19700418412685394,
+      "learning_rate": 2.325e-06,
+      "loss": 1.4243,
+      "step": 1907
+    },
+    {
+      "epoch": 2.8508194424989495,
+      "grad_norm": 0.2682986259460449,
+      "learning_rate": 2.3e-06,
+      "loss": 1.2794,
+      "step": 1908
+    },
+    {
+      "epoch": 2.852313582667974,
+      "grad_norm": 0.20905745029449463,
+      "learning_rate": 2.2750000000000002e-06,
+      "loss": 1.2348,
+      "step": 1909
+    },
+    {
+      "epoch": 2.853807722836999,
+      "grad_norm": 0.3586556315422058,
+      "learning_rate": 2.25e-06,
+      "loss": 1.3826,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8553018630060234,
+      "grad_norm": 0.1797320693731308,
+      "learning_rate": 2.225e-06,
+      "loss": 1.2428,
+      "step": 1911
+    },
+    {
+      "epoch": 2.856796003175048,
+      "grad_norm": 0.21647144854068756,
+      "learning_rate": 2.2e-06,
+      "loss": 1.4246,
+      "step": 1912
+    },
+    {
+      "epoch": 2.8582901433440724,
+      "grad_norm": 0.42772310972213745,
+      "learning_rate": 2.175e-06,
+      "loss": 1.2403,
+      "step": 1913
+    },
+    {
+      "epoch": 2.859784283513097,
+      "grad_norm": 0.18217770755290985,
+      "learning_rate": 2.1499999999999997e-06,
+      "loss": 1.2052,
+      "step": 1914
+    },
+    {
+      "epoch": 2.861278423682122,
+      "grad_norm": 0.2103743702173233,
+      "learning_rate": 2.1250000000000004e-06,
+      "loss": 1.2553,
+      "step": 1915
+    },
+    {
+      "epoch": 2.8627725638511463,
+      "grad_norm": 0.25712189078330994,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.3833,
+      "step": 1916
+    },
+    {
+      "epoch": 2.864266704020171,
+      "grad_norm": 0.31101569533348083,
+      "learning_rate": 2.075e-06,
+      "loss": 1.3011,
+      "step": 1917
+    },
+    {
+      "epoch": 2.8657608441891957,
+      "grad_norm": 0.4393167197704315,
+      "learning_rate": 2.0500000000000003e-06,
+      "loss": 1.4259,
+      "step": 1918
+    },
+    {
+      "epoch": 2.8672549843582202,
+      "grad_norm": 0.2310882806777954,
+      "learning_rate": 2.025e-06,
+      "loss": 1.2061,
+      "step": 1919
+    },
+    {
+      "epoch": 2.8687491245272447,
+      "grad_norm": 0.22133032977581024,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.255,
+      "step": 1920
+    },
+    {
+      "epoch": 2.8702432646962692,
+      "grad_norm": 0.31930020451545715,
+      "learning_rate": 1.975e-06,
+      "loss": 1.4847,
+      "step": 1921
+    },
+    {
+      "epoch": 2.8717374048652937,
+      "grad_norm": 0.1897219717502594,
+      "learning_rate": 1.95e-06,
+      "loss": 1.4619,
+      "step": 1922
+    },
+    {
+      "epoch": 2.8732315450343187,
+      "grad_norm": 0.2287951558828354,
+      "learning_rate": 1.925e-06,
+      "loss": 1.1942,
+      "step": 1923
+    },
+    {
+      "epoch": 2.874725685203343,
+      "grad_norm": 0.1839810609817505,
+      "learning_rate": 1.9e-06,
+      "loss": 1.401,
+      "step": 1924
+    },
+    {
+      "epoch": 2.8762198253723676,
+      "grad_norm": 0.2828529179096222,
+      "learning_rate": 1.875e-06,
+      "loss": 1.373,
+      "step": 1925
+    },
+    {
+      "epoch": 2.8777139655413926,
+      "grad_norm": 0.18922413885593414,
+      "learning_rate": 1.85e-06,
+      "loss": 1.322,
+      "step": 1926
+    },
+    {
+      "epoch": 2.879208105710417,
+      "grad_norm": 0.27389007806777954,
+      "learning_rate": 1.8249999999999999e-06,
+      "loss": 1.1936,
+      "step": 1927
+    },
+    {
+      "epoch": 2.8807022458794416,
+      "grad_norm": 0.16232287883758545,
+      "learning_rate": 1.8e-06,
+      "loss": 1.1672,
+      "step": 1928
+    },
+    {
+      "epoch": 2.882196386048466,
+      "grad_norm": 0.2973114550113678,
+      "learning_rate": 1.775e-06,
+      "loss": 1.2634,
+      "step": 1929
+    },
+    {
+      "epoch": 2.8836905262174906,
+      "grad_norm": 0.357994019985199,
+      "learning_rate": 1.7500000000000002e-06,
+      "loss": 1.3833,
+      "step": 1930
+    },
+    {
+      "epoch": 2.8851846663865155,
