{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9998200683740177,
  "eval_steps": 100,
  "global_step": 6252,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002399088346428357,
      "grad_norm": 34.25,
      "learning_rate": 2.6652452025586355e-08,
      "loss": 1.3138,
      "step": 5
    },
    {
      "epoch": 0.004798176692856714,
      "grad_norm": 42.5,
      "learning_rate": 5.330490405117271e-08,
      "loss": 1.3611,
      "step": 10
    },
    {
      "epoch": 0.007197265039285071,
      "grad_norm": 103.5,
      "learning_rate": 7.995735607675907e-08,
      "loss": 1.2608,
      "step": 15
    },
    {
      "epoch": 0.009596353385713428,
      "grad_norm": 9.6875,
      "learning_rate": 1.0660980810234542e-07,
      "loss": 1.2714,
      "step": 20
    },
    {
      "epoch": 0.011995441732141785,
      "grad_norm": 7.5,
      "learning_rate": 1.3326226012793176e-07,
      "loss": 1.4005,
      "step": 25
    },
    {
      "epoch": 0.014394530078570143,
      "grad_norm": 11.0,
      "learning_rate": 1.5991471215351813e-07,
      "loss": 1.3235,
      "step": 30
    },
    {
      "epoch": 0.0167936184249985,
      "grad_norm": 10.625,
      "learning_rate": 1.8656716417910447e-07,
      "loss": 1.2774,
      "step": 35
    },
    {
      "epoch": 0.019192706771426857,
      "grad_norm": 6.46875,
      "learning_rate": 2.1321961620469084e-07,
      "loss": 1.1795,
      "step": 40
    },
    {
      "epoch": 0.021591795117855216,
      "grad_norm": 8.0625,
      "learning_rate": 2.398720682302772e-07,
      "loss": 1.345,
      "step": 45
    },
    {
      "epoch": 0.02399088346428357,
      "grad_norm": 10.625,
      "learning_rate": 2.665245202558635e-07,
      "loss": 1.3542,
      "step": 50
    },
    {
      "epoch": 0.02638997181071193,
      "grad_norm": 10.75,
      "learning_rate": 2.931769722814499e-07,
      "loss": 1.2539,
      "step": 55
    },
    {
      "epoch": 0.028789060157140285,
      "grad_norm": 9.1875,
      "learning_rate": 3.1982942430703626e-07,
      "loss": 1.4054,
      "step": 60
    },
    {
      "epoch": 0.031188148503568644,
      "grad_norm": 7.125,
      "learning_rate": 3.4648187633262263e-07,
      "loss": 1.4234,
      "step": 65
    },
    {
      "epoch": 0.033587236849997,
      "grad_norm": 8.6875,
      "learning_rate": 3.7313432835820895e-07,
      "loss": 1.1964,
      "step": 70
    },
    {
      "epoch": 0.03598632519642536,
      "grad_norm": 11.75,
      "learning_rate": 3.9978678038379537e-07,
      "loss": 1.3206,
      "step": 75
    },
    {
      "epoch": 0.038385413542853714,
      "grad_norm": 9.8125,
      "learning_rate": 4.264392324093817e-07,
      "loss": 1.1716,
      "step": 80
    },
    {
      "epoch": 0.04078450188928207,
      "grad_norm": 20.0,
      "learning_rate": 4.53091684434968e-07,
      "loss": 1.3047,
      "step": 85
    },
    {
      "epoch": 0.04318359023571043,
      "grad_norm": 8.3125,
      "learning_rate": 4.797441364605544e-07,
      "loss": 1.2356,
      "step": 90
    },
    {
      "epoch": 0.04558267858213879,
      "grad_norm": 17.875,
      "learning_rate": 5.063965884861407e-07,
      "loss": 1.2631,
      "step": 95
    },
    {
      "epoch": 0.04798176692856714,
      "grad_norm": 8.8125,
      "learning_rate": 5.33049040511727e-07,
      "loss": 1.3332,
      "step": 100
    },
    {
      "epoch": 0.04798176692856714,
      "eval_loss": 1.3139781951904297,
      "eval_runtime": 177.7725,
      "eval_samples_per_second": 41.682,
      "eval_steps_per_second": 10.423,
      "step": 100
    },
    {
      "epoch": 0.0503808552749955,
      "grad_norm": 13.75,
      "learning_rate": 5.597014925373135e-07,
      "loss": 1.2867,
      "step": 105
    },
    {
      "epoch": 0.05277994362142386,
      "grad_norm": 9.3125,
      "learning_rate": 5.863539445628998e-07,
      "loss": 1.2629,
      "step": 110
    },
    {
      "epoch": 0.05517903196785222,
      "grad_norm": 11.625,
      "learning_rate": 6.130063965884862e-07,
      "loss": 1.3406,
      "step": 115
    },
    {
      "epoch": 0.05757812031428057,
      "grad_norm": 8.375,
      "learning_rate": 6.396588486140725e-07,
      "loss": 1.2352,
      "step": 120
    },
    {
      "epoch": 0.05997720866070893,
      "grad_norm": 18.25,
      "learning_rate": 6.663113006396589e-07,
      "loss": 1.3535,
      "step": 125
    },
    {
      "epoch": 0.06237629700713729,
      "grad_norm": 11.375,
      "learning_rate": 6.929637526652453e-07,
      "loss": 1.299,
      "step": 130
    },
    {
      "epoch": 0.06477538535356564,
      "grad_norm": 7.15625,
      "learning_rate": 7.196162046908316e-07,
      "loss": 1.2223,
      "step": 135
    },
    {
      "epoch": 0.067174473699994,
      "grad_norm": 9.0,
      "learning_rate": 7.462686567164179e-07,
      "loss": 1.2659,
      "step": 140
    },
    {
      "epoch": 0.06957356204642236,
      "grad_norm": 140.0,
      "learning_rate": 7.729211087420044e-07,
      "loss": 1.3394,
      "step": 145
    },
    {
      "epoch": 0.07197265039285072,
      "grad_norm": 11.0,
      "learning_rate": 7.995735607675907e-07,
      "loss": 1.3174,
      "step": 150
    },
    {
      "epoch": 0.07437173873927908,
      "grad_norm": 49.25,
      "learning_rate": 8.26226012793177e-07,
      "loss": 1.2811,
      "step": 155
    },
    {
      "epoch": 0.07677082708570743,
      "grad_norm": 8.625,
      "learning_rate": 8.528784648187634e-07,
      "loss": 1.2707,
      "step": 160
    },
    {
      "epoch": 0.07916991543213579,
      "grad_norm": 7.625,
      "learning_rate": 8.795309168443497e-07,
      "loss": 1.2574,
      "step": 165
    },
    {
      "epoch": 0.08156900377856414,
      "grad_norm": 18.125,
      "learning_rate": 9.06183368869936e-07,
      "loss": 1.3314,
      "step": 170
    },
    {
      "epoch": 0.0839680921249925,
      "grad_norm": 8.0625,
      "learning_rate": 9.328358208955225e-07,
      "loss": 1.3046,
      "step": 175
    },
    {
      "epoch": 0.08636718047142086,
      "grad_norm": 6.875,
      "learning_rate": 9.594882729211088e-07,
      "loss": 1.3634,
      "step": 180
    },
    {
      "epoch": 0.08876626881784921,
      "grad_norm": 85.0,
      "learning_rate": 9.861407249466952e-07,
      "loss": 1.2065,
      "step": 185
    },
    {
      "epoch": 0.09116535716427758,
      "grad_norm": 13.6875,
      "learning_rate": 1.0127931769722815e-06,
      "loss": 1.188,
      "step": 190
    },
    {
      "epoch": 0.09356444551070593,
      "grad_norm": 31.125,
      "learning_rate": 1.0394456289978678e-06,
      "loss": 1.3453,
      "step": 195
    },
    {
      "epoch": 0.09596353385713428,
      "grad_norm": 16.75,
      "learning_rate": 1.066098081023454e-06,
      "loss": 1.2185,
      "step": 200
    },
    {
      "epoch": 0.09596353385713428,
      "eval_loss": 1.2878996133804321,
      "eval_runtime": 176.065,
      "eval_samples_per_second": 42.087,
      "eval_steps_per_second": 10.525,
      "step": 200
    },
    {
      "epoch": 0.09836262220356265,
      "grad_norm": 12.75,
      "learning_rate": 1.0927505330490406e-06,
      "loss": 1.2153,
      "step": 205
    },
    {
      "epoch": 0.100761710549991,
      "grad_norm": 56.0,
      "learning_rate": 1.119402985074627e-06,
      "loss": 1.3089,
      "step": 210
    },
    {
      "epoch": 0.10316079889641937,
      "grad_norm": 17.5,
      "learning_rate": 1.1460554371002133e-06,
      "loss": 1.228,
      "step": 215
    },
    {
      "epoch": 0.10555988724284772,
      "grad_norm": 14.5,
      "learning_rate": 1.1727078891257996e-06,
      "loss": 1.293,
      "step": 220
    },
    {
      "epoch": 0.10795897558927607,
      "grad_norm": 9.1875,
      "learning_rate": 1.199360341151386e-06,
      "loss": 1.2792,
      "step": 225
    },
    {
      "epoch": 0.11035806393570444,
      "grad_norm": 8.1875,
      "learning_rate": 1.2260127931769724e-06,
      "loss": 1.3112,
      "step": 230
    },
    {
      "epoch": 0.11275715228213279,
      "grad_norm": 11.125,
      "learning_rate": 1.2526652452025587e-06,
      "loss": 1.2376,
      "step": 235
    },
    {
      "epoch": 0.11515624062856114,
      "grad_norm": 7.9375,
      "learning_rate": 1.279317697228145e-06,
      "loss": 1.2712,
      "step": 240
    },
    {
      "epoch": 0.1175553289749895,
      "grad_norm": 8.5,
      "learning_rate": 1.3059701492537314e-06,
      "loss": 1.2524,
      "step": 245
    },
    {
      "epoch": 0.11995441732141786,
      "grad_norm": 19.5,
      "learning_rate": 1.3326226012793179e-06,
      "loss": 1.3747,
      "step": 250
    },
    {
      "epoch": 0.12235350566784622,
      "grad_norm": 13.8125,
      "learning_rate": 1.3592750533049042e-06,
      "loss": 1.1628,
      "step": 255
    },
    {
      "epoch": 0.12475259401427458,
      "grad_norm": 7.96875,
      "learning_rate": 1.3859275053304905e-06,
      "loss": 1.2235,
      "step": 260
    },
    {
      "epoch": 0.12715168236070293,
      "grad_norm": 13.625,
      "learning_rate": 1.412579957356077e-06,
      "loss": 1.2642,
      "step": 265
    },
    {
      "epoch": 0.12955077070713128,
      "grad_norm": 5.21875,
      "learning_rate": 1.4392324093816632e-06,
      "loss": 1.2196,
      "step": 270
    },
    {
      "epoch": 0.13194985905355966,
      "grad_norm": 12.875,
      "learning_rate": 1.4658848614072497e-06,
      "loss": 1.2465,
      "step": 275
    },
    {
      "epoch": 0.134348947399988,
      "grad_norm": 14.1875,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 1.2708,
      "step": 280
    },
    {
      "epoch": 0.13674803574641636,
      "grad_norm": 10.625,
      "learning_rate": 1.5191897654584223e-06,
      "loss": 1.1976,
      "step": 285
    },
    {
      "epoch": 0.13914712409284472,
      "grad_norm": 19.875,
      "learning_rate": 1.5458422174840088e-06,
      "loss": 1.2336,
      "step": 290
    },
    {
      "epoch": 0.14154621243927307,
      "grad_norm": 13.9375,
      "learning_rate": 1.572494669509595e-06,
      "loss": 1.2791,
      "step": 295
    },
    {
      "epoch": 0.14394530078570145,
      "grad_norm": 32.75,
      "learning_rate": 1.5991471215351815e-06,
      "loss": 1.1976,
      "step": 300
    },
    {
      "epoch": 0.14394530078570145,
      "eval_loss": 1.2532644271850586,
      "eval_runtime": 176.1055,
      "eval_samples_per_second": 42.077,
      "eval_steps_per_second": 10.522,
      "step": 300
    },
    {
      "epoch": 0.1463443891321298,
      "grad_norm": 5.25,
      "learning_rate": 1.6257995735607676e-06,
      "loss": 1.2925,
      "step": 305
    },
    {
      "epoch": 0.14874347747855815,
      "grad_norm": 6.78125,
      "learning_rate": 1.652452025586354e-06,
      "loss": 1.2051,
      "step": 310
    },
    {
      "epoch": 0.1511425658249865,
      "grad_norm": 8.9375,
      "learning_rate": 1.6791044776119406e-06,
      "loss": 1.3741,
      "step": 315
    },
    {
      "epoch": 0.15354165417141485,
      "grad_norm": 83.5,
      "learning_rate": 1.7057569296375267e-06,
      "loss": 1.219,
      "step": 320
    },
    {
      "epoch": 0.1559407425178432,
      "grad_norm": 9.75,
      "learning_rate": 1.7324093816631133e-06,
      "loss": 1.233,
      "step": 325
    },
    {
      "epoch": 0.15833983086427159,
      "grad_norm": 12.1875,
      "learning_rate": 1.7590618336886994e-06,
      "loss": 1.179,
      "step": 330
    },
    {
      "epoch": 0.16073891921069994,
      "grad_norm": 14.4375,
      "learning_rate": 1.7857142857142859e-06,
      "loss": 1.2121,
      "step": 335
    },
    {
      "epoch": 0.1631380075571283,
      "grad_norm": 7.9375,
      "learning_rate": 1.812366737739872e-06,
      "loss": 1.1753,
      "step": 340
    },
    {
      "epoch": 0.16553709590355664,
      "grad_norm": 4.625,
      "learning_rate": 1.8390191897654585e-06,
      "loss": 1.2174,
      "step": 345
    },
    {
      "epoch": 0.167936184249985,
      "grad_norm": 4.28125,
      "learning_rate": 1.865671641791045e-06,
      "loss": 1.219,
      "step": 350
    },
    {
      "epoch": 0.17033527259641337,
      "grad_norm": 9.6875,
      "learning_rate": 1.8923240938166312e-06,
      "loss": 1.1994,
      "step": 355
    },
    {
      "epoch": 0.17273436094284172,
      "grad_norm": 25.5,
      "learning_rate": 1.9189765458422177e-06,
      "loss": 1.2914,
      "step": 360
    },
    {
      "epoch": 0.17513344928927008,
      "grad_norm": 15.5625,
      "learning_rate": 1.945628997867804e-06,
      "loss": 1.2052,
      "step": 365
    },
    {
      "epoch": 0.17753253763569843,
      "grad_norm": 96.0,
      "learning_rate": 1.9722814498933903e-06,
      "loss": 1.2929,
      "step": 370
    },
    {
      "epoch": 0.17993162598212678,
      "grad_norm": 5.09375,
      "learning_rate": 1.9989339019189766e-06,
      "loss": 1.2518,
      "step": 375
    },
    {
      "epoch": 0.18233071432855516,
      "grad_norm": 6.9375,
      "learning_rate": 2.025586353944563e-06,
      "loss": 1.1894,
      "step": 380
    },
    {
      "epoch": 0.1847298026749835,
      "grad_norm": 8.375,
      "learning_rate": 2.0522388059701497e-06,
      "loss": 1.2302,
      "step": 385
    },
    {
      "epoch": 0.18712889102141186,
      "grad_norm": 5.875,
      "learning_rate": 2.0788912579957356e-06,
      "loss": 1.2647,
      "step": 390
    },
    {
      "epoch": 0.18952797936784022,
      "grad_norm": 4.6875,
      "learning_rate": 2.1055437100213223e-06,
      "loss": 1.1564,
      "step": 395
    },
    {
      "epoch": 0.19192706771426857,
      "grad_norm": 5.03125,
      "learning_rate": 2.132196162046908e-06,
      "loss": 1.1627,
      "step": 400
    },
    {
      "epoch": 0.19192706771426857,
      "eval_loss": 1.2169007062911987,
      "eval_runtime": 176.1168,
      "eval_samples_per_second": 42.074,
      "eval_steps_per_second": 10.521,
      "step": 400
    },
    {
      "epoch": 0.19432615606069695,
      "grad_norm": 40.5,
      "learning_rate": 2.158848614072495e-06,
      "loss": 1.25,
      "step": 405
    },
    {
      "epoch": 0.1967252444071253,
      "grad_norm": 7.53125,
      "learning_rate": 2.1855010660980813e-06,
      "loss": 1.1966,
      "step": 410
    },
    {
      "epoch": 0.19912433275355365,
      "grad_norm": 10.0625,
      "learning_rate": 2.2121535181236676e-06,
      "loss": 1.1397,
      "step": 415
    },
    {
      "epoch": 0.201523421099982,
      "grad_norm": 7.25,
      "learning_rate": 2.238805970149254e-06,
      "loss": 1.2399,
      "step": 420
    },
    {
      "epoch": 0.20392250944641035,
      "grad_norm": 6.4375,
      "learning_rate": 2.26545842217484e-06,
      "loss": 1.1008,
      "step": 425
    },
    {
      "epoch": 0.20632159779283873,
      "grad_norm": 6.09375,
      "learning_rate": 2.2921108742004265e-06,
      "loss": 1.2229,
      "step": 430
    },
    {
      "epoch": 0.2087206861392671,
      "grad_norm": 12.75,
      "learning_rate": 2.318763326226013e-06,
      "loss": 1.1546,
      "step": 435
    },
    {
      "epoch": 0.21111977448569544,
      "grad_norm": 5.875,
      "learning_rate": 2.345415778251599e-06,
      "loss": 1.1509,
      "step": 440
    },
    {
      "epoch": 0.2135188628321238,
      "grad_norm": 3.859375,
      "learning_rate": 2.372068230277186e-06,
      "loss": 1.1714,
      "step": 445
    },
    {
      "epoch": 0.21591795117855214,
      "grad_norm": 4.5,
      "learning_rate": 2.398720682302772e-06,
      "loss": 1.2313,
      "step": 450
    },
    {
      "epoch": 0.2183170395249805,
      "grad_norm": 7.875,
      "learning_rate": 2.4253731343283585e-06,
      "loss": 1.1727,
      "step": 455
    },
    {
      "epoch": 0.22071612787140887,
      "grad_norm": 4.25,
      "learning_rate": 2.452025586353945e-06,
      "loss": 1.2308,
      "step": 460
    },
    {
      "epoch": 0.22311521621783723,
      "grad_norm": 5.78125,
      "learning_rate": 2.478678038379531e-06,
      "loss": 1.1464,
      "step": 465
    },
    {
      "epoch": 0.22551430456426558,
      "grad_norm": 9.0625,
      "learning_rate": 2.5053304904051175e-06,
      "loss": 1.0961,
      "step": 470
    },
    {
      "epoch": 0.22791339291069393,
      "grad_norm": 9.3125,
      "learning_rate": 2.531982942430704e-06,
      "loss": 1.2813,
      "step": 475
    },
    {
      "epoch": 0.23031248125712228,
      "grad_norm": 17.375,
      "learning_rate": 2.55863539445629e-06,