+      "grad_norm": 0.1879359632730484,
+      "learning_rate": 1.7250000000000002e-06,
+      "loss": 1.2167,
+      "step": 1931
+    },
+    {
+      "epoch": 2.88667880655554,
+      "grad_norm": 0.24511118233203888,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.3985,
+      "step": 1932
+    },
+    {
+      "epoch": 2.8881729467245645,
+      "grad_norm": 0.19905032217502594,
+      "learning_rate": 1.6750000000000003e-06,
+      "loss": 1.26,
+      "step": 1933
+    },
+    {
+      "epoch": 2.8896670868935894,
+      "grad_norm": 0.22167952358722687,
+      "learning_rate": 1.65e-06,
+      "loss": 1.3348,
+      "step": 1934
+    },
+    {
+      "epoch": 2.891161227062614,
+      "grad_norm": 0.41242149472236633,
+      "learning_rate": 1.6250000000000001e-06,
+      "loss": 1.3367,
+      "step": 1935
+    },
+    {
+      "epoch": 2.8926553672316384,
+      "grad_norm": 0.26797640323638916,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.1586,
+      "step": 1936
+    },
+    {
+      "epoch": 2.894149507400663,
+      "grad_norm": 0.20665213465690613,
+      "learning_rate": 1.5750000000000002e-06,
+      "loss": 1.2999,
+      "step": 1937
+    },
+    {
+      "epoch": 2.8956436475696874,
+      "grad_norm": 0.2320278286933899,
+      "learning_rate": 1.55e-06,
+      "loss": 1.2405,
+      "step": 1938
+    },
+    {
+      "epoch": 2.8971377877387123,
+      "grad_norm": 0.2435215711593628,
+      "learning_rate": 1.525e-06,
+      "loss": 1.4335,
+      "step": 1939
+    },
+    {
+      "epoch": 2.898631927907737,
+      "grad_norm": 0.202729731798172,
+      "learning_rate": 1.5e-06,
+      "loss": 1.2331,
+      "step": 1940
+    },
+    {
+      "epoch": 2.9001260680767613,
+      "grad_norm": 0.2189522087574005,
+      "learning_rate": 1.475e-06,
+      "loss": 1.2774,
+      "step": 1941
+    },
+    {
+      "epoch": 2.9016202082457863,
+      "grad_norm": 0.19130611419677734,
+      "learning_rate": 1.45e-06,
+      "loss": 1.2951,
+      "step": 1942
+    },
+    {
+      "epoch": 2.9031143484148108,
+      "grad_norm": 0.2006605565547943,
+      "learning_rate": 1.4250000000000001e-06,
+      "loss": 1.2765,
+      "step": 1943
+    },
+    {
+      "epoch": 2.9046084885838352,
+      "grad_norm": 0.7907626032829285,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.353,
+      "step": 1944
+    },
+    {
+      "epoch": 2.9061026287528597,
+      "grad_norm": 0.19130118191242218,
+      "learning_rate": 1.3750000000000002e-06,
+      "loss": 1.2826,
+      "step": 1945
+    },
+    {
+      "epoch": 2.9075967689218842,
+      "grad_norm": 0.2727353870868683,
+      "learning_rate": 1.35e-06,
+      "loss": 1.1702,
+      "step": 1946
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.28280338644981384,
+      "learning_rate": 1.325e-06,
+      "loss": 1.4216,
+      "step": 1947
+    },
+    {
+      "epoch": 2.9105850492599337,
+      "grad_norm": 0.17979927361011505,
+      "learning_rate": 1.3e-06,
+      "loss": 1.3568,
+      "step": 1948
+    },
+    {
+      "epoch": 2.912079189428958,
+      "grad_norm": 0.2216491550207138,
+      "learning_rate": 1.275e-06,
+      "loss": 1.2296,
+      "step": 1949
+    },
+    {
+      "epoch": 2.913573329597983,
+      "grad_norm": 1.3488661050796509,
+      "learning_rate": 1.25e-06,
+      "loss": 1.2513,
+      "step": 1950
+    },
+    {
+      "epoch": 2.9150674697670076,
+      "grad_norm": 2.417159080505371,
+      "learning_rate": 1.2250000000000001e-06,
+      "loss": 1.5253,
+      "step": 1951
+    },
+    {
+      "epoch": 2.916561609936032,
+      "grad_norm": 0.21374452114105225,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.3452,
+      "step": 1952
+    },
+    {
+      "epoch": 2.9180557501050566,
+      "grad_norm": 0.20700985193252563,
+      "learning_rate": 1.175e-06,
+      "loss": 1.2113,
+      "step": 1953
+    },
+    {
+      "epoch": 2.919549890274081,
+      "grad_norm": 0.19350235164165497,
+      "learning_rate": 1.15e-06,
+      "loss": 1.345,
+      "step": 1954
+    },
+    {
+      "epoch": 2.921044030443106,