      "loss": 1.1666,
      "step": 480
    },
    {
      "epoch": 0.23271156960355066,
      "grad_norm": 4.875,
      "learning_rate": 2.5852878464818764e-06,
      "loss": 1.1806,
      "step": 485
    },
    {
      "epoch": 0.235110657949979,
      "grad_norm": 6.125,
      "learning_rate": 2.6119402985074627e-06,
      "loss": 1.2986,
      "step": 490
    },
    {
      "epoch": 0.23750974629640736,
      "grad_norm": 8.5,
      "learning_rate": 2.6385927505330495e-06,
      "loss": 1.1637,
      "step": 495
    },
    {
      "epoch": 0.23990883464283572,
      "grad_norm": 12.625,
      "learning_rate": 2.6652452025586358e-06,
      "loss": 1.178,
      "step": 500
    },
    {
      "epoch": 0.23990883464283572,
      "eval_loss": 1.1765925884246826,
      "eval_runtime": 176.0206,
      "eval_samples_per_second": 42.097,
      "eval_steps_per_second": 10.527,
      "step": 500
    },
    {
      "epoch": 0.24230792298926407,
      "grad_norm": 6.5625,
      "learning_rate": 2.6918976545842217e-06,
      "loss": 1.0748,
      "step": 505
    },
    {
      "epoch": 0.24470701133569245,
      "grad_norm": 5.34375,
      "learning_rate": 2.7185501066098084e-06,
      "loss": 1.0757,
      "step": 510
    },
    {
      "epoch": 0.2471060996821208,
      "grad_norm": 15.9375,
      "learning_rate": 2.7452025586353947e-06,
      "loss": 1.2197,
      "step": 515
    },
    {
      "epoch": 0.24950518802854915,
      "grad_norm": 6.1875,
      "learning_rate": 2.771855010660981e-06,
      "loss": 1.2274,
      "step": 520
    },
    {
      "epoch": 0.25190427637497753,
      "grad_norm": 6.34375,
      "learning_rate": 2.798507462686567e-06,
      "loss": 1.1705,
      "step": 525
    },
    {
      "epoch": 0.25430336472140586,
      "grad_norm": 3.9375,
      "learning_rate": 2.825159914712154e-06,
      "loss": 1.179,
      "step": 530
    },
    {
      "epoch": 0.25670245306783424,
      "grad_norm": 7.46875,
      "learning_rate": 2.85181236673774e-06,
      "loss": 1.1046,
      "step": 535
    },
    {
      "epoch": 0.25910154141426256,
      "grad_norm": 5.625,
      "learning_rate": 2.8784648187633263e-06,
      "loss": 1.1889,
      "step": 540
    },
    {
      "epoch": 0.26150062976069094,
      "grad_norm": 6.25,
      "learning_rate": 2.905117270788913e-06,
      "loss": 1.136,
      "step": 545
    },
    {
      "epoch": 0.2638997181071193,
      "grad_norm": 20.75,
      "learning_rate": 2.9317697228144994e-06,
      "loss": 1.1492,
      "step": 550
    },
    {
      "epoch": 0.26629880645354764,
      "grad_norm": 7.84375,
      "learning_rate": 2.9584221748400853e-06,
      "loss": 1.1257,
      "step": 555
    },
    {
      "epoch": 0.268697894799976,
      "grad_norm": 4.78125,
      "learning_rate": 2.9850746268656716e-06,
      "loss": 1.0821,
      "step": 560
    },
    {
      "epoch": 0.27109698314640435,
      "grad_norm": 4.875,
      "learning_rate": 3.0117270788912583e-06,
      "loss": 1.1404,
      "step": 565
    },
    {
      "epoch": 0.2734960714928327,
      "grad_norm": 7.8125,
      "learning_rate": 3.0383795309168446e-06,
      "loss": 1.0904,
      "step": 570
    },
    {
      "epoch": 0.2758951598392611,
      "grad_norm": 4.59375,
      "learning_rate": 3.065031982942431e-06,
      "loss": 1.0591,
      "step": 575
    },
    {
      "epoch": 0.27829424818568943,
      "grad_norm": 8.875,
      "learning_rate": 3.0916844349680177e-06,
      "loss": 1.0304,
      "step": 580
    },
    {
      "epoch": 0.2806933365321178,
      "grad_norm": 5.0625,
      "learning_rate": 3.1183368869936036e-06,
      "loss": 1.1397,
      "step": 585
    },
    {
      "epoch": 0.28309242487854613,
      "grad_norm": 12.875,
      "learning_rate": 3.14498933901919e-06,
      "loss": 1.1217,
      "step": 590
    },
    {
      "epoch": 0.2854915132249745,
      "grad_norm": 4.9375,
      "learning_rate": 3.1716417910447766e-06,
      "loss": 1.2383,
      "step": 595
    },
    {
      "epoch": 0.2878906015714029,
      "grad_norm": 6.9375,
      "learning_rate": 3.198294243070363e-06,
      "loss": 1.133,
      "step": 600
    },
    {
      "epoch": 0.2878906015714029,
      "eval_loss": 1.1296346187591553,
      "eval_runtime": 175.841,
      "eval_samples_per_second": 42.14,
      "eval_steps_per_second": 10.538,
      "step": 600
    },
    {
      "epoch": 0.2902896899178312,
      "grad_norm": 4.9375,
      "learning_rate": 3.224946695095949e-06,
      "loss": 1.0715,
      "step": 605
    },
    {
      "epoch": 0.2926887782642596,
      "grad_norm": 5.09375,
      "learning_rate": 3.251599147121535e-06,
      "loss": 1.0086,
      "step": 610
    },
    {
      "epoch": 0.2950878666106879,
      "grad_norm": 7.3125,
      "learning_rate": 3.278251599147122e-06,
      "loss": 1.0881,
      "step": 615
    },
    {
      "epoch": 0.2974869549571163,
      "grad_norm": 10.125,
      "learning_rate": 3.304904051172708e-06,
      "loss": 1.0767,
      "step": 620
    },
    {
      "epoch": 0.2998860433035447,
      "grad_norm": 6.90625,
      "learning_rate": 3.3315565031982945e-06,
      "loss": 1.0721,
      "step": 625
    },
    {
      "epoch": 0.302285131649973,
      "grad_norm": 6.875,
      "learning_rate": 3.3582089552238813e-06,
      "loss": 1.1838,
      "step": 630
    },
    {
      "epoch": 0.3046842199964014,
      "grad_norm": 7.75,
      "learning_rate": 3.384861407249467e-06,
      "loss": 1.0761,
      "step": 635
    },
    {
      "epoch": 0.3070833083428297,
      "grad_norm": 7.59375,
      "learning_rate": 3.4115138592750535e-06,
      "loss": 1.0424,
      "step": 640
    },
    {
      "epoch": 0.3094823966892581,
      "grad_norm": 37.0,
      "learning_rate": 3.4381663113006398e-06,
      "loss": 1.0834,
      "step": 645
    },
    {
      "epoch": 0.3118814850356864,
      "grad_norm": 5.21875,
      "learning_rate": 3.4648187633262265e-06,
      "loss": 1.1291,
      "step": 650
    },
    {
      "epoch": 0.3142805733821148,
      "grad_norm": 4.5625,
      "learning_rate": 3.491471215351813e-06,
      "loss": 1.1013,
      "step": 655
    },
    {
      "epoch": 0.31667966172854317,
      "grad_norm": 6.1875,
      "learning_rate": 3.5181236673773987e-06,
      "loss": 1.0895,
      "step": 660
    },
    {
      "epoch": 0.3190787500749715,
      "grad_norm": 7.4375,
      "learning_rate": 3.5447761194029855e-06,
      "loss": 1.1223,
      "step": 665
    },
    {
      "epoch": 0.3214778384213999,
      "grad_norm": 4.3125,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 1.0967,
      "step": 670
    },
    {
      "epoch": 0.3238769267678282,
      "grad_norm": 4.40625,
      "learning_rate": 3.598081023454158e-06,
      "loss": 1.1102,
      "step": 675
    },
    {
      "epoch": 0.3262760151142566,
      "grad_norm": 6.40625,
      "learning_rate": 3.624733475479744e-06,
      "loss": 1.1283,
      "step": 680
    },
    {
      "epoch": 0.32867510346068496,
      "grad_norm": 4.5,
      "learning_rate": 3.651385927505331e-06,
      "loss": 1.0848,
      "step": 685
    },
    {
      "epoch": 0.3310741918071133,
      "grad_norm": 4.5625,
      "learning_rate": 3.678038379530917e-06,
      "loss": 1.0805,
      "step": 690
    },
    {
      "epoch": 0.33347328015354166,
      "grad_norm": 4.28125,
      "learning_rate": 3.7046908315565034e-06,
      "loss": 1.0819,
      "step": 695
    },
    {
      "epoch": 0.33587236849997,
      "grad_norm": 7.84375,
      "learning_rate": 3.73134328358209e-06,
      "loss": 1.0466,
      "step": 700
    },
    {
      "epoch": 0.33587236849997,
      "eval_loss": 1.0983320474624634,
      "eval_runtime": 233.7274,
      "eval_samples_per_second": 31.704,
      "eval_steps_per_second": 7.928,
      "step": 700
    },
    {
      "epoch": 0.33827145684639837,
      "grad_norm": 15.3125,
      "learning_rate": 3.7579957356076764e-06,
      "loss": 1.1632,
      "step": 705
    },
    {
      "epoch": 0.34067054519282675,
      "grad_norm": 9.125,
      "learning_rate": 3.7846481876332623e-06,
      "loss": 1.0408,
      "step": 710
    },
    {
      "epoch": 0.34306963353925507,
      "grad_norm": 6.34375,
      "learning_rate": 3.8113006396588486e-06,
      "loss": 1.0748,
      "step": 715
    },
    {
      "epoch": 0.34546872188568345,
      "grad_norm": 3.78125,
      "learning_rate": 3.837953091684435e-06,
      "loss": 1.0707,
      "step": 720
    },
    {
      "epoch": 0.3478678102321118,
      "grad_norm": 4.28125,
      "learning_rate": 3.864605543710022e-06,
      "loss": 1.1292,
      "step": 725
    },
    {
      "epoch": 0.35026689857854015,
      "grad_norm": 4.28125,
      "learning_rate": 3.891257995735608e-06,
      "loss": 1.0883,
      "step": 730
    },
    {
      "epoch": 0.35266598692496853,
      "grad_norm": 6.15625,
      "learning_rate": 3.917910447761194e-06,
      "loss": 1.0613,
      "step": 735
    },
    {
      "epoch": 0.35506507527139686,
      "grad_norm": 4.125,
      "learning_rate": 3.944562899786781e-06,
      "loss": 1.1056,
      "step": 740
    },
    {
      "epoch": 0.35746416361782524,
      "grad_norm": 5.09375,
      "learning_rate": 3.971215351812367e-06,
      "loss": 1.1064,
      "step": 745
    },
    {
      "epoch": 0.35986325196425356,
      "grad_norm": 4.3125,
      "learning_rate": 3.997867803837953e-06,
      "loss": 1.0194,
      "step": 750
    },
    {
      "epoch": 0.36226234031068194,
      "grad_norm": 4.75,
      "learning_rate": 4.0245202558635396e-06,
      "loss": 1.0052,
      "step": 755
    },
    {
      "epoch": 0.3646614286571103,
      "grad_norm": 3.46875,
      "learning_rate": 4.051172707889126e-06,
      "loss": 1.0548,
      "step": 760
    },
    {
      "epoch": 0.36706051700353864,
      "grad_norm": 7.625,
      "learning_rate": 4.077825159914712e-06,
      "loss": 1.127,
      "step": 765
    },
    {
      "epoch": 0.369459605349967,
      "grad_norm": 4.65625,
      "learning_rate": 4.104477611940299e-06,
      "loss": 0.9835,
      "step": 770
    },
    {
      "epoch": 0.37185869369639535,
      "grad_norm": 3.828125,
      "learning_rate": 4.131130063965885e-06,
      "loss": 1.1199,
      "step": 775
    },
    {
      "epoch": 0.37425778204282373,
      "grad_norm": 5.3125,
      "learning_rate": 4.157782515991471e-06,
      "loss": 1.0222,
      "step": 780
    },
    {
      "epoch": 0.3766568703892521,
      "grad_norm": 4.90625,
      "learning_rate": 4.184434968017058e-06,
      "loss": 1.1066,
      "step": 785
    },
    {
      "epoch": 0.37905595873568043,
      "grad_norm": 3.515625,
      "learning_rate": 4.211087420042645e-06,
      "loss": 1.0559,
      "step": 790
    },
    {
      "epoch": 0.3814550470821088,
      "grad_norm": 6.3125,
      "learning_rate": 4.237739872068231e-06,
      "loss": 0.974,
      "step": 795
    },
    {
      "epoch": 0.38385413542853714,
      "grad_norm": 8.0625,
      "learning_rate": 4.264392324093816e-06,
      "loss": 1.0657,
      "step": 800
    },
    {
      "epoch": 0.38385413542853714,
      "eval_loss": 1.0770481824874878,
      "eval_runtime": 177.0808,
      "eval_samples_per_second": 41.845,
      "eval_steps_per_second": 10.464,
      "step": 800
    },
    {
      "epoch": 0.3862532237749655,
      "grad_norm": 3.8125,
      "learning_rate": 4.2910447761194036e-06,
      "loss": 1.0495,
      "step": 805
    },
    {
      "epoch": 0.3886523121213939,
      "grad_norm": 4.21875,
      "learning_rate": 4.31769722814499e-06,
      "loss": 1.0879,
      "step": 810
    },
    {
      "epoch": 0.3910514004678222,
      "grad_norm": 8.75,
      "learning_rate": 4.344349680170576e-06,
      "loss": 1.0887,
      "step": 815
    },
    {
      "epoch": 0.3934504888142506,
      "grad_norm": 4.6875,
      "learning_rate": 4.3710021321961625e-06,
      "loss": 1.0342,
      "step": 820
    },
    {
      "epoch": 0.3958495771606789,
      "grad_norm": 5.21875,
      "learning_rate": 4.397654584221749e-06,
      "loss": 1.0058,
      "step": 825
    },
    {
      "epoch": 0.3982486655071073,
      "grad_norm": 6.6875,
      "learning_rate": 4.424307036247335e-06,
      "loss": 0.9424,
      "step": 830
    },
    {
      "epoch": 0.4006477538535357,
      "grad_norm": 8.125,
      "learning_rate": 4.4509594882729215e-06,
      "loss": 1.0973,
      "step": 835
    },
    {
      "epoch": 0.403046842199964,
      "grad_norm": 5.3125,
      "learning_rate": 4.477611940298508e-06,
      "loss": 1.1379,
      "step": 840
    },
    {
      "epoch": 0.4054459305463924,
      "grad_norm": 4.28125,
      "learning_rate": 4.504264392324094e-06,
      "loss": 1.0758,
      "step": 845
    },
    {
      "epoch": 0.4078450188928207,
      "grad_norm": 4.625,
      "learning_rate": 4.53091684434968e-06,
      "loss": 1.1237,
      "step": 850
    },
    {
      "epoch": 0.4102441072392491,
      "grad_norm": 3.859375,
      "learning_rate": 4.557569296375267e-06,
      "loss": 1.083,
      "step": 855
    },
    {
      "epoch": 0.41264319558567747,
      "grad_norm": 3.84375,
      "learning_rate": 4.584221748400853e-06,
      "loss": 1.0514,
      "step": 860
    },
    {
      "epoch": 0.4150422839321058,
      "grad_norm": 3.921875,
      "learning_rate": 4.610874200426439e-06,
      "loss": 1.0582,
      "step": 865
    },
    {
      "epoch": 0.4174413722785342,
      "grad_norm": 5.15625,
      "learning_rate": 4.637526652452026e-06,
      "loss": 0.9442,
      "step": 870
    },
    {
      "epoch": 0.4198404606249625,
      "grad_norm": 6.03125,
      "learning_rate": 4.664179104477613e-06,
      "loss": 0.9391,
      "step": 875
    },
    {
      "epoch": 0.4222395489713909,
      "grad_norm": 3.546875,
      "learning_rate": 4.690831556503198e-06,
      "loss": 1.016,
      "step": 880
    },
    {
      "epoch": 0.42463863731781926,
      "grad_norm": 4.9375,
      "learning_rate": 4.717484008528785e-06,
      "loss": 1.0797,
      "step": 885
    },
    {
      "epoch": 0.4270377256642476,
      "grad_norm": 6.78125,
      "learning_rate": 4.744136460554372e-06,
      "loss": 1.0632,
      "step": 890
    },
    {
      "epoch": 0.42943681401067596,
      "grad_norm": 4.28125,
      "learning_rate": 4.770788912579958e-06,
      "loss": 1.0721,
      "step": 895
    },
    {
      "epoch": 0.4318359023571043,
      "grad_norm": 4.59375,
      "learning_rate": 4.797441364605544e-06,
      "loss": 1.054,
      "step": 900
    },
    {
      "epoch": 0.4318359023571043,
      "eval_loss": 1.0616637468338013,
      "eval_runtime": 175.5614,
      "eval_samples_per_second": 42.207,
      "eval_steps_per_second": 10.555,
      "step": 900
    },
    {
      "epoch": 0.43423499070353266,
      "grad_norm": 5.4375,
      "learning_rate": 4.82409381663113e-06,
      "loss": 1.0619,
      "step": 905
    },
    {
      "epoch": 0.436634079049961,
      "grad_norm": 4.40625,
      "learning_rate": 4.850746268656717e-06,
      "loss": 0.9793,
      "step": 910
    },
    {
      "epoch": 0.43903316739638937,
      "grad_norm": 4.78125,
      "learning_rate": 4.877398720682303e-06,
      "loss": 1.031,
      "step": 915
    },
    {
      "epoch": 0.44143225574281775,
      "grad_norm": 3.953125,
      "learning_rate": 4.90405117270789e-06,
      "loss": 1.0687,
      "step": 920
    },
    {
      "epoch": 0.44383134408924607,
      "grad_norm": 3.59375,
      "learning_rate": 4.930703624733476e-06,
      "loss": 1.0311,
      "step": 925
    },
    {
      "epoch": 0.44623043243567445,
      "grad_norm": 3.265625,
      "learning_rate": 4.957356076759062e-06,
      "loss": 0.9699,
      "step": 930
    },
    {
      "epoch": 0.4486295207821028,
      "grad_norm": 4.78125,
      "learning_rate": 4.984008528784649e-06,
      "loss": 1.1171,
      "step": 935
    },
    {
      "epoch": 0.45102860912853115,
      "grad_norm": 4.0625,
      "learning_rate": 5.010660980810235e-06,
      "loss": 0.9621,
      "step": 940
    },
    {
      "epoch": 0.45342769747495953,
      "grad_norm": 3.84375,
      "learning_rate": 5.037313432835821e-06,
      "loss": 1.0444,
      "step": 945
    },
    {
      "epoch": 0.45582678582138786,
      "grad_norm": 5.0,
      "learning_rate": 5.063965884861408e-06,
      "loss": 0.9907,
      "step": 950
    },
    {
      "epoch": 0.45822587416781624,
      "grad_norm": 3.578125,