+      "grad_norm": 0.1589677780866623,
+      "learning_rate": 1.125e-06,
+      "loss": 1.1717,
+      "step": 1955
+    },
+    {
+      "epoch": 2.9225381706121305,
+      "grad_norm": 0.22525258362293243,
+      "learning_rate": 1.1e-06,
+      "loss": 1.184,
+      "step": 1956
+    },
+    {
+      "epoch": 2.924032310781155,
+      "grad_norm": 0.17294123768806458,
+      "learning_rate": 1.0749999999999999e-06,
+      "loss": 1.2835,
+      "step": 1957
+    },
+    {
+      "epoch": 2.92552645095018,
+      "grad_norm": 0.21778512001037598,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.3665,
+      "step": 1958
+    },
+    {
+      "epoch": 2.9270205911192044,
+      "grad_norm": 0.17190255224704742,
+      "learning_rate": 1.0250000000000001e-06,
+      "loss": 1.2313,
+      "step": 1959
+    },
+    {
+      "epoch": 2.928514731288229,
+      "grad_norm": 0.20868921279907227,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.2845,
+      "step": 1960
+    },
+    {
+      "epoch": 2.930008871457254,
+      "grad_norm": 0.19007806479930878,
+      "learning_rate": 9.75e-07,
+      "loss": 1.1979,
+      "step": 1961
+    },
+    {
+      "epoch": 2.9315030116262784,
+      "grad_norm": 0.48530903458595276,
+      "learning_rate": 9.5e-07,
+      "loss": 1.3446,
+      "step": 1962
+    },
+    {
+      "epoch": 2.932997151795303,
+      "grad_norm": 0.3450883626937866,
+      "learning_rate": 9.25e-07,
+      "loss": 1.3834,
+      "step": 1963
+    },
+    {
+      "epoch": 2.9344912919643273,
+      "grad_norm": 0.23902513086795807,
+      "learning_rate": 9e-07,
+      "loss": 1.286,
+      "step": 1964
+    },
+    {
+      "epoch": 2.935985432133352,
+      "grad_norm": 0.44902661442756653,
+      "learning_rate": 8.750000000000001e-07,
+      "loss": 1.2347,
+      "step": 1965
+    },
+    {
+      "epoch": 2.9374795723023768,
+      "grad_norm": 0.24165555834770203,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 1.254,
+      "step": 1966
+    },
+    {
+      "epoch": 2.9389737124714013,
+      "grad_norm": 0.20330724120140076,
+      "learning_rate": 8.25e-07,
+      "loss": 1.4177,
+      "step": 1967
+    },
+    {
+      "epoch": 2.9404678526404258,
+      "grad_norm": 0.25209009647369385,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.4941,
+      "step": 1968
+    },
+    {
+      "epoch": 2.9419619928094507,
+      "grad_norm": 0.240138441324234,
+      "learning_rate": 7.75e-07,
+      "loss": 1.3458,
+      "step": 1969
+    },
+    {
+      "epoch": 2.943456132978475,
+      "grad_norm": 0.34439221024513245,
+      "learning_rate": 7.5e-07,
+      "loss": 1.4309,
+      "step": 1970
+    },
+    {
+      "epoch": 2.9449502731474997,
+      "grad_norm": 0.22321772575378418,
+      "learning_rate": 7.25e-07,
+      "loss": 1.4066,
+      "step": 1971
+    },
+    {
+      "epoch": 2.946444413316524,
+      "grad_norm": 0.2187803089618683,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.3323,
+      "step": 1972
+    },
+    {
+      "epoch": 2.9479385534855487,
+      "grad_norm": 0.2270226627588272,
+      "learning_rate": 6.75e-07,
+      "loss": 1.243,
+      "step": 1973
+    },
+    {
+      "epoch": 2.9494326936545736,
+      "grad_norm": 0.20585761964321136,
+      "learning_rate": 6.5e-07,
+      "loss": 1.2544,
+      "step": 1974
+    },
+    {
+      "epoch": 2.950926833823598,
+      "grad_norm": 0.18699559569358826,
+      "learning_rate": 6.25e-07,
+      "loss": 1.1969,
+      "step": 1975
+    },
+    {
+      "epoch": 2.9524209739926226,
+      "grad_norm": 0.5145549178123474,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.1359,
+      "step": 1976
+    },
+    {
+      "epoch": 2.9539151141616475,
+      "grad_norm": 0.23603378236293793,
+      "learning_rate": 5.75e-07,
+      "loss": 1.4074,
+      "step": 1977
+    },
+    {
+      "epoch": 2.955409254330672,
+      "grad_norm": 0.24418418109416962,
+      "learning_rate": 5.5e-07,
+      "loss": 1.4617,
+      "step": 1978
+    },
+    {
+      "epoch": 2.9569033944996965,
+      "grad_norm": 0.2115854173898697,