      "learning_rate": 5.090618336886995e-06,
      "loss": 1.0314,
      "step": 955
    },
    {
      "epoch": 0.46062496251424456,
      "grad_norm": 3.78125,
      "learning_rate": 5.11727078891258e-06,
      "loss": 1.0468,
      "step": 960
    },
    {
      "epoch": 0.46302405086067294,
      "grad_norm": 4.625,
      "learning_rate": 5.1439232409381665e-06,
      "loss": 1.0273,
      "step": 965
    },
    {
      "epoch": 0.4654231392071013,
      "grad_norm": 3.546875,
      "learning_rate": 5.170575692963753e-06,
      "loss": 1.0078,
      "step": 970
    },
    {
      "epoch": 0.46782222755352965,
      "grad_norm": 4.25,
      "learning_rate": 5.197228144989339e-06,
      "loss": 0.9977,
      "step": 975
    },
    {
      "epoch": 0.470221315899958,
      "grad_norm": 5.0625,
      "learning_rate": 5.2238805970149255e-06,
      "loss": 0.9653,
      "step": 980
    },
    {
      "epoch": 0.47262040424638635,
      "grad_norm": 5.0,
      "learning_rate": 5.250533049040513e-06,
      "loss": 1.0362,
      "step": 985
    },
    {
      "epoch": 0.47501949259281473,
      "grad_norm": 3.765625,
      "learning_rate": 5.277185501066099e-06,
      "loss": 1.0136,
      "step": 990
    },
    {
      "epoch": 0.4774185809392431,
      "grad_norm": 3.015625,
      "learning_rate": 5.303837953091685e-06,
      "loss": 0.9821,
      "step": 995
    },
    {
      "epoch": 0.47981766928567143,
      "grad_norm": 3.40625,
      "learning_rate": 5.3304904051172716e-06,
      "loss": 1.0744,
      "step": 1000
    },
    {
      "epoch": 0.47981766928567143,
      "eval_loss": 1.048653483390808,
      "eval_runtime": 175.6017,
      "eval_samples_per_second": 42.198,
      "eval_steps_per_second": 10.552,
      "step": 1000
    },
    {
      "epoch": 0.4822167576320998,
      "grad_norm": 3.859375,
      "learning_rate": 5.357142857142857e-06,
      "loss": 1.0825,
      "step": 1005
    },
    {
      "epoch": 0.48461584597852814,
      "grad_norm": 3.359375,
      "learning_rate": 5.383795309168443e-06,
      "loss": 1.0803,
      "step": 1010
    },
    {
      "epoch": 0.4870149343249565,
      "grad_norm": 5.53125,
      "learning_rate": 5.41044776119403e-06,
      "loss": 0.9915,
      "step": 1015
    },
    {
      "epoch": 0.4894140226713849,
      "grad_norm": 5.5,
      "learning_rate": 5.437100213219617e-06,
      "loss": 1.0955,
      "step": 1020
    },
    {
      "epoch": 0.4918131110178132,
      "grad_norm": 3.15625,
      "learning_rate": 5.463752665245203e-06,
      "loss": 0.9685,
      "step": 1025
    },
    {
      "epoch": 0.4942121993642416,
      "grad_norm": 4.65625,
      "learning_rate": 5.4904051172707895e-06,
      "loss": 1.0203,
      "step": 1030
    },
    {
      "epoch": 0.4966112877106699,
      "grad_norm": 5.34375,
      "learning_rate": 5.517057569296376e-06,
      "loss": 1.1175,
      "step": 1035
    },
    {
      "epoch": 0.4990103760570983,
      "grad_norm": 5.09375,
      "learning_rate": 5.543710021321962e-06,
      "loss": 1.0268,
      "step": 1040
    },
    {
      "epoch": 0.5014094644035266,
      "grad_norm": 4.09375,
      "learning_rate": 5.570362473347548e-06,
      "loss": 0.9042,
      "step": 1045
    },
    {
      "epoch": 0.5038085527499551,
      "grad_norm": 4.34375,
      "learning_rate": 5.597014925373134e-06,
      "loss": 1.0338,
      "step": 1050
    },
    {
      "epoch": 0.5062076410963834,
      "grad_norm": 3.390625,
      "learning_rate": 5.623667377398722e-06,
      "loss": 1.0724,
      "step": 1055
    },
    {
      "epoch": 0.5086067294428117,
      "grad_norm": 3.703125,
      "learning_rate": 5.650319829424308e-06,
      "loss": 1.1108,
      "step": 1060
    },
    {
      "epoch": 0.51100581778924,
      "grad_norm": 3.4375,
      "learning_rate": 5.676972281449894e-06,
      "loss": 1.0471,
      "step": 1065
    },
    {
      "epoch": 0.5134049061356685,
      "grad_norm": 2.875,
      "learning_rate": 5.70362473347548e-06,
      "loss": 1.0467,
      "step": 1070
    },
    {
      "epoch": 0.5158039944820968,
      "grad_norm": 4.28125,
      "learning_rate": 5.730277185501066e-06,
      "loss": 1.0706,
      "step": 1075
    },
    {
      "epoch": 0.5182030828285251,
      "grad_norm": 3.59375,
      "learning_rate": 5.756929637526653e-06,
      "loss": 1.0565,
      "step": 1080
    },
    {
      "epoch": 0.5206021711749536,
      "grad_norm": 4.0625,
      "learning_rate": 5.783582089552239e-06,
      "loss": 0.9996,
      "step": 1085
    },
    {
      "epoch": 0.5230012595213819,
      "grad_norm": 4.15625,
      "learning_rate": 5.810234541577826e-06,
      "loss": 1.0384,
      "step": 1090
    },
    {
      "epoch": 0.5254003478678102,
      "grad_norm": 3.875,
      "learning_rate": 5.836886993603412e-06,
      "loss": 1.0247,
      "step": 1095
    },
    {
      "epoch": 0.5277994362142386,
      "grad_norm": 4.0,
      "learning_rate": 5.863539445628999e-06,
      "loss": 0.9977,
      "step": 1100
    },
    {
      "epoch": 0.5277994362142386,
      "eval_loss": 1.0383222103118896,
      "eval_runtime": 175.7302,
      "eval_samples_per_second": 42.167,
      "eval_steps_per_second": 10.545,
      "step": 1100
    },
    {
      "epoch": 0.530198524560667,
      "grad_norm": 3.421875,
      "learning_rate": 5.890191897654585e-06,
      "loss": 1.0028,
      "step": 1105
    },
    {
      "epoch": 0.5325976129070953,
      "grad_norm": 3.3125,
      "learning_rate": 5.9168443496801705e-06,
      "loss": 1.0242,
      "step": 1110
    },
    {
      "epoch": 0.5349967012535236,
      "grad_norm": 3.359375,
      "learning_rate": 5.943496801705757e-06,
      "loss": 1.0978,
      "step": 1115
    },
    {
      "epoch": 0.537395789599952,
      "grad_norm": 3.609375,
      "learning_rate": 5.970149253731343e-06,
      "loss": 0.9951,
      "step": 1120
    },
    {
      "epoch": 0.5397948779463804,
      "grad_norm": 3.65625,
      "learning_rate": 5.99680170575693e-06,
      "loss": 1.0309,
      "step": 1125
    },
    {
      "epoch": 0.5421939662928087,
      "grad_norm": 4.375,
      "learning_rate": 6.023454157782517e-06,
      "loss": 1.0587,
      "step": 1130
    },
    {
      "epoch": 0.5445930546392371,
      "grad_norm": 4.03125,
      "learning_rate": 6.050106609808103e-06,
      "loss": 1.0184,
      "step": 1135
    },
    {
      "epoch": 0.5469921429856655,
      "grad_norm": 3.984375,
      "learning_rate": 6.076759061833689e-06,
      "loss": 1.0146,
      "step": 1140
    },
    {
      "epoch": 0.5493912313320938,
      "grad_norm": 4.03125,
      "learning_rate": 6.1034115138592756e-06,
      "loss": 1.1346,
      "step": 1145
    },
    {
      "epoch": 0.5517903196785222,
      "grad_norm": 3.28125,
      "learning_rate": 6.130063965884862e-06,
      "loss": 0.9866,
      "step": 1150
    },
    {
      "epoch": 0.5541894080249505,
      "grad_norm": 5.21875,
      "learning_rate": 6.156716417910447e-06,
      "loss": 1.0329,
      "step": 1155
    },
    {
      "epoch": 0.5565884963713789,
      "grad_norm": 4.1875,
      "learning_rate": 6.183368869936035e-06,
      "loss": 1.0053,
      "step": 1160
    },
    {
      "epoch": 0.5589875847178072,
      "grad_norm": 3.21875,
      "learning_rate": 6.210021321961621e-06,
      "loss": 1.0875,
      "step": 1165
    },
    {
      "epoch": 0.5613866730642356,
      "grad_norm": 3.84375,
      "learning_rate": 6.236673773987207e-06,
      "loss": 1.007,
      "step": 1170
    },
    {
      "epoch": 0.5637857614106639,
      "grad_norm": 3.59375,
      "learning_rate": 6.2633262260127935e-06,
      "loss": 0.982,
      "step": 1175
    },
    {
      "epoch": 0.5661848497570923,
      "grad_norm": 3.5,
      "learning_rate": 6.28997867803838e-06,
      "loss": 0.9737,
      "step": 1180
    },
    {
      "epoch": 0.5685839381035207,
      "grad_norm": 3.578125,
      "learning_rate": 6.316631130063966e-06,
      "loss": 1.0252,
      "step": 1185
    },
    {
      "epoch": 0.570983026449949,
      "grad_norm": 4.15625,
      "learning_rate": 6.343283582089553e-06,
      "loss": 1.1008,
      "step": 1190
    },
    {
      "epoch": 0.5733821147963774,
      "grad_norm": 2.734375,
      "learning_rate": 6.3699360341151396e-06,
      "loss": 1.1048,
      "step": 1195
    },
    {
      "epoch": 0.5757812031428058,
      "grad_norm": 4.40625,
      "learning_rate": 6.396588486140726e-06,
      "loss": 0.9778,
      "step": 1200
    },
    {
      "epoch": 0.5757812031428058,
      "eval_loss": 1.0289998054504395,
      "eval_runtime": 175.7204,
      "eval_samples_per_second": 42.169,
      "eval_steps_per_second": 10.545,
      "step": 1200
    },
    {
      "epoch": 0.5781802914892341,
      "grad_norm": 2.796875,
      "learning_rate": 6.423240938166312e-06,
      "loss": 0.9643,
      "step": 1205
    },
    {
      "epoch": 0.5805793798356624,
      "grad_norm": 3.03125,
      "learning_rate": 6.449893390191898e-06,
      "loss": 0.9933,
      "step": 1210
    },
    {
      "epoch": 0.5829784681820908,
      "grad_norm": 3.6875,
      "learning_rate": 6.476545842217484e-06,
      "loss": 1.0294,
      "step": 1215
    },
    {
      "epoch": 0.5853775565285192,
      "grad_norm": 3.984375,
      "learning_rate": 6.50319829424307e-06,
      "loss": 1.0155,
      "step": 1220
    },
    {
      "epoch": 0.5877766448749475,
      "grad_norm": 3.203125,
      "learning_rate": 6.5298507462686575e-06,
      "loss": 1.0275,
      "step": 1225
    },
    {
      "epoch": 0.5901757332213758,
      "grad_norm": 5.78125,
      "learning_rate": 6.556503198294244e-06,
      "loss": 1.0395,
      "step": 1230
    },
    {
      "epoch": 0.5925748215678043,
      "grad_norm": 3.59375,
      "learning_rate": 6.58315565031983e-06,
      "loss": 1.0462,
      "step": 1235
    },
    {
      "epoch": 0.5949739099142326,
      "grad_norm": 3.84375,
      "learning_rate": 6.609808102345416e-06,
      "loss": 1.0042,
      "step": 1240
    },
    {
      "epoch": 0.5973729982606609,
      "grad_norm": 4.0625,
      "learning_rate": 6.636460554371003e-06,
      "loss": 0.9548,
      "step": 1245
    },
    {
      "epoch": 0.5997720866070894,
      "grad_norm": 3.609375,
      "learning_rate": 6.663113006396589e-06,
      "loss": 1.0012,
      "step": 1250
    },
    {
      "epoch": 0.6021711749535177,
      "grad_norm": 3.046875,
      "learning_rate": 6.689765458422175e-06,
      "loss": 0.8987,
      "step": 1255
    },
    {
      "epoch": 0.604570263299946,
      "grad_norm": 4.3125,
      "learning_rate": 6.7164179104477625e-06,
      "loss": 1.0123,
      "step": 1260
    },
    {
      "epoch": 0.6069693516463743,
      "grad_norm": 4.40625,
      "learning_rate": 6.743070362473349e-06,
      "loss": 0.9884,
      "step": 1265
    },
    {
      "epoch": 0.6093684399928028,
      "grad_norm": 3.140625,
      "learning_rate": 6.769722814498934e-06,
      "loss": 1.1241,
      "step": 1270
    },
    {
      "epoch": 0.6117675283392311,
      "grad_norm": 4.3125,
      "learning_rate": 6.796375266524521e-06,
      "loss": 1.0129,
      "step": 1275
    },
    {
      "epoch": 0.6141666166856594,
      "grad_norm": 3.96875,
      "learning_rate": 6.823027718550107e-06,
      "loss": 1.0729,
      "step": 1280
    },
    {
      "epoch": 0.6165657050320879,
      "grad_norm": 4.0625,
      "learning_rate": 6.849680170575693e-06,
      "loss": 1.0007,
      "step": 1285
    },
    {
      "epoch": 0.6189647933785162,
      "grad_norm": 4.65625,
      "learning_rate": 6.8763326226012796e-06,
      "loss": 0.9614,
      "step": 1290
    },
    {
      "epoch": 0.6213638817249445,
      "grad_norm": 3.765625,
      "learning_rate": 6.902985074626867e-06,
      "loss": 0.9357,
      "step": 1295
    },
    {
      "epoch": 0.6237629700713728,
      "grad_norm": 2.9375,
      "learning_rate": 6.929637526652453e-06,
      "loss": 1.0187,
      "step": 1300
    },
    {
      "epoch": 0.6237629700713728,
      "eval_loss": 1.0210597515106201,
      "eval_runtime": 175.7208,
      "eval_samples_per_second": 42.169,
      "eval_steps_per_second": 10.545,
      "step": 1300
    },
    {
      "epoch": 0.6261620584178013,
      "grad_norm": 3.40625,
      "learning_rate": 6.956289978678039e-06,
      "loss": 0.9008,
      "step": 1305
    },
    {
      "epoch": 0.6285611467642296,
      "grad_norm": 3.0625,
      "learning_rate": 6.982942430703626e-06,
      "loss": 1.0302,
      "step": 1310
    },
    {
      "epoch": 0.6309602351106579,
      "grad_norm": 3.0,
      "learning_rate": 7.009594882729211e-06,
      "loss": 0.9546,
      "step": 1315
    },
    {
      "epoch": 0.6333593234570863,
      "grad_norm": 3.78125,
      "learning_rate": 7.0362473347547975e-06,
      "loss": 1.0431,
      "step": 1320
    },
    {
      "epoch": 0.6357584118035147,
      "grad_norm": 3.84375,
      "learning_rate": 7.062899786780384e-06,
      "loss": 0.9423,
      "step": 1325
    },
    {
      "epoch": 0.638157500149943,
      "grad_norm": 3.890625,
      "learning_rate": 7.089552238805971e-06,
      "loss": 1.0365,
      "step": 1330
    },
    {
      "epoch": 0.6405565884963714,
      "grad_norm": 3.390625,
      "learning_rate": 7.116204690831557e-06,
      "loss": 0.973,
      "step": 1335
    },
    {
      "epoch": 0.6429556768427998,
      "grad_norm": 3.34375,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.8795,
      "step": 1340
    },
    {
      "epoch": 0.6453547651892281,
      "grad_norm": 3.359375,
      "learning_rate": 7.16950959488273e-06,
      "loss": 0.9596,
      "step": 1345
    },
    {
      "epoch": 0.6477538535356564,
      "grad_norm": 3.734375,
      "learning_rate": 7.196162046908316e-06,
      "loss": 0.9984,
      "step": 1350
    },
    {
      "epoch": 0.6501529418820848,
      "grad_norm": 4.09375,
      "learning_rate": 7.2228144989339025e-06,
      "loss": 0.9752,
      "step": 1355
    },
    {
      "epoch": 0.6525520302285132,
      "grad_norm": 5.59375,
      "learning_rate": 7.249466950959488e-06,
      "loss": 0.9701,
      "step": 1360
    },
    {
      "epoch": 0.6549511185749415,
      "grad_norm": 3.40625,
      "learning_rate": 7.276119402985076e-06,
      "loss": 1.0054,
      "step": 1365
    },
    {
      "epoch": 0.6573502069213699,
      "grad_norm": 3.25,
      "learning_rate": 7.302771855010662e-06,
      "loss": 0.9919,
      "step": 1370
    },
    {
      "epoch": 0.6597492952677982,
      "grad_norm": 3.515625,
      "learning_rate": 7.329424307036248e-06,
      "loss": 0.9389,
      "step": 1375
    },
    {
      "epoch": 0.6621483836142266,
      "grad_norm": 3.953125,
      "learning_rate": 7.356076759061834e-06,
      "loss": 0.9607,
      "step": 1380
    },
    {
      "epoch": 0.664547471960655,
      "grad_norm": 3.046875,
      "learning_rate": 7.38272921108742e-06,
      "loss": 1.084,
      "step": 1385
    },
    {
      "epoch": 0.6669465603070833,
      "grad_norm": 3.203125,
      "learning_rate": 7.409381663113007e-06,
      "loss": 0.9887,
      "step": 1390
    },
    {
      "epoch": 0.6693456486535116,
      "grad_norm": 3.953125,
      "learning_rate": 7.436034115138593e-06,
      "loss": 0.9656,
      "step": 1395
    },
    {
      "epoch": 0.67174473699994,
      "grad_norm": 2.46875,
      "learning_rate": 7.46268656716418e-06,
      "loss": 1.085,
      "step": 1400
    },
    {
      "epoch": 0.67174473699994,
      "eval_loss": 1.0130821466445923,
      "eval_runtime": 178.5199,
      "eval_samples_per_second": 41.508,
      "eval_steps_per_second": 10.38,
      "step": 1400
    },
    {
      "epoch": 0.6741438253463684,
      "grad_norm": 3.546875,
      "learning_rate": 7.4893390191897665e-06,
      "loss": 1.0731,
      "step": 1405
    },
    {
      "epoch": 0.6765429136927967,
      "grad_norm": 3.125,
      "learning_rate": 7.515991471215353e-06,
      "loss": 1.0096,
      "step": 1410
    },
    {
      "epoch": 0.6789420020392251,
      "grad_norm": 2.828125,
      "learning_rate": 7.542643923240939e-06,
      "loss": 0.9554,
      "step": 1415
    },
    {
      "epoch": 0.6813410903856535,
      "grad_norm": 3.09375,
      "learning_rate": 7.569296375266525e-06,
      "loss": 1.0688,
      "step": 1420
    },
    {
      "epoch": 0.6837401787320818,
      "grad_norm": 3.359375,
      "learning_rate": 7.595948827292111e-06,
      "loss": 0.9863,
      "step": 1425