+      "learning_rate": 5.250000000000001e-07,
+      "loss": 1.3637,
+      "step": 1979
+    },
+    {
+      "epoch": 2.958397534668721,
+      "grad_norm": 0.2542995810508728,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.2526,
+      "step": 1980
+    },
+    {
+      "epoch": 2.9598916748377455,
+      "grad_norm": 0.20433180034160614,
+      "learning_rate": 4.75e-07,
+      "loss": 1.2907,
+      "step": 1981
+    },
+    {
+      "epoch": 2.9613858150067704,
+      "grad_norm": 0.19650673866271973,
+      "learning_rate": 4.5e-07,
+      "loss": 1.2711,
+      "step": 1982
+    },
+    {
+      "epoch": 2.962879955175795,
+      "grad_norm": 0.4737726151943207,
+      "learning_rate": 4.2500000000000006e-07,
+      "loss": 1.2896,
+      "step": 1983
+    },
+    {
+      "epoch": 2.9643740953448194,
+      "grad_norm": 0.23286136984825134,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.219,
+      "step": 1984
+    },
+    {
+      "epoch": 2.9658682355138444,
+      "grad_norm": 0.17285366356372833,
+      "learning_rate": 3.75e-07,
+      "loss": 1.2118,
+      "step": 1985
+    },
+    {
+      "epoch": 2.967362375682869,
+      "grad_norm": 0.15881465375423431,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.4014,
+      "step": 1986
+    },
+    {
+      "epoch": 2.9688565158518934,
+      "grad_norm": 0.19492295384407043,
+      "learning_rate": 3.25e-07,
+      "loss": 1.2168,
+      "step": 1987
+    },
+    {
+      "epoch": 2.970350656020918,
+      "grad_norm": 0.22486300766468048,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.4393,
+      "step": 1988
+    },
+    {
+      "epoch": 2.9718447961899424,
+      "grad_norm": 0.24522030353546143,
+      "learning_rate": 2.75e-07,
+      "loss": 1.3018,
+      "step": 1989
+    },
+    {
+      "epoch": 2.9733389363589673,
+      "grad_norm": 0.34732580184936523,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.3166,
+      "step": 1990
+    },
+    {
+      "epoch": 2.974833076527992,
+      "grad_norm": 0.22350691258907318,
+      "learning_rate": 2.25e-07,
+      "loss": 1.3465,
+      "step": 1991
+    },
+    {
+      "epoch": 2.9763272166970163,
+      "grad_norm": 0.20143184065818787,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.2512,
+      "step": 1992
+    },
+    {
+      "epoch": 2.977821356866041,
+      "grad_norm": 0.31710419058799744,
+      "learning_rate": 1.7500000000000002e-07,
+      "loss": 1.1942,
+      "step": 1993
+    },
+    {
+      "epoch": 2.9793154970350657,
+      "grad_norm": 0.60710209608078,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.2493,
+      "step": 1994
+    },
+    {
+      "epoch": 2.98080963720409,
+      "grad_norm": 0.27234354615211487,
+      "learning_rate": 1.2500000000000002e-07,
+      "loss": 1.2728,
+      "step": 1995
+    },
+    {
+      "epoch": 2.9823037773731147,
+      "grad_norm": 0.2213677167892456,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.3398,
+      "step": 1996
+    },
+    {
+      "epoch": 2.983797917542139,
+      "grad_norm": 0.187736377120018,
+      "learning_rate": 7.500000000000001e-08,
+      "loss": 1.3615,
+      "step": 1997
+    },
+    {
+      "epoch": 2.985292057711164,
+      "grad_norm": 0.1614825576543808,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.1303,
+      "step": 1998
+    },
+    {
+      "epoch": 2.9867861978801886,
+      "grad_norm": 0.18150056898593903,
+      "learning_rate": 2.5000000000000002e-08,
+      "loss": 1.2191,
+      "step": 1999
+    },
+    {
+      "epoch": 2.988280338049213,
+      "grad_norm": 0.38465365767478943,
+      "learning_rate": 0.0,
+      "loss": 1.3653,
+      "step": 2000
+    },
+    {
+      "epoch": 2.988280338049213,
+      "step": 2000,
+      "total_flos": 3.3909236563968e+16,
+      "train_loss": 1.523793903529644,
+      "train_runtime": 5754.2327,
+      "train_samples_per_second": 11.122,
+      "train_steps_per_second": 0.348
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.3909236563968e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}