    },
    {
      "epoch": 0.6861392670785101,
      "grad_norm": 4.3125,
      "learning_rate": 7.622601279317697e-06,
      "loss": 0.9846,
      "step": 1430
    },
    {
      "epoch": 0.6885383554249386,
      "grad_norm": 3.578125,
      "learning_rate": 7.649253731343284e-06,
      "loss": 0.9898,
      "step": 1435
    },
    {
      "epoch": 0.6909374437713669,
      "grad_norm": 3.53125,
      "learning_rate": 7.67590618336887e-06,
      "loss": 1.0163,
      "step": 1440
    },
    {
      "epoch": 0.6933365321177952,
      "grad_norm": 3.25,
      "learning_rate": 7.702558635394457e-06,
      "loss": 1.0098,
      "step": 1445
    },
    {
      "epoch": 0.6957356204642235,
      "grad_norm": 3.359375,
      "learning_rate": 7.729211087420043e-06,
      "loss": 0.9562,
      "step": 1450
    },
    {
      "epoch": 0.698134708810652,
      "grad_norm": 3.578125,
      "learning_rate": 7.75586353944563e-06,
      "loss": 0.9638,
      "step": 1455
    },
    {
      "epoch": 0.7005337971570803,
      "grad_norm": 3.421875,
      "learning_rate": 7.782515991471216e-06,
      "loss": 0.9566,
      "step": 1460
    },
    {
      "epoch": 0.7029328855035086,
      "grad_norm": 3.609375,
      "learning_rate": 7.809168443496802e-06,
      "loss": 1.0047,
      "step": 1465
    },
    {
      "epoch": 0.7053319738499371,
      "grad_norm": 5.0,
      "learning_rate": 7.835820895522389e-06,
      "loss": 0.9726,
      "step": 1470
    },
    {
      "epoch": 0.7077310621963654,
      "grad_norm": 3.53125,
      "learning_rate": 7.862473347547975e-06,
      "loss": 0.951,
      "step": 1475
    },
    {
      "epoch": 0.7101301505427937,
      "grad_norm": 2.8125,
      "learning_rate": 7.889125799573561e-06,
      "loss": 0.9984,
      "step": 1480
    },
    {
      "epoch": 0.7125292388892221,
      "grad_norm": 3.3125,
      "learning_rate": 7.915778251599148e-06,
      "loss": 0.9609,
      "step": 1485
    },
    {
      "epoch": 0.7149283272356505,
      "grad_norm": 3.1875,
      "learning_rate": 7.942430703624734e-06,
      "loss": 0.915,
      "step": 1490
    },
    {
      "epoch": 0.7173274155820788,
      "grad_norm": 3.25,
      "learning_rate": 7.96908315565032e-06,
      "loss": 1.0487,
      "step": 1495
    },
    {
      "epoch": 0.7197265039285071,
      "grad_norm": 3.296875,
      "learning_rate": 7.995735607675907e-06,
      "loss": 0.958,
      "step": 1500
    },
    {
      "epoch": 0.7197265039285071,
      "eval_loss": 1.0071519613265991,
      "eval_runtime": 175.8655,
      "eval_samples_per_second": 42.134,
      "eval_steps_per_second": 10.536,
      "step": 1500
    },
    {
      "epoch": 0.7221255922749356,
      "grad_norm": 3.578125,
      "learning_rate": 8.022388059701493e-06,
      "loss": 1.0321,
      "step": 1505
    },
    {
      "epoch": 0.7245246806213639,
      "grad_norm": 3.625,
      "learning_rate": 8.049040511727079e-06,
      "loss": 0.9943,
      "step": 1510
    },
    {
      "epoch": 0.7269237689677922,
      "grad_norm": 2.78125,
      "learning_rate": 8.075692963752665e-06,
      "loss": 0.9148,
      "step": 1515
    },
    {
      "epoch": 0.7293228573142206,
      "grad_norm": 3.625,
      "learning_rate": 8.102345415778252e-06,
      "loss": 0.9026,
      "step": 1520
    },
    {
      "epoch": 0.731721945660649,
      "grad_norm": 3.15625,
      "learning_rate": 8.128997867803838e-06,
      "loss": 1.023,
      "step": 1525
    },
    {
      "epoch": 0.7341210340070773,
      "grad_norm": 3.671875,
      "learning_rate": 8.155650319829424e-06,
      "loss": 1.0484,
      "step": 1530
    },
    {
      "epoch": 0.7365201223535056,
      "grad_norm": 3.078125,
      "learning_rate": 8.182302771855012e-06,
      "loss": 1.0762,
      "step": 1535
    },
    {
      "epoch": 0.738919210699934,
      "grad_norm": 3.890625,
      "learning_rate": 8.208955223880599e-06,
      "loss": 1.0513,
      "step": 1540
    },
    {
      "epoch": 0.7413182990463624,
      "grad_norm": 3.515625,
      "learning_rate": 8.235607675906185e-06,
      "loss": 0.9738,
      "step": 1545
    },
    {
      "epoch": 0.7437173873927907,
      "grad_norm": 4.3125,
      "learning_rate": 8.26226012793177e-06,
      "loss": 1.0299,
      "step": 1550
    },
    {
      "epoch": 0.7461164757392191,
      "grad_norm": 3.28125,
      "learning_rate": 8.288912579957356e-06,
      "loss": 0.9971,
      "step": 1555
    },
    {
      "epoch": 0.7485155640856475,
      "grad_norm": 5.75,
      "learning_rate": 8.315565031982942e-06,
      "loss": 0.9628,
      "step": 1560
    },
    {
      "epoch": 0.7509146524320758,
      "grad_norm": 4.0625,
      "learning_rate": 8.342217484008529e-06,
      "loss": 1.0038,
      "step": 1565
    },
    {
      "epoch": 0.7533137407785042,
      "grad_norm": 3.828125,
      "learning_rate": 8.368869936034117e-06,
      "loss": 0.9872,
      "step": 1570
    },
    {
      "epoch": 0.7557128291249325,
      "grad_norm": 3.640625,
      "learning_rate": 8.395522388059703e-06,
      "loss": 0.9728,
      "step": 1575
    },
    {
      "epoch": 0.7581119174713609,
      "grad_norm": 3.03125,
      "learning_rate": 8.42217484008529e-06,
      "loss": 0.9373,
      "step": 1580
    },
    {
      "epoch": 0.7605110058177892,
      "grad_norm": 3.65625,
      "learning_rate": 8.448827292110876e-06,
      "loss": 0.9553,
      "step": 1585
    },
    {
      "epoch": 0.7629100941642176,
      "grad_norm": 3.09375,
      "learning_rate": 8.475479744136462e-06,
      "loss": 0.8991,
      "step": 1590
    },
    {
      "epoch": 0.765309182510646,
      "grad_norm": 3.125,
      "learning_rate": 8.502132196162046e-06,
      "loss": 1.0188,
      "step": 1595
    },
    {
      "epoch": 0.7677082708570743,
      "grad_norm": 3.53125,
      "learning_rate": 8.528784648187633e-06,
      "loss": 1.0482,
      "step": 1600
    },
    {
      "epoch": 0.7677082708570743,
      "eval_loss": 1.0007458925247192,
      "eval_runtime": 175.7117,
      "eval_samples_per_second": 42.171,
      "eval_steps_per_second": 10.546,
      "step": 1600
    },
    {
      "epoch": 0.7701073592035027,
      "grad_norm": 2.75,
      "learning_rate": 8.55543710021322e-06,
      "loss": 0.9521,
      "step": 1605
    },
    {
      "epoch": 0.772506447549931,
      "grad_norm": 2.875,
      "learning_rate": 8.582089552238807e-06,
      "loss": 0.9527,
      "step": 1610
    },
    {
      "epoch": 0.7749055358963594,
      "grad_norm": 3.046875,
      "learning_rate": 8.608742004264393e-06,
      "loss": 1.0012,
      "step": 1615
    },
    {
      "epoch": 0.7773046242427878,
      "grad_norm": 3.203125,
      "learning_rate": 8.63539445628998e-06,
      "loss": 0.9729,
      "step": 1620
    },
    {
      "epoch": 0.7797037125892161,
      "grad_norm": 4.1875,
      "learning_rate": 8.662046908315566e-06,
      "loss": 0.9014,
      "step": 1625
    },
    {
      "epoch": 0.7821028009356444,
      "grad_norm": 3.03125,
      "learning_rate": 8.688699360341152e-06,
      "loss": 1.0124,
      "step": 1630
    },
    {
      "epoch": 0.7845018892820728,
      "grad_norm": 3.078125,
      "learning_rate": 8.715351812366739e-06,
      "loss": 0.9066,
      "step": 1635
    },
    {
      "epoch": 0.7869009776285012,
      "grad_norm": 3.421875,
      "learning_rate": 8.742004264392325e-06,
      "loss": 0.9101,
      "step": 1640
    },
    {
      "epoch": 0.7893000659749295,
      "grad_norm": 3.796875,
      "learning_rate": 8.768656716417911e-06,
      "loss": 0.9386,
      "step": 1645
    },
    {
      "epoch": 0.7916991543213578,
      "grad_norm": 3.6875,
      "learning_rate": 8.795309168443498e-06,
      "loss": 0.9139,
      "step": 1650
    },
    {
      "epoch": 0.7940982426677863,
      "grad_norm": 3.40625,
      "learning_rate": 8.821961620469084e-06,
      "loss": 0.9178,
      "step": 1655
    },
    {
      "epoch": 0.7964973310142146,
      "grad_norm": 3.203125,
      "learning_rate": 8.84861407249467e-06,
      "loss": 0.9809,
      "step": 1660
    },
    {
      "epoch": 0.7988964193606429,
      "grad_norm": 3.34375,
      "learning_rate": 8.875266524520257e-06,
      "loss": 1.0115,
      "step": 1665
    },
    {
      "epoch": 0.8012955077070714,
      "grad_norm": 2.890625,
      "learning_rate": 8.901918976545843e-06,
      "loss": 0.9548,
      "step": 1670
    },
    {
      "epoch": 0.8036945960534997,
      "grad_norm": 2.625,
      "learning_rate": 8.92857142857143e-06,
      "loss": 0.9575,
      "step": 1675
    },
    {
      "epoch": 0.806093684399928,
      "grad_norm": 3.203125,
      "learning_rate": 8.955223880597016e-06,
      "loss": 1.0284,
      "step": 1680
    },
    {
      "epoch": 0.8084927727463563,
      "grad_norm": 3.265625,
      "learning_rate": 8.981876332622602e-06,
      "loss": 0.9645,
      "step": 1685
    },
    {
      "epoch": 0.8108918610927848,
      "grad_norm": 4.09375,
      "learning_rate": 9.008528784648188e-06,
      "loss": 0.9457,
      "step": 1690
    },
    {
      "epoch": 0.8132909494392131,
      "grad_norm": 4.21875,
      "learning_rate": 9.035181236673775e-06,
      "loss": 0.9795,
      "step": 1695
    },
    {
      "epoch": 0.8156900377856414,
      "grad_norm": 4.09375,
      "learning_rate": 9.06183368869936e-06,
      "loss": 0.9447,
      "step": 1700
    },
    {
      "epoch": 0.8156900377856414,
      "eval_loss": 0.9945608377456665,
      "eval_runtime": 175.6059,
      "eval_samples_per_second": 42.197,
      "eval_steps_per_second": 10.552,
      "step": 1700
    },
    {
      "epoch": 0.8180891261320699,
      "grad_norm": 3.765625,
      "learning_rate": 9.088486140724947e-06,
      "loss": 1.0695,
      "step": 1705
    },
    {
      "epoch": 0.8204882144784982,
      "grad_norm": 3.15625,
      "learning_rate": 9.115138592750533e-06,
      "loss": 0.9614,
      "step": 1710
    },
    {
      "epoch": 0.8228873028249265,
      "grad_norm": 3.734375,
      "learning_rate": 9.14179104477612e-06,
      "loss": 1.055,
      "step": 1715
    },
    {
      "epoch": 0.8252863911713549,
      "grad_norm": 3.25,
      "learning_rate": 9.168443496801706e-06,
      "loss": 0.9108,
      "step": 1720
    },
    {
      "epoch": 0.8276854795177833,
      "grad_norm": 5.09375,
      "learning_rate": 9.195095948827292e-06,
      "loss": 0.9487,
      "step": 1725
    },
    {
      "epoch": 0.8300845678642116,
      "grad_norm": 3.109375,
      "learning_rate": 9.221748400852879e-06,
      "loss": 0.9393,
      "step": 1730
    },
    {
      "epoch": 0.8324836562106399,
      "grad_norm": 2.953125,
      "learning_rate": 9.248400852878465e-06,
      "loss": 0.9992,
      "step": 1735
    },
    {
      "epoch": 0.8348827445570683,
      "grad_norm": 3.015625,
      "learning_rate": 9.275053304904051e-06,
      "loss": 1.0015,
      "step": 1740
    },
    {
      "epoch": 0.8372818329034967,
      "grad_norm": 3.234375,
      "learning_rate": 9.30170575692964e-06,
      "loss": 0.977,
      "step": 1745
    },
    {
      "epoch": 0.839680921249925,
      "grad_norm": 3.53125,
      "learning_rate": 9.328358208955226e-06,
      "loss": 1.0238,
      "step": 1750
    },
    {
      "epoch": 0.8420800095963534,
      "grad_norm": 2.78125,
      "learning_rate": 9.35501066098081e-06,
      "loss": 0.9109,
      "step": 1755
    },
    {
      "epoch": 0.8444790979427818,
      "grad_norm": 3.78125,
      "learning_rate": 9.381663113006397e-06,
      "loss": 0.9748,
      "step": 1760
    },
    {
      "epoch": 0.8468781862892101,
      "grad_norm": 2.890625,
      "learning_rate": 9.408315565031983e-06,
      "loss": 0.9892,
      "step": 1765
    },
    {
      "epoch": 0.8492772746356385,
      "grad_norm": 3.4375,
      "learning_rate": 9.43496801705757e-06,
      "loss": 0.9563,
      "step": 1770
    },
    {
      "epoch": 0.8516763629820668,
      "grad_norm": 3.1875,
      "learning_rate": 9.461620469083156e-06,
      "loss": 1.0417,
      "step": 1775
    },
    {
      "epoch": 0.8540754513284952,
      "grad_norm": 2.828125,
      "learning_rate": 9.488272921108744e-06,
      "loss": 0.921,
      "step": 1780
    },
    {
      "epoch": 0.8564745396749235,
      "grad_norm": 3.53125,
      "learning_rate": 9.51492537313433e-06,
      "loss": 1.0076,
      "step": 1785
    },
    {
      "epoch": 0.8588736280213519,
      "grad_norm": 3.515625,
      "learning_rate": 9.541577825159916e-06,
      "loss": 1.0417,
      "step": 1790
    },
    {
      "epoch": 0.8612727163677802,
      "grad_norm": 3.75,
      "learning_rate": 9.568230277185503e-06,
      "loss": 0.934,
      "step": 1795
    },
    {
      "epoch": 0.8636718047142086,
      "grad_norm": 3.78125,
      "learning_rate": 9.594882729211089e-06,
      "loss": 1.0,
      "step": 1800
    },
    {
      "epoch": 0.8636718047142086,
      "eval_loss": 0.9894086718559265,
      "eval_runtime": 175.7987,
      "eval_samples_per_second": 42.15,
      "eval_steps_per_second": 10.54,
      "step": 1800
    },
    {
      "epoch": 0.866070893060637,
      "grad_norm": 2.765625,
      "learning_rate": 9.621535181236673e-06,
      "loss": 0.9134,
      "step": 1805
    },
    {
      "epoch": 0.8684699814070653,
      "grad_norm": 3.078125,
      "learning_rate": 9.64818763326226e-06,
      "loss": 1.033,
      "step": 1810
    },
    {
      "epoch": 0.8708690697534937,
      "grad_norm": 3.5,
      "learning_rate": 9.674840085287848e-06,
      "loss": 1.0698,
      "step": 1815
    },
    {
      "epoch": 0.873268158099922,
      "grad_norm": 2.890625,
      "learning_rate": 9.701492537313434e-06,
      "loss": 1.0128,
      "step": 1820
    },
    {
      "epoch": 0.8756672464463504,
      "grad_norm": 3.171875,
      "learning_rate": 9.72814498933902e-06,
      "loss": 0.9641,
      "step": 1825
    },
    {
      "epoch": 0.8780663347927787,
      "grad_norm": 3.53125,
      "learning_rate": 9.754797441364607e-06,
      "loss": 1.0036,
      "step": 1830
    },
    {
      "epoch": 0.8804654231392071,
      "grad_norm": 3.515625,
      "learning_rate": 9.781449893390193e-06,
      "loss": 1.0096,
      "step": 1835
    },
    {
      "epoch": 0.8828645114856355,
      "grad_norm": 3.78125,
      "learning_rate": 9.80810234541578e-06,
      "loss": 0.9844,
      "step": 1840
    },
    {
      "epoch": 0.8852635998320638,
      "grad_norm": 2.796875,
      "learning_rate": 9.834754797441366e-06,
      "loss": 0.9803,
      "step": 1845
    },
    {
      "epoch": 0.8876626881784921,
      "grad_norm": 3.828125,
      "learning_rate": 9.861407249466952e-06,
      "loss": 0.9482,
      "step": 1850
    },
    {
      "epoch": 0.8900617765249206,
      "grad_norm": 2.71875,
      "learning_rate": 9.888059701492538e-06,
      "loss": 0.8553,
      "step": 1855
    },
    {
      "epoch": 0.8924608648713489,
      "grad_norm": 3.0,
      "learning_rate": 9.914712153518125e-06,
      "loss": 0.9406,
      "step": 1860
    },
    {
      "epoch": 0.8948599532177772,
      "grad_norm": 2.875,
      "learning_rate": 9.941364605543711e-06,
      "loss": 1.0665,
      "step": 1865
    },
    {
      "epoch": 0.8972590415642056,
      "grad_norm": 2.5625,
      "learning_rate": 9.968017057569297e-06,
      "loss": 0.9742,
      "step": 1870
    },
    {
      "epoch": 0.899658129910634,
      "grad_norm": 3.046875,
      "learning_rate": 9.994669509594884e-06,
      "loss": 0.9881,
      "step": 1875
    },
    {
      "epoch": 0.9020572182570623,
      "grad_norm": 3.4375,
      "learning_rate": 9.999979383980725e-06,
      "loss": 0.8914,
      "step": 1880
    },
    {
      "epoch": 0.9044563066034906,
      "grad_norm": 3.203125,
      "learning_rate": 9.999895631693786e-06,
      "loss": 0.9904,
      "step": 1885
    },
    {
      "epoch": 0.9068553949499191,
      "grad_norm": 3.203125,
      "learning_rate": 9.999747455716298e-06,
      "loss": 1.0054,
      "step": 1890
    },
    {
      "epoch": 0.9092544832963474,
      "grad_norm": 3.40625,
      "learning_rate": 9.999534857957508e-06,
      "loss": 1.0812,
      "step": 1895
    },
    {
      "epoch": 0.9116535716427757,
      "grad_norm": 3.75,
      "learning_rate": 9.999257841156743e-06,
      "loss": 0.9685,
      "step": 1900
    },
    {
      "epoch": 0.9116535716427757,
      "eval_loss": 0.9848875999450684,
      "eval_runtime": 177.1556,
      "eval_samples_per_second": 41.828,
      "eval_steps_per_second": 10.46,
      "step": 1900
    },
    {
      "epoch": 0.9140526599892042,
      "grad_norm": 3.71875,
      "learning_rate": 9.998916408883365e-06,
      "loss": 1.0029,
      "step": 1905
    },
    {
      "epoch": 0.9164517483356325,
      "grad_norm": 3.125,
      "learning_rate": 9.99851056553673e-06,
      "loss": 0.9063,
      "step": 1910
    },
    {
      "epoch": 0.9188508366820608,
      "grad_norm": 3.53125,
      "learning_rate": 9.998040316346134e-06,
      "loss": 1.0225,
      "step": 1915
    },
    {
      "epoch": 0.9212499250284891,
      "grad_norm": 7.5625,
      "learning_rate": 9.99750566737074e-06,
      "loss": 0.9892,
      "step": 1920
    },
    {
      "epoch": 0.9236490133749176,
      "grad_norm": 3.28125,
      "learning_rate": 9.996906625499504e-06,
      "loss": 0.9066,
      "step": 1925
    },
    {
      "epoch": 0.9260481017213459,
      "grad_norm": 3.21875,
      "learning_rate": 9.996243198451085e-06,
      "loss": 1.0039,
      "step": 1930
    },
    {
      "epoch": 0.9284471900677742,
      "grad_norm": 3.390625,
      "learning_rate": 9.995515394773744e-06,
      "loss": 0.9956,
      "step": 1935
    },
    {
      "epoch": 0.9308462784142026,
      "grad_norm": 3.546875,
      "learning_rate": 9.99472322384524e-06,
      "loss": 0.9169,
      "step": 1940
    },
    {
      "epoch": 0.933245366760631,
      "grad_norm": 3.15625,
      "learning_rate": 9.993866695872699e-06,
      "loss": 0.9753,
      "step": 1945
    },
    {
      "epoch": 0.9356444551070593,
      "grad_norm": 3.65625,
      "learning_rate": 9.992945821892488e-06,
      "loss": 1.0543,
      "step": 1950
    },
    {
      "epoch": 0.9380435434534877,
      "grad_norm": 3.375,
      "learning_rate": 9.991960613770078e-06,
      "loss": 1.039,
      "step": 1955
    },
    {
      "epoch": 0.940442631799916,
      "grad_norm": 3.25,
      "learning_rate": 9.990911084199879e-06,
      "loss": 0.9921,
      "step": 1960
    },
    {
      "epoch": 0.9428417201463444,
      "grad_norm": 3.25,
      "learning_rate": 9.98979724670509e-06,
      "loss": 1.0042,
      "step": 1965
    },
    {
      "epoch": 0.9452408084927727,
      "grad_norm": 4.25,
      "learning_rate": 9.988619115637514e-06,
      "loss": 0.9578,
      "step": 1970
    },
    {
      "epoch": 0.9476398968392011,
      "grad_norm": 4.71875,
      "learning_rate": 9.98737670617738e-06,
      "loss": 0.9997,
      "step": 1975
    },
    {
      "epoch": 0.9500389851856295,
      "grad_norm": 3.5,
      "learning_rate": 9.98607003433314e-06,
      "loss": 1.0004,
      "step": 1980
    },
    {
      "epoch": 0.9524380735320578,
      "grad_norm": 3.15625,
      "learning_rate": 9.98469911694127e-06,
      "loss": 0.8219,
      "step": 1985
    },
    {
      "epoch": 0.9548371618784862,
      "grad_norm": 2.90625,
      "learning_rate": 9.983263971666051e-06,
      "loss": 1.0284,
      "step": 1990
    },
    {
      "epoch": 0.9572362502249145,
      "grad_norm": 3.703125,
      "learning_rate": 9.981764616999339e-06,
      "loss": 1.022,
      "step": 1995
    },
    {
      "epoch": 0.9596353385713429,
      "grad_norm": 3.3125,
      "learning_rate": 9.980201072260332e-06,
      "loss": 0.8576,
      "step": 2000
    },
    {
      "epoch": 0.9596353385713429,
      "eval_loss": 0.9806957244873047,
      "eval_runtime": 176.763,
      "eval_samples_per_second": 41.921,
      "eval_steps_per_second": 10.483,
      "step": 2000
    },
    {
      "epoch": 0.9620344269177713,
      "grad_norm": 3.15625,
      "learning_rate": 9.978573357595314e-06,
      "loss": 1.0717,
      "step": 2005
    },
    {
      "epoch": 0.9644335152641996,
      "grad_norm": 3.328125,
      "learning_rate": 9.9768814939774e-06,
      "loss": 0.9899,
      "step": 2010
    },
    {
      "epoch": 0.966832603610628,
      "grad_norm": 3.546875,
      "learning_rate": 9.975125503206262e-06,
      "loss": 1.047,
      "step": 2015
    },
    {
      "epoch": 0.9692316919570563,
      "grad_norm": 2.90625,
      "learning_rate": 9.973305407907856e-06,
      "loss": 0.8933,
      "step": 2020
    },
    {
      "epoch": 0.9716307803034847,
      "grad_norm": 3.296875,
      "learning_rate": 9.971421231534123e-06,
      "loss": 0.9153,
      "step": 2025
    },
    {
      "epoch": 0.974029868649913,
      "grad_norm": 3.34375,
      "learning_rate": 9.96947299836269e-06,
      "loss": 0.9508,
      "step": 2030
    },
    {
      "epoch": 0.9764289569963414,
      "grad_norm": 3.59375,
      "learning_rate": 9.967460733496552e-06,
      "loss": 0.945,
      "step": 2035
    },
    {
      "epoch": 0.9788280453427698,
      "grad_norm": 3.359375,
      "learning_rate": 9.965384462863757e-06,
      "loss": 0.9655,
      "step": 2040
    },
    {
      "epoch": 0.9812271336891981,
      "grad_norm": 3.28125,
      "learning_rate": 9.96324421321707e-06,
      "loss": 1.094,
      "step": 2045
    },
    {
      "epoch": 0.9836262220356264,
      "grad_norm": 3.1875,
      "learning_rate": 9.961040012133618e-06,
      "loss": 0.8919,
      "step": 2050
    },
    {
      "epoch": 0.9860253103820548,
      "grad_norm": 3.421875,
      "learning_rate": 9.958771888014549e-06,
      "loss": 1.0294,
      "step": 2055
    },
    {
      "epoch": 0.9884243987284832,
      "grad_norm": 4.03125,
      "learning_rate": 9.95643987008466e-06,
      "loss": 1.0486,
      "step": 2060
    },
    {
      "epoch": 0.9908234870749115,
      "grad_norm": 2.71875,
      "learning_rate": 9.954043988392017e-06,
      "loss": 1.0447,
      "step": 2065
    },
    {
      "epoch": 0.9932225754213398,
      "grad_norm": 3.578125,
      "learning_rate": 9.951584273807574e-06,
      "loss": 0.9619,
      "step": 2070
    },
    {
      "epoch": 0.9956216637677683,
      "grad_norm": 3.296875,
      "learning_rate": 9.949060758024768e-06,
      "loss": 0.9278,
      "step": 2075
    },
    {
      "epoch": 0.9980207521141966,
      "grad_norm": 3.453125,
      "learning_rate": 9.946473473559122e-06,
      "loss": 0.9724,
      "step": 2080
    },
    {
      "epoch": 1.000419840460625,
      "grad_norm": 3.296875,
      "learning_rate": 9.943822453747811e-06,
      "loss": 0.9471,
      "step": 2085
    },
    {
      "epoch": 1.0028189288070533,
      "grad_norm": 4.03125,
      "learning_rate": 9.941107732749247e-06,
      "loss": 0.8855,
      "step": 2090
    },
    {
      "epoch": 1.0052180171534817,
      "grad_norm": 3.109375,
      "learning_rate": 9.938329345542626e-06,
      "loss": 1.013,
      "step": 2095
    },
    {
      "epoch": 1.0076171054999101,
      "grad_norm": 2.578125,
      "learning_rate": 9.935487327927487e-06,
      "loss": 0.8853,
      "step": 2100
    },
    {
      "epoch": 1.0076171054999101,
      "eval_loss": 0.9774662256240845,
      "eval_runtime": 176.4723,
      "eval_samples_per_second": 41.99,
      "eval_steps_per_second": 10.5,
      "step": 2100
    },
    {
      "epoch": 1.0100161938463383,
      "grad_norm": 3.0,
      "learning_rate": 9.93258171652325e-06,
      "loss": 1.0145,
      "step": 2105
    },
    {
      "epoch": 1.0124152821927668,
      "grad_norm": 2.546875,
      "learning_rate": 9.929612548768735e-06,
      "loss": 1.0052,
      "step": 2110
    },
    {
      "epoch": 1.0148143705391952,
      "grad_norm": 3.421875,
      "learning_rate": 9.926579862921693e-06,
      "loss": 0.9061,
      "step": 2115
    },
    {
      "epoch": 1.0172134588856234,
      "grad_norm": 3.15625,
      "learning_rate": 9.923483698058301e-06,
      "loss": 0.7924,
      "step": 2120
    },
    {
      "epoch": 1.0196125472320519,
      "grad_norm": 3.0625,
      "learning_rate": 9.920324094072663e-06,
      "loss": 0.7767,
      "step": 2125
    },
    {
      "epoch": 1.02201163557848,
      "grad_norm": 4.15625,
      "learning_rate": 9.917101091676302e-06,
      "loss": 0.9295,
      "step": 2130
    },
    {
      "epoch": 1.0244107239249085,
      "grad_norm": 3.5,
      "learning_rate": 9.913814732397624e-06,
      "loss": 0.8664,
      "step": 2135
    },
    {
      "epoch": 1.026809812271337,
      "grad_norm": 3.140625,
      "learning_rate": 9.910465058581395e-06,
      "loss": 1.0206,
      "step": 2140
    },
    {
      "epoch": 1.0292089006177652,
      "grad_norm": 2.9375,
      "learning_rate": 9.907052113388183e-06,
      "loss": 0.9652,
      "step": 2145
    },
    {
      "epoch": 1.0316079889641936,
      "grad_norm": 7.40625,
      "learning_rate": 9.90357594079381e-06,
      "loss": 0.795,
      "step": 2150
    },
    {
      "epoch": 1.034007077310622,
      "grad_norm": 3.5,
      "learning_rate": 9.900036585588788e-06,
      "loss": 0.8602,
      "step": 2155
    },
    {
      "epoch": 1.0364061656570502,
      "grad_norm": 2.90625,
      "learning_rate": 9.89643409337773e-06,
      "loss": 0.9133,
      "step": 2160
    },
    {
      "epoch": 1.0388052540034787,
      "grad_norm": 2.828125,
      "learning_rate": 9.892768510578777e-06,
      "loss": 0.9104,
      "step": 2165
    },
    {
      "epoch": 1.041204342349907,
      "grad_norm": 2.734375,
      "learning_rate": 9.889039884422989e-06,
      "loss": 0.9898,
      "step": 2170
    },
    {
      "epoch": 1.0436034306963353,
      "grad_norm": 2.984375,
      "learning_rate": 9.885248262953736e-06,
      "loss": 1.031,
      "step": 2175
    },
    {
      "epoch": 1.0460025190427638,
      "grad_norm": 3.234375,
      "learning_rate": 9.88139369502609e-06,
      "loss": 0.9867,
      "step": 2180
    },
    {
      "epoch": 1.0484016073891922,
      "grad_norm": 2.921875,
      "learning_rate": 9.87747623030619e-06,
      "loss": 0.9899,
      "step": 2185
    },
    {
      "epoch": 1.0508006957356204,
      "grad_norm": 2.828125,
      "learning_rate": 9.873495919270593e-06,
      "loss": 0.8685,
      "step": 2190
    },
    {
      "epoch": 1.0531997840820488,
      "grad_norm": 2.859375,
      "learning_rate": 9.869452813205632e-06,
      "loss": 0.8505,
      "step": 2195
    },
    {
      "epoch": 1.0555988724284773,
      "grad_norm": 3.078125,
      "learning_rate": 9.865346964206762e-06,
      "loss": 0.947,
      "step": 2200
    },
    {
      "epoch": 1.0555988724284773,
      "eval_loss": 0.9739471673965454,
      "eval_runtime": 175.7002,
      "eval_samples_per_second": 42.174,
      "eval_steps_per_second": 10.546,
      "step": 2200
    },
    {
      "epoch": 1.0579979607749055,
      "grad_norm": 2.640625,
      "learning_rate": 9.861178425177874e-06,
      "loss": 0.9071,
      "step": 2205
    },
    {
      "epoch": 1.060397049121334,
      "grad_norm": 3.578125,
      "learning_rate": 9.856947249830624e-06,
      "loss": 0.9053,
      "step": 2210
    },
    {
      "epoch": 1.0627961374677621,
      "grad_norm": 3.03125,
      "learning_rate": 9.852653492683735e-06,
      "loss": 0.974,
      "step": 2215
    },
    {
      "epoch": 1.0651952258141906,
      "grad_norm": 3.25,
      "learning_rate": 9.848297209062299e-06,
      "loss": 0.8943,
      "step": 2220
    },
    {
      "epoch": 1.067594314160619,
      "grad_norm": 2.84375,
      "learning_rate": 9.843878455097061e-06,
      "loss": 0.976,
      "step": 2225
    },
    {
      "epoch": 1.0699934025070472,
      "grad_norm": 2.828125,
      "learning_rate": 9.839397287723695e-06,
      "loss": 0.8868,
      "step": 2230
    },
    {
      "epoch": 1.0723924908534757,
      "grad_norm": 3.375,
      "learning_rate": 9.83485376468208e-06,
      "loss": 0.9177,
      "step": 2235
    },
    {
      "epoch": 1.074791579199904,
      "grad_norm": 2.671875,
      "learning_rate": 9.830247944515536e-06,
      "loss": 0.9299,
      "step": 2240
    },
    {
      "epoch": 1.0771906675463323,
      "grad_norm": 3.046875,
      "learning_rate": 9.825579886570094e-06,
      "loss": 0.9351,
      "step": 2245
    },
    {
      "epoch": 1.0795897558927607,
      "grad_norm": 3.453125,
      "learning_rate": 9.820849650993709e-06,
      "loss": 0.8633,
      "step": 2250
    },
    {
      "epoch": 1.0819888442391892,
      "grad_norm": 3.921875,
      "learning_rate": 9.816057298735501e-06,
      "loss": 1.0134,
      "step": 2255
    },
    {
      "epoch": 1.0843879325856174,
      "grad_norm": 3.0625,
      "learning_rate": 9.811202891544965e-06,
      "loss": 0.9708,
      "step": 2260
    },
    {
      "epoch": 1.0867870209320458,
      "grad_norm": 2.890625,
      "learning_rate": 9.80628649197117e-06,
      "loss": 0.868,
      "step": 2265
    },
    {
      "epoch": 1.0891861092784743,
      "grad_norm": 3.046875,
      "learning_rate": 9.80130816336196e-06,
      "loss": 0.9277,
      "step": 2270
    },
    {
      "epoch": 1.0915851976249025,
      "grad_norm": 2.78125,
      "learning_rate": 9.796267969863134e-06,
      "loss": 0.9102,
      "step": 2275
    },
    {
      "epoch": 1.093984285971331,
      "grad_norm": 4.15625,
      "learning_rate": 9.791165976417621e-06,
      "loss": 0.8154,
      "step": 2280
    },
    {
      "epoch": 1.0963833743177593,
      "grad_norm": 3.25,
      "learning_rate": 9.786002248764642e-06,
      "loss": 0.8024,
      "step": 2285
    },
    {
      "epoch": 1.0987824626641876,
      "grad_norm": 3.03125,
      "learning_rate": 9.780776853438863e-06,
      "loss": 0.9436,
      "step": 2290
    },
    {
      "epoch": 1.101181551010616,
      "grad_norm": 11.1875,
      "learning_rate": 9.775489857769544e-06,
      "loss": 0.888,
      "step": 2295
    },
    {
      "epoch": 1.1035806393570444,
      "grad_norm": 3.34375,
      "learning_rate": 9.770141329879658e-06,
      "loss": 0.9207,
      "step": 2300
    },
    {
      "epoch": 1.1035806393570444,
      "eval_loss": 0.9713129997253418,
      "eval_runtime": 175.6734,
      "eval_samples_per_second": 42.181,
      "eval_steps_per_second": 10.548,
      "step": 2300
    },
    {
      "epoch": 1.1059797277034726,
      "grad_norm": 2.84375,
      "learning_rate": 9.764731338685026e-06,
      "loss": 0.9329,
      "step": 2305
    },
    {
      "epoch": 1.108378816049901,
      "grad_norm": 3.515625,
      "learning_rate": 9.75925995389342e-06,
      "loss": 0.9009,
      "step": 2310
    },
    {
      "epoch": 1.1107779043963295,
      "grad_norm": 3.125,
      "learning_rate": 9.753727246003677e-06,
      "loss": 0.9371,
      "step": 2315
    },
    {
      "epoch": 1.1131769927427577,
      "grad_norm": 4.09375,
      "learning_rate": 9.748133286304774e-06,
      "loss": 0.9952,
      "step": 2320
    },
    {
      "epoch": 1.1155760810891862,
      "grad_norm": 3.15625,
      "learning_rate": 9.74247814687492e-06,
      "loss": 0.8928,
      "step": 2325
    },
    {
      "epoch": 1.1179751694356144,
      "grad_norm": 3.453125,
      "learning_rate": 9.73676190058063e-06,
      "loss": 0.7709,
      "step": 2330
    },
    {
      "epoch": 1.1203742577820428,
      "grad_norm": 4.65625,
      "learning_rate": 9.730984621075777e-06,
      "loss": 0.9633,
      "step": 2335
    },
    {
      "epoch": 1.1227733461284712,
      "grad_norm": 3.578125,
      "learning_rate": 9.725146382800644e-06,
      "loss": 0.9883,
      "step": 2340
    },
    {
      "epoch": 1.1251724344748995,
      "grad_norm": 2.796875,
      "learning_rate": 9.719247260980977e-06,
      "loss": 0.8631,
      "step": 2345
    },
    {
      "epoch": 1.1275715228213279,
      "grad_norm": 3.296875,
      "learning_rate": 9.713287331627002e-06,
      "loss": 0.9008,
      "step": 2350
    },
    {
      "epoch": 1.1299706111677563,
      "grad_norm": 3.21875,
      "learning_rate": 9.70726667153245e-06,
      "loss": 0.8715,
      "step": 2355
    },
    {
      "epoch": 1.1323696995141845,
      "grad_norm": 3.078125,
      "learning_rate": 9.701185358273568e-06,
      "loss": 0.9063,
      "step": 2360
    },
    {
      "epoch": 1.134768787860613,
      "grad_norm": 3.15625,
      "learning_rate": 9.69504347020812e-06,
      "loss": 0.7194,
      "step": 2365
    },
    {
      "epoch": 1.1371678762070414,
      "grad_norm": 3.0,
      "learning_rate": 9.688841086474381e-06,
      "loss": 0.8225,
      "step": 2370
    },
    {
      "epoch": 1.1395669645534696,
      "grad_norm": 2.921875,
      "learning_rate": 9.682578286990105e-06,
      "loss": 1.033,
      "step": 2375
    },
    {
      "epoch": 1.141966052899898,
      "grad_norm": 3.484375,
      "learning_rate": 9.676255152451508e-06,
      "loss": 0.9202,
      "step": 2380
    },
    {
      "epoch": 1.1443651412463265,
      "grad_norm": 2.890625,
      "learning_rate": 9.669871764332226e-06,
      "loss": 0.9405,
      "step": 2385
    },
    {
      "epoch": 1.1467642295927547,
      "grad_norm": 3.515625,
      "learning_rate": 9.663428204882258e-06,
      "loss": 0.9197,
      "step": 2390
    },
    {
      "epoch": 1.1491633179391831,
      "grad_norm": 2.953125,
      "learning_rate": 9.656924557126913e-06,
      "loss": 0.9512,
      "step": 2395
    },
    {
      "epoch": 1.1515624062856116,
      "grad_norm": 2.96875,
      "learning_rate": 9.650360904865738e-06,
      "loss": 0.8596,
      "step": 2400
    },
    {
      "epoch": 1.1515624062856116,
      "eval_loss": 0.9691145420074463,
      "eval_runtime": 175.6179,
      "eval_samples_per_second": 42.194,
      "eval_steps_per_second": 10.551,
      "step": 2400
    },
    {
      "epoch": 1.1539614946320398,
      "grad_norm": 2.796875,
      "learning_rate": 9.643737332671441e-06,
      "loss": 1.0536,
      "step": 2405
    },
    {
      "epoch": 1.1563605829784682,
      "grad_norm": 2.59375,
      "learning_rate": 9.637053925888793e-06,
      "loss": 0.8452,
      "step": 2410
    },
    {
      "epoch": 1.1587596713248964,
      "grad_norm": 2.640625,
      "learning_rate": 9.630310770633542e-06,
      "loss": 0.7487,
      "step": 2415
    },
    {
      "epoch": 1.1611587596713249,
      "grad_norm": 2.640625,
      "learning_rate": 9.623507953791287e-06,
      "loss": 1.0087,
      "step": 2420
    },
    {
      "epoch": 1.1635578480177533,
      "grad_norm": 3.265625,
      "learning_rate": 9.616645563016373e-06,
      "loss": 0.9177,
      "step": 2425
    },
    {
      "epoch": 1.1659569363641815,
      "grad_norm": 2.765625,
      "learning_rate": 9.609723686730754e-06,
      "loss": 0.8388,
      "step": 2430
    },
    {
      "epoch": 1.16835602471061,
      "grad_norm": 3.484375,
      "learning_rate": 9.602742414122855e-06,
      "loss": 0.962,
      "step": 2435
    },
    {
      "epoch": 1.1707551130570384,
      "grad_norm": 3.46875,
      "learning_rate": 9.59570183514642e-06,
      "loss": 0.9372,
      "step": 2440
    },
    {
      "epoch": 1.1731542014034666,
      "grad_norm": 3.53125,
      "learning_rate": 9.588602040519363e-06,
      "loss": 0.9363,
      "step": 2445
    },
    {
      "epoch": 1.175553289749895,
      "grad_norm": 3.21875,
      "learning_rate": 9.581443121722585e-06,
      "loss": 0.8577,
      "step": 2450
    },
    {
      "epoch": 1.1779523780963235,
      "grad_norm": 3.578125,
      "learning_rate": 9.574225170998807e-06,
      "loss": 0.9581,
      "step": 2455
    },
    {
      "epoch": 1.1803514664427517,
      "grad_norm": 2.890625,
      "learning_rate": 9.566948281351373e-06,
      "loss": 0.9223,
      "step": 2460
    },
    {
      "epoch": 1.1827505547891801,
      "grad_norm": 3.296875,
      "learning_rate": 9.55961254654306e-06,
      "loss": 1.054,
      "step": 2465
    },
    {
      "epoch": 1.1851496431356086,
      "grad_norm": 4.03125,
      "learning_rate": 9.552218061094863e-06,
      "loss": 0.8736,
      "step": 2470
    },
    {
      "epoch": 1.1875487314820368,
      "grad_norm": 2.6875,
      "learning_rate": 9.544764920284775e-06,
      "loss": 0.8971,
      "step": 2475
    },
    {
      "epoch": 1.1899478198284652,
      "grad_norm": 2.734375,
      "learning_rate": 9.537253220146574e-06,
      "loss": 0.961,
      "step": 2480
    },
    {
      "epoch": 1.1923469081748936,
      "grad_norm": 5.6875,
      "learning_rate": 9.529683057468564e-06,
      "loss": 0.979,
      "step": 2485
    },
    {
      "epoch": 1.1947459965213219,
      "grad_norm": 3.25,
      "learning_rate": 9.522054529792348e-06,
      "loss": 0.8174,
      "step": 2490
    },
    {
      "epoch": 1.1971450848677503,
      "grad_norm": 3.15625,
      "learning_rate": 9.514367735411558e-06,
      "loss": 0.7918,
      "step": 2495
    },
    {
      "epoch": 1.1995441732141785,
      "grad_norm": 2.65625,
      "learning_rate": 9.506622773370595e-06,
      "loss": 1.0277,
      "step": 2500
    },
    {
      "epoch": 1.1995441732141785,
      "eval_loss": 0.9654711484909058,
      "eval_runtime": 175.623,
      "eval_samples_per_second": 42.193,
      "eval_steps_per_second": 10.551,
      "step": 2500
    },
    {
      "epoch": 1.201943261560607,
      "grad_norm": 6.3125,
      "learning_rate": 9.498819743463347e-06,
      "loss": 1.0265,
      "step": 2505
    },
    {
      "epoch": 1.2043423499070354,
      "grad_norm": 4.09375,
      "learning_rate": 9.490958746231911e-06,
      "loss": 0.9911,
      "step": 2510
    },
    {
      "epoch": 1.2067414382534638,
      "grad_norm": 2.609375,
      "learning_rate": 9.483039882965293e-06,
      "loss": 0.9705,
      "step": 2515
    },
    {
      "epoch": 1.209140526599892,
      "grad_norm": 3.421875,
      "learning_rate": 9.4750632556981e-06,
      "loss": 0.9417,
      "step": 2520
    },
    {
      "epoch": 1.2115396149463205,
      "grad_norm": 4.59375,
      "learning_rate": 9.467028967209232e-06,
      "loss": 0.9103,
      "step": 2525
    },
    {
      "epoch": 1.2139387032927487,
      "grad_norm": 2.734375,
      "learning_rate": 9.458937121020555e-06,
      "loss": 0.8767,
      "step": 2530
    },
    {
      "epoch": 1.216337791639177,
      "grad_norm": 3.1875,
      "learning_rate": 9.45078782139556e-06,
      "loss": 0.8681,
      "step": 2535
    },
    {
      "epoch": 1.2187368799856055,
      "grad_norm": 2.578125,
      "learning_rate": 9.442581173338032e-06,
      "loss": 0.8802,
      "step": 2540
    },
    {
      "epoch": 1.2211359683320337,
      "grad_norm": 3.125,
      "learning_rate": 9.43431728259069e-06,
      "loss": 0.8075,
      "step": 2545
    },
    {
      "epoch": 1.2235350566784622,
      "grad_norm": 3.4375,
      "learning_rate": 9.425996255633825e-06,
      "loss": 0.8549,
      "step": 2550
    },
    {
      "epoch": 1.2259341450248906,
      "grad_norm": 2.984375,
      "learning_rate": 9.417618199683926e-06,
      "loss": 0.8986,
      "step": 2555
    },
    {
      "epoch": 1.2283332333713188,
      "grad_norm": 2.703125,
      "learning_rate": 9.409183222692307e-06,
      "loss": 0.9454,
      "step": 2560
    },
    {
      "epoch": 1.2307323217177473,
      "grad_norm": 3.203125,
      "learning_rate": 9.4006914333437e-06,
      "loss": 0.7339,
      "step": 2565
    },
    {
      "epoch": 1.2331314100641757,
      "grad_norm": 3.734375,
      "learning_rate": 9.392142941054878e-06,
      "loss": 0.9632,
      "step": 2570
    },
    {
      "epoch": 1.235530498410604,
      "grad_norm": 3.0625,
      "learning_rate": 9.38353785597322e-06,
      "loss": 0.9105,
      "step": 2575
    },
    {
      "epoch": 1.2379295867570324,
      "grad_norm": 2.875,
      "learning_rate": 9.374876288975307e-06,
      "loss": 0.8167,
      "step": 2580
    },
    {
      "epoch": 1.2403286751034606,
      "grad_norm": 3.0625,
      "learning_rate": 9.366158351665495e-06,
      "loss": 0.871,
      "step": 2585
    },
    {
      "epoch": 1.242727763449889,
      "grad_norm": 3.171875,
      "learning_rate": 9.357384156374465e-06,
      "loss": 0.8765,
      "step": 2590
    },
    {
      "epoch": 1.2451268517963174,
      "grad_norm": 2.8125,
      "learning_rate": 9.348553816157785e-06,
      "loss": 1.032,
      "step": 2595
    },
    {
      "epoch": 1.2475259401427459,
      "grad_norm": 4.84375,
      "learning_rate": 9.339667444794456e-06,
      "loss": 0.9646,
      "step": 2600
    },
    {
      "epoch": 1.2475259401427459,
      "eval_loss": 0.9630805850028992,
      "eval_runtime": 175.7116,
      "eval_samples_per_second": 42.171,
      "eval_steps_per_second": 10.546,
      "step": 2600
    },
    {
      "epoch": 1.249925028489174,
      "grad_norm": 2.609375,
      "learning_rate": 9.33072515678543e-06,
      "loss": 0.8976,
      "step": 2605
    },
    {
      "epoch": 1.2523241168356025,
      "grad_norm": 2.90625,
      "learning_rate": 9.321727067352153e-06,
      "loss": 0.9049,
      "step": 2610
    },
    {
      "epoch": 1.2547232051820307,
      "grad_norm": 3.4375,
      "learning_rate": 9.312673292435073e-06,
      "loss": 0.9314,
      "step": 2615
    },
    {
      "epoch": 1.2571222935284592,
      "grad_norm": 2.734375,
      "learning_rate": 9.30356394869214e-06,
      "loss": 0.9375,
      "step": 2620
    },
    {
      "epoch": 1.2595213818748876,
      "grad_norm": 4.34375,
      "learning_rate": 9.294399153497316e-06,
      "loss": 0.886,
      "step": 2625
    },
    {
      "epoch": 1.261920470221316,
      "grad_norm": 3.453125,
      "learning_rate": 9.28517902493905e-06,
      "loss": 0.8899,
      "step": 2630
    },
    {
      "epoch": 1.2643195585677443,
      "grad_norm": 3.46875,
      "learning_rate": 9.275903681818763e-06,
      "loss": 0.8431,
      "step": 2635
    },
    {
      "epoch": 1.2667186469141727,
      "grad_norm": 3.328125,
      "learning_rate": 9.26657324364932e-06,
      "loss": 0.9004,
      "step": 2640
    },
    {
      "epoch": 1.269117735260601,
      "grad_norm": 3.078125,
      "learning_rate": 9.257187830653478e-06,
      "loss": 0.8216,
      "step": 2645
    },
    {
      "epoch": 1.2715168236070293,
      "grad_norm": 3.21875,
      "learning_rate": 9.247747563762353e-06,
      "loss": 0.8482,
      "step": 2650
    },
    {
      "epoch": 1.2739159119534578,
      "grad_norm": 2.9375,
      "learning_rate": 9.23825256461385e-06,
      "loss": 0.903,
      "step": 2655
    },
    {
      "epoch": 1.276315000299886,
      "grad_norm": 3.1875,
      "learning_rate": 9.228702955551101e-06,
      "loss": 0.8568,
      "step": 2660
    },
    {
      "epoch": 1.2787140886463144,
      "grad_norm": 3.109375,
      "learning_rate": 9.219098859620884e-06,
      "loss": 0.9839,
      "step": 2665
    },
    {
      "epoch": 1.2811131769927426,
      "grad_norm": 3.671875,
      "learning_rate": 9.209440400572045e-06,
      "loss": 0.9935,
      "step": 2670
    },
    {
      "epoch": 1.283512265339171,
      "grad_norm": 2.78125,
      "learning_rate": 9.199727702853896e-06,
      "loss": 0.9291,
      "step": 2675
    },
    {
      "epoch": 1.2859113536855995,
      "grad_norm": 3.375,
      "learning_rate": 9.189960891614616e-06,
      "loss": 0.8867,
      "step": 2680
    },
    {
      "epoch": 1.288310442032028,
      "grad_norm": 3.546875,
      "learning_rate": 9.180140092699636e-06,
      "loss": 0.9109,
      "step": 2685
    },
    {
      "epoch": 1.2907095303784561,
      "grad_norm": 2.84375,
      "learning_rate": 9.17026543265002e-06,
      "loss": 0.9201,
      "step": 2690
    },
    {
      "epoch": 1.2931086187248846,
      "grad_norm": 3.703125,
      "learning_rate": 9.160337038700834e-06,
      "loss": 0.9066,
      "step": 2695
    },
    {
      "epoch": 1.2955077070713128,
      "grad_norm": 3.34375,
      "learning_rate": 9.150355038779504e-06,
      "loss": 0.8583,
      "step": 2700
    },
    {
      "epoch": 1.2955077070713128,
      "eval_loss": 0.9613306522369385,
      "eval_runtime": 180.9335,
      "eval_samples_per_second": 40.954,
      "eval_steps_per_second": 10.241,
      "step": 2700
    },
    {
      "epoch": 1.2979067954177412,
      "grad_norm": 2.796875,
      "learning_rate": 9.140319561504168e-06,
      "loss": 0.8036,
      "step": 2705
    },
    {
      "epoch": 1.3003058837641697,
      "grad_norm": 4.125,
      "learning_rate": 9.13023073618202e-06,
      "loss": 0.9962,
      "step": 2710
    },
    {
      "epoch": 1.302704972110598,
      "grad_norm": 3.0625,
      "learning_rate": 9.12008869280765e-06,
      "loss": 0.8352,
      "step": 2715
    },
    {
      "epoch": 1.3051040604570263,
      "grad_norm": 2.859375,
      "learning_rate": 9.109893562061353e-06,
      "loss": 0.992,
      "step": 2720
    },
    {
      "epoch": 1.3075031488034548,
      "grad_norm": 3.390625,
      "learning_rate": 9.099645475307468e-06,
      "loss": 0.8995,
      "step": 2725
    },
    {
      "epoch": 1.309902237149883,
      "grad_norm": 2.796875,
      "learning_rate": 9.089344564592659e-06,
      "loss": 0.9768,
      "step": 2730
    },
    {
      "epoch": 1.3123013254963114,
      "grad_norm": 3.3125,
      "learning_rate": 9.078990962644237e-06,
      "loss": 0.8367,
      "step": 2735
    },
    {
      "epoch": 1.3147004138427398,
      "grad_norm": 3.03125,
      "learning_rate": 9.068584802868434e-06,
      "loss": 0.9742,
      "step": 2740
    },
    {
      "epoch": 1.317099502189168,
      "grad_norm": 2.875,
      "learning_rate": 9.058126219348692e-06,
      "loss": 0.8591,
      "step": 2745
    },
    {
      "epoch": 1.3194985905355965,
      "grad_norm": 2.71875,
      "learning_rate": 9.047615346843938e-06,
      "loss": 0.9576,
      "step": 2750
    },
    {
      "epoch": 1.3218976788820247,
      "grad_norm": 2.859375,
      "learning_rate": 9.037052320786833e-06,
      "loss": 1.0105,
      "step": 2755
    },
    {
      "epoch": 1.3242967672284531,
      "grad_norm": 3.765625,
      "learning_rate": 9.026437277282044e-06,
      "loss": 0.9007,
      "step": 2760
    },
    {
      "epoch": 1.3266958555748816,
      "grad_norm": 4.53125,
      "learning_rate": 9.015770353104482e-06,
      "loss": 0.9398,
      "step": 2765
    },
    {
      "epoch": 1.32909494392131,
      "grad_norm": 3.078125,
      "learning_rate": 9.005051685697544e-06,
      "loss": 0.9709,
      "step": 2770
    },
    {
      "epoch": 1.3314940322677382,
      "grad_norm": 2.9375,
      "learning_rate": 8.99428141317133e-06,
      "loss": 0.8677,
      "step": 2775
    },
    {
      "epoch": 1.3338931206141666,
      "grad_norm": 2.875,
      "learning_rate": 8.983459674300877e-06,
      "loss": 0.8374,
      "step": 2780
    },
    {
      "epoch": 1.3362922089605949,
      "grad_norm": 3.109375,
      "learning_rate": 8.972586608524371e-06,
      "loss": 0.8641,
      "step": 2785
    },
    {
      "epoch": 1.3386912973070233,
      "grad_norm": 2.875,
      "learning_rate": 8.961662355941339e-06,
      "loss": 0.9468,
      "step": 2790
    },
    {
      "epoch": 1.3410903856534517,
      "grad_norm": 3.1875,
      "learning_rate": 8.950687057310854e-06,
      "loss": 0.9041,
      "step": 2795
    },
    {
      "epoch": 1.3434894739998802,
      "grad_norm": 3.3125,
      "learning_rate": 8.939660854049716e-06,
      "loss": 0.9367,
      "step": 2800
    },
    {
      "epoch": 1.3434894739998802,
      "eval_loss": 0.9589128494262695,
      "eval_runtime": 186.8641,
      "eval_samples_per_second": 39.654,
      "eval_steps_per_second": 9.916,
      "step": 2800
    },
    {
      "epoch": 1.3458885623463084,
      "grad_norm": 3.359375,
      "learning_rate": 8.928583888230632e-06,
      "loss": 0.9278,
      "step": 2805
    },
    {
      "epoch": 1.3482876506927368,
      "grad_norm": 2.859375,
      "learning_rate": 8.917456302580384e-06,
      "loss": 0.8432,
      "step": 2810
    },
    {
      "epoch": 1.350686739039165,
      "grad_norm": 3.078125,
      "learning_rate": 8.906278240477993e-06,
      "loss": 0.8627,
      "step": 2815
    },
    {
      "epoch": 1.3530858273855935,
      "grad_norm": 2.75,
      "learning_rate": 8.895049845952868e-06,
      "loss": 0.8152,
      "step": 2820
    },
    {
      "epoch": 1.355484915732022,
      "grad_norm": 3.34375,
      "learning_rate": 8.883771263682949e-06,
      "loss": 0.909,
      "step": 2825
    },
    {
      "epoch": 1.3578840040784501,
      "grad_norm": 3.046875,
      "learning_rate": 8.872442638992853e-06,
      "loss": 0.8528,
      "step": 2830
    },
    {
      "epoch": 1.3602830924248785,
      "grad_norm": 4.125,
      "learning_rate": 8.861064117851987e-06,
      "loss": 0.8544,
      "step": 2835
    },
    {
      "epoch": 1.362682180771307,
      "grad_norm": 2.859375,
      "learning_rate": 8.849635846872675e-06,
      "loss": 0.905,
      "step": 2840
    },
    {
      "epoch": 1.3650812691177352,
      "grad_norm": 2.984375,
      "learning_rate": 8.83815797330827e-06,
      "loss": 0.8494,
      "step": 2845
    },
    {
      "epoch": 1.3674803574641636,
      "grad_norm": 3.453125,
      "learning_rate": 8.826630645051254e-06,
      "loss": 0.8055,
      "step": 2850
    },
    {
      "epoch": 1.369879445810592,
      "grad_norm": 2.671875,
      "learning_rate": 8.815054010631336e-06,
      "loss": 0.8326,
      "step": 2855
    },
    {
      "epoch": 1.3722785341570203,
      "grad_norm": 3.265625,
      "learning_rate": 8.803428219213527e-06,
      "loss": 0.8647,
      "step": 2860
    },
    {
      "epoch": 1.3746776225034487,
      "grad_norm": 4.375,
      "learning_rate": 8.791753420596237e-06,
      "loss": 0.9319,
      "step": 2865
    },
    {
      "epoch": 1.377076710849877,
      "grad_norm": 3.375,
      "learning_rate": 8.780029765209324e-06,
      "loss": 0.9639,
      "step": 2870
    },
    {
      "epoch": 1.3794757991963054,
      "grad_norm": 3.578125,
      "learning_rate": 8.768257404112175e-06,
      "loss": 0.9156,
      "step": 2875
    },
    {
      "epoch": 1.3818748875427338,
      "grad_norm": 3.484375,
      "learning_rate": 8.756436488991743e-06,
      "loss": 0.8767,
      "step": 2880
    },
    {
      "epoch": 1.3842739758891622,
      "grad_norm": 2.859375,
      "learning_rate": 8.744567172160601e-06,
      "loss": 0.8061,
      "step": 2885
    },
    {
      "epoch": 1.3866730642355904,
      "grad_norm": 3.90625,
      "learning_rate": 8.732649606554983e-06,
      "loss": 0.9058,
      "step": 2890
    },
    {
      "epoch": 1.3890721525820189,
      "grad_norm": 3.71875,
      "learning_rate": 8.720683945732807e-06,
      "loss": 0.9113,
      "step": 2895
    },
    {
      "epoch": 1.391471240928447,
      "grad_norm": 3.03125,
      "learning_rate": 8.708670343871697e-06,
      "loss": 0.9146,
      "step": 2900
    },
    {
      "epoch": 1.391471240928447,
      "eval_loss": 0.9569985866546631,
      "eval_runtime": 175.8282,
      "eval_samples_per_second": 42.143,
      "eval_steps_per_second": 10.539,
      "step": 2900
    },
    {
      "epoch": 1.3938703292748755,
      "grad_norm": 3.25,
      "learning_rate": 8.696608955766995e-06,
      "loss": 0.8575,
      "step": 2905
    },
    {
      "epoch": 1.396269417621304,
      "grad_norm": 3.109375,
      "learning_rate": 8.684499936829773e-06,
      "loss": 0.8904,
      "step": 2910
    },
    {
      "epoch": 1.3986685059677322,
      "grad_norm": 2.953125,
      "learning_rate": 8.67234344308483e-06,
      "loss": 0.8498,
      "step": 2915
    },
    {
      "epoch": 1.4010675943141606,
      "grad_norm": 2.6875,
      "learning_rate": 8.660139631168668e-06,
      "loss": 0.9119,
      "step": 2920
    },
    {
      "epoch": 1.403466682660589,
      "grad_norm": 3.203125,
      "learning_rate": 8.647888658327491e-06,
      "loss": 0.9382,
      "step": 2925
    },
    {
      "epoch": 1.4058657710070173,
      "grad_norm": 3.5,
      "learning_rate": 8.635590682415172e-06,
      "loss": 0.9829,
      "step": 2930
    },
    {
      "epoch": 1.4082648593534457,
      "grad_norm": 2.78125,
      "learning_rate": 8.623245861891217e-06,
      "loss": 0.9363,
      "step": 2935
    },
    {
      "epoch": 1.4106639476998741,
      "grad_norm": 3.046875,
      "learning_rate": 8.610854355818727e-06,
      "loss": 0.9634,
      "step": 2940
    },
    {
      "epoch": 1.4130630360463023,
      "grad_norm": 3.40625,
      "learning_rate": 8.598416323862344e-06,
      "loss": 0.8603,
      "step": 2945
    },
    {
      "epoch": 1.4154621243927308,
      "grad_norm": 2.828125,
      "learning_rate": 8.585931926286197e-06,
      "loss": 0.9382,
      "step": 2950
    },
    {
      "epoch": 1.417861212739159,
      "grad_norm": 2.796875,
      "learning_rate": 8.573401323951838e-06,
      "loss": 0.9435,
      "step": 2955
    },
    {
      "epoch": 1.4202603010855874,
      "grad_norm": 2.640625,
      "learning_rate": 8.560824678316166e-06,
      "loss": 0.9,
      "step": 2960
    },
    {
      "epoch": 1.4226593894320159,
      "grad_norm": 2.8125,
      "learning_rate": 8.548202151429351e-06,
      "loss": 0.8678,
      "step": 2965
    },
    {
      "epoch": 1.4250584777784443,
      "grad_norm": 3.265625,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.868,
      "step": 2970
    },
    {
      "epoch": 1.4274575661248725,
      "grad_norm": 3.140625,
      "learning_rate": 8.522820105056762e-06,
      "loss": 0.9313,
      "step": 2975
    },
    {
      "epoch": 1.429856654471301,
      "grad_norm": 2.96875,
      "learning_rate": 8.510060912618836e-06,
      "loss": 0.861,
      "step": 2980
    },
    {
      "epoch": 1.4322557428177292,
      "grad_norm": 2.8125,
      "learning_rate": 8.497256493021247e-06,
      "loss": 0.8741,
      "step": 2985
    },
    {
      "epoch": 1.4346548311641576,
      "grad_norm": 3.265625,
      "learning_rate": 8.484407011249027e-06,
      "loss": 0.9683,
      "step": 2990
    },
    {
      "epoch": 1.437053919510586,
      "grad_norm": 3.09375,
      "learning_rate": 8.471512632867844e-06,
      "loss": 0.8919,
      "step": 2995
    },
    {
      "epoch": 1.4394530078570145,
      "grad_norm": 3.34375,
      "learning_rate": 8.458573524021854e-06,
      "loss": 0.9697,
      "step": 3000
    },
    {
      "epoch": 1.4394530078570145,
      "eval_loss": 0.9555841088294983,
      "eval_runtime": 175.8483,
      "eval_samples_per_second": 42.139,
      "eval_steps_per_second": 10.537,
      "step": 3000
    },
    {
      "epoch": 1.4418520962034427,
      "grad_norm": 3.796875,
      "learning_rate": 8.445589851431563e-06,
      "loss": 0.9467,
      "step": 3005
    },
    {
      "epoch": 1.4442511845498711,
      "grad_norm": 3.734375,
      "learning_rate": 8.432561782391687e-06,
      "loss": 1.0304,
      "step": 3010
    },
    {
      "epoch": 1.4466502728962993,
      "grad_norm": 4.3125,
      "learning_rate": 8.419489484768988e-06,
      "loss": 0.8586,
      "step": 3015
    },
    {
      "epoch": 1.4490493612427278,
      "grad_norm": 3.53125,
      "learning_rate": 8.406373127000111e-06,
      "loss": 0.8946,
      "step": 3020
    },
    {
      "epoch": 1.4514484495891562,
      "grad_norm": 3.46875,
      "learning_rate": 8.393212878089418e-06,
      "loss": 0.9761,
      "step": 3025
    },
    {
      "epoch": 1.4538475379355844,
      "grad_norm": 3.4375,
      "learning_rate": 8.380008907606814e-06,
      "loss": 0.9166,
      "step": 3030
    },
    {
      "epoch": 1.4562466262820128,
      "grad_norm": 2.890625,
      "learning_rate": 8.366761385685547e-06,
      "loss": 0.9467,
      "step": 3035
    },
    {
      "epoch": 1.458645714628441,
      "grad_norm": 2.84375,
      "learning_rate": 8.353470483020032e-06,
      "loss": 0.9424,
      "step": 3040
    },
    {
      "epoch": 1.4610448029748695,
      "grad_norm": 3.140625,
      "learning_rate": 8.340136370863644e-06,
      "loss": 0.9584,
      "step": 3045
    },
    {
      "epoch": 1.463443891321298,
      "grad_norm": 2.671875,
      "learning_rate": 8.326759221026513e-06,
      "loss": 0.9652,
      "step": 3050
    },
    {
      "epoch": 1.4658429796677264,
      "grad_norm": 3.390625,
      "learning_rate": 8.31333920587331e-06,
      "loss": 1.0062,
      "step": 3055
    },
    {
      "epoch": 1.4682420680141546,
      "grad_norm": 3.09375,
      "learning_rate": 8.299876498321022e-06,
      "loss": 0.8903,
      "step": 3060
    },
    {
      "epoch": 1.470641156360583,
      "grad_norm": 3.0625,
      "learning_rate": 8.286371271836734e-06,
      "loss": 0.8911,
      "step": 3065
    },
    {
      "epoch": 1.4730402447070112,
      "grad_norm": 3.09375,
      "learning_rate": 8.272823700435382e-06,
      "loss": 0.8508,
      "step": 3070
    },
    {
      "epoch": 1.4754393330534397,
      "grad_norm": 3.5625,
      "learning_rate": 8.259233958677522e-06,
      "loss": 0.9078,
      "step": 3075
    },
    {
      "epoch": 1.477838421399868,
      "grad_norm": 3.203125,
      "learning_rate": 8.245602221667069e-06,
      "loss": 0.9197,
      "step": 3080
    },
    {
      "epoch": 1.4802375097462965,
      "grad_norm": 3.15625,
      "learning_rate": 8.231928665049057e-06,
      "loss": 0.8263,
      "step": 3085
    },
    {
      "epoch": 1.4826365980927247,
      "grad_norm": 3.40625,
      "learning_rate": 8.218213465007352e-06,
      "loss": 0.9468,
      "step": 3090
    },
    {
      "epoch": 1.4850356864391532,
      "grad_norm": 2.84375,
      "learning_rate": 8.204456798262408e-06,
      "loss": 0.9964,
      "step": 3095
    },
    {
      "epoch": 1.4874347747855814,
      "grad_norm": 3.765625,
      "learning_rate": 8.190658842068973e-06,
      "loss": 0.8713,
      "step": 3100
    },
    {
      "epoch": 1.4874347747855814,
      "eval_loss": 0.9542006850242615,
      "eval_runtime": 175.8535,
      "eval_samples_per_second": 42.137,
      "eval_steps_per_second": 10.537,
      "step": 3100
    },
    {
      "epoch": 1.4898338631320098,
      "grad_norm": 3.484375,
      "learning_rate": 8.176819774213807e-06,
      "loss": 0.9638,
      "step": 3105
    },
    {
      "epoch": 1.4922329514784383,
      "grad_norm": 3.015625,
      "learning_rate": 8.162939773013404e-06,
      "loss": 0.889,
      "step": 3110
    },
    {
      "epoch": 1.4946320398248665,
      "grad_norm": 2.390625,
      "learning_rate": 8.14901901731167e-06,
      "loss": 0.8743,
      "step": 3115
    },
    {
      "epoch": 1.497031128171295,
      "grad_norm": 3.359375,
      "learning_rate": 8.135057686477644e-06,
      "loss": 0.911,
      "step": 3120
    },
    {
      "epoch": 1.4994302165177233,
      "grad_norm": 2.9375,
      "learning_rate": 8.121055960403172e-06,
      "loss": 0.8063,
      "step": 3125
    },
    {
      "epoch": 1.5018293048641516,
      "grad_norm": 3.5,
      "learning_rate": 8.107014019500593e-06,
      "loss": 0.8361,
      "step": 3130
    },
    {
      "epoch": 1.50422839321058,
      "grad_norm": 3.171875,
      "learning_rate": 8.092932044700416e-06,
      "loss": 0.8645,
      "step": 3135
    },
    {
      "epoch": 1.5066274815570084,
      "grad_norm": 2.734375,
      "learning_rate": 8.078810217448986e-06,
      "loss": 0.9313,
      "step": 3140
    },
    {
      "epoch": 1.5090265699034366,
      "grad_norm": 3.0625,
      "learning_rate": 8.064648719706145e-06,
      "loss": 0.8725,
      "step": 3145
    },
    {
      "epoch": 1.511425658249865,
      "grad_norm": 3.484375,
      "learning_rate": 8.050447733942892e-06,
      "loss": 0.9236,
      "step": 3150
    },
    {
      "epoch": 1.5138247465962933,
      "grad_norm": 3.5,
      "learning_rate": 8.03620744313903e-06,
      "loss": 0.8835,
      "step": 3155
    },
    {
      "epoch": 1.5162238349427217,
      "grad_norm": 3.265625,
      "learning_rate": 8.021928030780806e-06,
      "loss": 0.9518,
      "step": 3160
    },
    {
      "epoch": 1.5186229232891502,
      "grad_norm": 3.03125,
      "learning_rate": 8.00760968085855e-06,
      "loss": 0.9473,
      "step": 3165
    },
    {
      "epoch": 1.5210220116355786,
      "grad_norm": 3.390625,
      "learning_rate": 7.993252577864302e-06,
      "loss": 0.8956,
      "step": 3170
    },
    {
      "epoch": 1.5234210999820068,
      "grad_norm": 2.515625,
      "learning_rate": 7.978856906789433e-06,
      "loss": 0.9128,
      "step": 3175
    },
    {
      "epoch": 1.5258201883284352,
      "grad_norm": 3.265625,
      "learning_rate": 7.964422853122268e-06,
      "loss": 0.8885,
      "step": 3180
    },
    {
      "epoch": 1.5282192766748635,
      "grad_norm": 2.765625,
      "learning_rate": 7.949950602845692e-06,
      "loss": 0.9026,
      "step": 3185
    },
    {
      "epoch": 1.530618365021292,
      "grad_norm": 2.703125,
      "learning_rate": 7.935440342434751e-06,
      "loss": 0.9181,
      "step": 3190
    },
    {
      "epoch": 1.5330174533677203,
      "grad_norm": 3.234375,
      "learning_rate": 7.920892258854252e-06,
      "loss": 0.9653,
      "step": 3195
    },
    {
      "epoch": 1.5354165417141488,
      "grad_norm": 3.34375,
      "learning_rate": 7.906306539556354e-06,
      "loss": 0.9855,
      "step": 3200
    },
    {
      "epoch": 1.5354165417141488,
      "eval_loss": 0.9524497985839844,
      "eval_runtime": 175.7768,
      "eval_samples_per_second": 42.156,
      "eval_steps_per_second": 10.542,
      "step": 3200
    },
    {
      "epoch": 1.537815630060577,
      "grad_norm": 3.203125,
      "learning_rate": 7.891683372478157e-06,
      "loss": 0.9054,
      "step": 3205
    },
    {
      "epoch": 1.5402147184070052,
      "grad_norm": 3.078125,
      "learning_rate": 7.87702294603927e-06,
      "loss": 0.7991,
      "step": 3210
    },
    {
      "epoch": 1.5426138067534336,
      "grad_norm": 3.640625,
      "learning_rate": 7.86232544913939e-06,
      "loss": 0.9486,
      "step": 3215
    },
    {
      "epoch": 1.545012895099862,
      "grad_norm": 3.375,
      "learning_rate": 7.847591071155871e-06,
      "loss": 0.8998,
      "step": 3220
    },
    {
      "epoch": 1.5474119834462905,
      "grad_norm": 3.203125,
      "learning_rate": 7.832820001941274e-06,
      "loss": 0.9656,
      "step": 3225
    },
    {
      "epoch": 1.549811071792719,
      "grad_norm": 2.984375,
      "learning_rate": 7.818012431820935e-06,
      "loss": 0.842,
      "step": 3230
    },
    {
      "epoch": 1.5522101601391471,
      "grad_norm": 2.859375,
      "learning_rate": 7.803168551590496e-06,
      "loss": 0.9687,
      "step": 3235
    },
    {
      "epoch": 1.5546092484855754,
      "grad_norm": 2.765625,
      "learning_rate": 7.788288552513459e-06,
      "loss": 0.9853,
      "step": 3240
    },
    {
      "epoch": 1.5570083368320038,
      "grad_norm": 2.984375,
      "learning_rate": 7.773372626318719e-06,
      "loss": 0.915,
      "step": 3245
    },
    {
      "epoch": 1.5594074251784322,
      "grad_norm": 3.109375,
      "learning_rate": 7.758420965198087e-06,
      "loss": 0.8403,
      "step": 3250
    },
    {
      "epoch": 1.5618065135248607,
      "grad_norm": 2.78125,
      "learning_rate": 7.743433761803826e-06,
      "loss": 0.8819,
      "step": 3255
    },
    {
      "epoch": 1.5642056018712889,
      "grad_norm": 3.09375,
      "learning_rate": 7.728411209246156e-06,
      "loss": 0.857,
      "step": 3260
    },
    {
      "epoch": 1.5666046902177173,
      "grad_norm": 2.8125,
      "learning_rate": 7.713353501090773e-06,
      "loss": 0.8918,
      "step": 3265
    },
    {
      "epoch": 1.5690037785641455,
      "grad_norm": 2.46875,
      "learning_rate": 7.698260831356352e-06,
      "loss": 0.8965,
      "step": 3270
    },
    {
      "epoch": 1.571402866910574,
      "grad_norm": 2.28125,
      "learning_rate": 7.683133394512053e-06,
      "loss": 0.8654,
      "step": 3275
    },
    {
      "epoch": 1.5738019552570024,
      "grad_norm": 2.9375,
      "learning_rate": 7.667971385475002e-06,
      "loss": 0.9612,
      "step": 3280
    },
    {
      "epoch": 1.5762010436034308,
      "grad_norm": 3.25,
      "learning_rate": 7.652774999607794e-06,
      "loss": 0.952,
      "step": 3285
    },
    {
      "epoch": 1.578600131949859,
      "grad_norm": 3.234375,
      "learning_rate": 7.63754443271597e-06,
      "loss": 0.8155,
      "step": 3290
    },
    {
      "epoch": 1.5809992202962873,
      "grad_norm": 3.875,
      "learning_rate": 7.622279881045489e-06,
      "loss": 0.9594,
      "step": 3295
    },
    {
      "epoch": 1.5833983086427157,
      "grad_norm": 3.5625,
      "learning_rate": 7.606981541280212e-06,
      "loss": 0.8651,
      "step": 3300
    },
    {
      "epoch": 1.5833983086427157,
      "eval_loss": 0.9511102437973022,
      "eval_runtime": 175.9414,
      "eval_samples_per_second": 42.116,
      "eval_steps_per_second": 10.532,
      "step": 3300
    },
    {
      "epoch": 1.5857973969891441,
      "grad_norm": 2.640625,
      "learning_rate": 7.591649610539349e-06,
      "loss": 0.9834,
      "step": 3305
    },
    {
      "epoch": 1.5881964853355726,
      "grad_norm": 2.953125,
      "learning_rate": 7.57628428637494e-06,
      "loss": 1.0254,
      "step": 3310
    },
    {
      "epoch": 1.590595573682001,
      "grad_norm": 3.234375,
      "learning_rate": 7.560885766769295e-06,
      "loss": 0.8831,
      "step": 3315
    },
    {
      "epoch": 1.5929946620284292,
      "grad_norm": 3.21875,
      "learning_rate": 7.5454542501324445e-06,
      "loss": 0.9003,
      "step": 3320
    },
    {
      "epoch": 1.5953937503748574,
      "grad_norm": 3.609375,
      "learning_rate": 7.529989935299595e-06,
      "loss": 0.8892,
      "step": 3325
    },
    {
      "epoch": 1.5977928387212859,
      "grad_norm": 3.625,
      "learning_rate": 7.514493021528548e-06,
      "loss": 0.9812,
      "step": 3330
    },
    {
      "epoch": 1.6001919270677143,
      "grad_norm": 2.4375,
      "learning_rate": 7.498963708497149e-06,
      "loss": 0.9269,
      "step": 3335
    },
    {
      "epoch": 1.6025910154141427,
      "grad_norm": 2.890625,
      "learning_rate": 7.483402196300705e-06,
      "loss": 0.8783,
      "step": 3340
    },
    {
      "epoch": 1.604990103760571,
      "grad_norm": 2.890625,
      "learning_rate": 7.467808685449413e-06,
      "loss": 0.8795,
      "step": 3345
    },
    {
      "epoch": 1.6073891921069994,
      "grad_norm": 4.1875,
      "learning_rate": 7.452183376865768e-06,
      "loss": 0.9221,
      "step": 3350
    },
    {
      "epoch": 1.6097882804534276,
      "grad_norm": 3.09375,
      "learning_rate": 7.436526471881982e-06,
      "loss": 1.0129,
      "step": 3355
    },
    {
      "epoch": 1.612187368799856,
      "grad_norm": 3.171875,
      "learning_rate": 7.420838172237388e-06,
      "loss": 0.9056,
      "step": 3360
    },
    {
      "epoch": 1.6145864571462845,
      "grad_norm": 2.765625,
      "learning_rate": 7.405118680075835e-06,
      "loss": 0.8635,
      "step": 3365
    },
    {
      "epoch": 1.616985545492713,
      "grad_norm": 2.921875,
      "learning_rate": 7.389368197943092e-06,
      "loss": 0.8995,
      "step": 3370
    },
    {
      "epoch": 1.619384633839141,
      "grad_norm": 2.828125,
      "learning_rate": 7.373586928784234e-06,
      "loss": 0.8698,
      "step": 3375
    },
    {
      "epoch": 1.6217837221855693,
      "grad_norm": 3.96875,
      "learning_rate": 7.357775075941025e-06,
      "loss": 0.922,
      "step": 3380
    },
    {
      "epoch": 1.6241828105319978,
      "grad_norm": 3.15625,
      "learning_rate": 7.341932843149298e-06,
      "loss": 0.9133,
      "step": 3385
    },
    {
      "epoch": 1.6265818988784262,
      "grad_norm": 3.09375,
      "learning_rate": 7.326060434536337e-06,
      "loss": 0.8743,
      "step": 3390
    },
    {
      "epoch": 1.6289809872248546,
      "grad_norm": 3.1875,
      "learning_rate": 7.31015805461824e-06,
      "loss": 0.8689,
      "step": 3395
    },
    {
      "epoch": 1.631380075571283,
      "grad_norm": 3.296875,
      "learning_rate": 7.294225908297281e-06,
      "loss": 0.9448,
      "step": 3400
    },
    {
      "epoch": 1.631380075571283,
      "eval_loss": 0.9495499730110168,
      "eval_runtime": 176.9541,
      "eval_samples_per_second": 41.875,
      "eval_steps_per_second": 10.472,
      "step": 3400
    },
    {
      "epoch": 1.6337791639177113,
      "grad_norm": 3.125,
      "learning_rate": 7.278264200859281e-06,
      "loss": 0.9375,
      "step": 3405
    },
    {
      "epoch": 1.6361782522641395,
      "grad_norm": 2.796875,
      "learning_rate": 7.262273137970953e-06,
      "loss": 0.9227,
      "step": 3410
    },
    {
      "epoch": 1.638577340610568,
      "grad_norm": 3.09375,
      "learning_rate": 7.246252925677253e-06,
      "loss": 0.811,
      "step": 3415
    },
    {
      "epoch": 1.6409764289569964,
      "grad_norm": 3.21875,
      "learning_rate": 7.230203770398734e-06,
      "loss": 1.0199,
      "step": 3420
    },
    {
      "epoch": 1.6433755173034248,
      "grad_norm": 2.828125,
      "learning_rate": 7.21412587892887e-06,
      "loss": 0.8974,
      "step": 3425
    },
    {
      "epoch": 1.645774605649853,
      "grad_norm": 3.4375,
      "learning_rate": 7.19801945843141e-06,
      "loss": 0.9508,
      "step": 3430
    },
    {
      "epoch": 1.6481736939962814,
      "grad_norm": 3.28125,
      "learning_rate": 7.181884716437694e-06,
      "loss": 0.9031,
      "step": 3435
    },
    {
      "epoch": 1.6505727823427097,
      "grad_norm": 3.171875,
      "learning_rate": 7.165721860843987e-06,
      "loss": 0.8852,
      "step": 3440
    },
    {
      "epoch": 1.652971870689138,
      "grad_norm": 3.15625,
      "learning_rate": 7.149531099908799e-06,
      "loss": 0.8157,
      "step": 3445
    },
    {
      "epoch": 1.6553709590355665,
      "grad_norm": 2.90625,
      "learning_rate": 7.1333126422501965e-06,
      "loss": 0.9073,
      "step": 3450
    },
    {
      "epoch": 1.657770047381995,
      "grad_norm": 3.25,
      "learning_rate": 7.1170666968431225e-06,
      "loss": 0.9776,
      "step": 3455
    },
    {
      "epoch": 1.6601691357284232,
      "grad_norm": 2.9375,
      "learning_rate": 7.100793473016699e-06,
      "loss": 0.9031,
      "step": 3460
    },
    {
      "epoch": 1.6625682240748516,
      "grad_norm": 2.78125,
      "learning_rate": 7.084493180451529e-06,
      "loss": 0.872,
      "step": 3465
    },
    {
      "epoch": 1.6649673124212798,
      "grad_norm": 2.109375,
      "learning_rate": 7.068166029176996e-06,
      "loss": 0.937,
      "step": 3470
    },
    {
      "epoch": 1.6673664007677083,
      "grad_norm": 3.3125,
      "learning_rate": 7.051812229568562e-06,
      "loss": 0.8244,
      "step": 3475
    },
    {
      "epoch": 1.6697654891141367,
      "grad_norm": 3.453125,
      "learning_rate": 7.035431992345051e-06,
      "loss": 0.8413,
      "step": 3480
    },
    {
      "epoch": 1.6721645774605651,
      "grad_norm": 3.28125,
      "learning_rate": 7.019025528565933e-06,
      "loss": 0.917,
      "step": 3485
    },
    {
      "epoch": 1.6745636658069933,
      "grad_norm": 3.375,
      "learning_rate": 7.002593049628611e-06,
      "loss": 0.8301,
      "step": 3490
    },
    {
      "epoch": 1.6769627541534216,
      "grad_norm": 5.25,
      "learning_rate": 6.986134767265693e-06,
      "loss": 0.917,
      "step": 3495
    },
    {
      "epoch": 1.67936184249985,
      "grad_norm": 2.859375,
      "learning_rate": 6.969650893542261e-06,
      "loss": 0.8997,
      "step": 3500
    },
    {
      "epoch": 1.67936184249985,
      "eval_loss": 0.9484797120094299,
      "eval_runtime": 175.9736,
      "eval_samples_per_second": 42.109,
      "eval_steps_per_second": 10.53,
      "step": 3500
    },
    {
      "epoch": 1.6817609308462784,
      "grad_norm": 3.890625,
      "learning_rate": 6.9531416408531475e-06,
      "loss": 0.7639,
      "step": 3505
    },
    {
      "epoch": 1.6841600191927069,
      "grad_norm": 3.96875,
      "learning_rate": 6.936607221920188e-06,
      "loss": 0.8797,
      "step": 3510
    },
    {
      "epoch": 1.6865591075391353,
      "grad_norm": 2.90625,
      "learning_rate": 6.920047849789488e-06,
      "loss": 0.7968,
      "step": 3515
    },
    {
      "epoch": 1.6889581958855635,
      "grad_norm": 3.84375,
      "learning_rate": 6.903463737828675e-06,
      "loss": 0.9295,
      "step": 3520
    },
    {
      "epoch": 1.6913572842319917,
      "grad_norm": 3.796875,
      "learning_rate": 6.886855099724148e-06,
      "loss": 0.8382,
      "step": 3525
    },
    {
      "epoch": 1.6937563725784202,
      "grad_norm": 2.96875,
      "learning_rate": 6.870222149478326e-06,
      "loss": 0.931,
      "step": 3530
    },
    {
      "epoch": 1.6961554609248486,
      "grad_norm": 2.546875,
      "learning_rate": 6.853565101406891e-06,
      "loss": 0.9034,
      "step": 3535
    },
    {
      "epoch": 1.698554549271277,
      "grad_norm": 3.046875,
      "learning_rate": 6.836884170136026e-06,
      "loss": 0.881,
      "step": 3540
    },
    {
      "epoch": 1.7009536376177052,
      "grad_norm": 3.21875,
      "learning_rate": 6.8201795705996465e-06,
      "loss": 0.8059,
      "step": 3545
    },
    {
      "epoch": 1.7033527259641337,
      "grad_norm": 3.296875,
      "learning_rate": 6.8034515180366366e-06,
      "loss": 0.7942,
      "step": 3550
    },
    {
      "epoch": 1.7057518143105619,
      "grad_norm": 4.21875,
      "learning_rate": 6.786700227988072e-06,
      "loss": 0.8774,
      "step": 3555
    },
    {
      "epoch": 1.7081509026569903,
      "grad_norm": 3.375,
      "learning_rate": 6.7699259162944445e-06,
      "loss": 0.8572,
      "step": 3560
    },
    {
      "epoch": 1.7105499910034188,
      "grad_norm": 2.859375,
      "learning_rate": 6.753128799092875e-06,
      "loss": 0.9758,
      "step": 3565
    },
    {
      "epoch": 1.7129490793498472,
      "grad_norm": 3.578125,
      "learning_rate": 6.7363090928143414e-06,
      "loss": 0.9013,
      "step": 3570
    },
    {
      "epoch": 1.7153481676962754,
      "grad_norm": 3.25,
      "learning_rate": 6.719467014180876e-06,
      "loss": 0.9169,
      "step": 3575
    },
    {
      "epoch": 1.7177472560427036,
      "grad_norm": 3.59375,
      "learning_rate": 6.702602780202779e-06,
      "loss": 0.7857,
      "step": 3580
    },
    {
      "epoch": 1.720146344389132,
      "grad_norm": 4.21875,
      "learning_rate": 6.68571660817583e-06,
      "loss": 0.9271,
      "step": 3585
    },
    {
      "epoch": 1.7225454327355605,
      "grad_norm": 2.90625,
      "learning_rate": 6.66880871567847e-06,
      "loss": 0.9313,
      "step": 3590
    },
    {
      "epoch": 1.724944521081989,
      "grad_norm": 3.0,
      "learning_rate": 6.651879320569015e-06,
      "loss": 0.9542,
      "step": 3595
    },
    {
      "epoch": 1.7273436094284174,
      "grad_norm": 3.59375,
      "learning_rate": 6.634928640982841e-06,
      "loss": 1.0446,
      "step": 3600
    },
    {
      "epoch": 1.7273436094284174,
      "eval_loss": 0.9475127458572388,
      "eval_runtime": 175.7691,
      "eval_samples_per_second": 42.158,
      "eval_steps_per_second": 10.542,
      "step": 3600
    },
    {
      "epoch": 1.7297426977748456,
      "grad_norm": 2.75,
      "learning_rate": 6.617956895329574e-06,
      "loss": 0.8341,
      "step": 3605
    },
    {
      "epoch": 1.7321417861212738,
      "grad_norm": 2.828125,
      "learning_rate": 6.600964302290275e-06,
      "loss": 0.8448,
      "step": 3610
    },
    {
      "epoch": 1.7345408744677022,
      "grad_norm": 2.265625,
      "learning_rate": 6.5839510808146276e-06,
      "loss": 0.887,
      "step": 3615
    },
    {
      "epoch": 1.7369399628141307,
      "grad_norm": 2.921875,
      "learning_rate": 6.566917450118109e-06,
      "loss": 0.9268,
      "step": 3620
    },
    {
      "epoch": 1.739339051160559,
      "grad_norm": 3.28125,
      "learning_rate": 6.549863629679174e-06,
      "loss": 0.873,
      "step": 3625
    },
    {
      "epoch": 1.7417381395069873,
      "grad_norm": 3.625,
      "learning_rate": 6.532789839236417e-06,
      "loss": 0.8447,
      "step": 3630
    },
    {
      "epoch": 1.7441372278534157,
      "grad_norm": 2.984375,
      "learning_rate": 6.5156962987857485e-06,
      "loss": 0.9165,
      "step": 3635
    },
    {
      "epoch": 1.746536316199844,
      "grad_norm": 3.046875,
      "learning_rate": 6.498583228577559e-06,
      "loss": 0.8892,
      "step": 3640
    },
    {
      "epoch": 1.7489354045462724,
      "grad_norm": 2.890625,
      "learning_rate": 6.48145084911388e-06,
      "loss": 0.8452,
      "step": 3645
    },
    {
      "epoch": 1.7513344928927008,
      "grad_norm": 3.125,
      "learning_rate": 6.464299381145539e-06,
      "loss": 0.8913,
      "step": 3650
    },
    {
      "epoch": 1.7537335812391293,
      "grad_norm": 3.21875,
      "learning_rate": 6.4471290456693245e-06,
      "loss": 0.9104,
      "step": 3655
    },
    {
      "epoch": 1.7561326695855575,
      "grad_norm": 3.234375,
      "learning_rate": 6.429940063925129e-06,
      "loss": 0.8567,
      "step": 3660
    },
    {
      "epoch": 1.7585317579319857,
      "grad_norm": 2.671875,
      "learning_rate": 6.412732657393104e-06,
      "loss": 0.8169,
      "step": 3665
    },
    {
      "epoch": 1.7609308462784141,
      "grad_norm": 3.03125,
      "learning_rate": 6.395507047790807e-06,
      "loss": 0.7973,
      "step": 3670
    },
    {
      "epoch": 1.7633299346248426,
      "grad_norm": 2.78125,
      "learning_rate": 6.378263457070334e-06,
      "loss": 0.9667,
      "step": 3675
    },
    {
      "epoch": 1.765729022971271,
      "grad_norm": 2.890625,
      "learning_rate": 6.361002107415478e-06,
      "loss": 0.9165,
      "step": 3680
    },
    {
      "epoch": 1.7681281113176994,
      "grad_norm": 3.078125,
      "learning_rate": 6.34372322123885e-06,
      "loss": 0.8751,
      "step": 3685
    },
    {
      "epoch": 1.7705271996641276,
      "grad_norm": 3.71875,
      "learning_rate": 6.32642702117902e-06,
      "loss": 0.9498,
      "step": 3690
    },
    {
      "epoch": 1.7729262880105559,
      "grad_norm": 2.96875,
      "learning_rate": 6.309113730097647e-06,
      "loss": 0.8921,
      "step": 3695
    },
    {
      "epoch": 1.7753253763569843,
      "grad_norm": 3.015625,
      "learning_rate": 6.291783571076612e-06,
      "loss": 0.8862,
      "step": 3700
    },
    {
      "epoch": 1.7753253763569843,
      "eval_loss": 0.9464648365974426,
      "eval_runtime": 175.7404,
      "eval_samples_per_second": 42.164,
      "eval_steps_per_second": 10.544,
      "step": 3700
    },
    {
      "epoch": 1.7777244647034127,
      "grad_norm": 2.609375,
      "learning_rate": 6.274436767415133e-06,
      "loss": 0.9401,
      "step": 3705
    },
    {
      "epoch": 1.7801235530498412,
      "grad_norm": 2.765625,
      "learning_rate": 6.257073542626899e-06,
      "loss": 0.9079,
      "step": 3710
    },
    {
      "epoch": 1.7825226413962694,
      "grad_norm": 2.734375,
      "learning_rate": 6.239694120437186e-06,
      "loss": 0.8473,
      "step": 3715
    },
    {
      "epoch": 1.7849217297426978,
      "grad_norm": 3.5625,
      "learning_rate": 6.2222987247799705e-06,
      "loss": 0.867,
      "step": 3720
    },
    {
      "epoch": 1.787320818089126,
      "grad_norm": 3.1875,
      "learning_rate": 6.204887579795046e-06,
      "loss": 0.8493,
      "step": 3725
    },
    {
      "epoch": 1.7897199064355545,
      "grad_norm": 3.296875,
      "learning_rate": 6.187460909825142e-06,
      "loss": 0.8569,
      "step": 3730
    },
    {
      "epoch": 1.7921189947819829,
      "grad_norm": 2.90625,
      "learning_rate": 6.170018939413024e-06,
      "loss": 0.8716,
      "step": 3735
    },
    {
      "epoch": 1.7945180831284113,
      "grad_norm": 3.078125,
      "learning_rate": 6.152561893298601e-06,
      "loss": 1.0129,
      "step": 3740
    },
    {
      "epoch": 1.7969171714748395,
      "grad_norm": 2.5625,
      "learning_rate": 6.135089996416039e-06,
      "loss": 0.8712,
      "step": 3745
    },
    {
      "epoch": 1.799316259821268,
      "grad_norm": 2.9375,
      "learning_rate": 6.1176034738908515e-06,
      "loss": 0.9311,
      "step": 3750
    },
    {
      "epoch": 1.8017153481676962,
      "grad_norm": 2.8125,
      "learning_rate": 6.100102551037003e-06,
      "loss": 0.948,
      "step": 3755
    },
    {
      "epoch": 1.8041144365141246,
      "grad_norm": 2.984375,
      "learning_rate": 6.082587453354012e-06,
      "loss": 0.956,
      "step": 3760
    },
    {
      "epoch": 1.806513524860553,
      "grad_norm": 3.34375,
      "learning_rate": 6.065058406524033e-06,
      "loss": 0.9149,
      "step": 3765
    },
    {
      "epoch": 1.8089126132069815,
      "grad_norm": 2.75,
      "learning_rate": 6.047515636408959e-06,
      "loss": 0.8932,
      "step": 3770
    },
    {
      "epoch": 1.8113117015534097,
      "grad_norm": 3.265625,
      "learning_rate": 6.029959369047507e-06,
      "loss": 0.7662,
      "step": 3775
    },
    {
      "epoch": 1.813710789899838,
      "grad_norm": 3.625,
      "learning_rate": 6.012389830652307e-06,
      "loss": 0.9215,
      "step": 3780
    },
    {
      "epoch": 1.8161098782462664,
      "grad_norm": 3.546875,
      "learning_rate": 5.994807247606984e-06,
      "loss": 0.8813,
      "step": 3785
    },
    {
      "epoch": 1.8185089665926948,
      "grad_norm": 3.171875,
      "learning_rate": 5.977211846463243e-06,
      "loss": 0.8829,
      "step": 3790
    },
    {
      "epoch": 1.8209080549391232,
      "grad_norm": 4.0,
      "learning_rate": 5.959603853937958e-06,
      "loss": 1.0723,
      "step": 3795
    },
    {
      "epoch": 1.8233071432855517,
      "grad_norm": 3.03125,
      "learning_rate": 5.941983496910232e-06,
      "loss": 0.873,
      "step": 3800
    },
    {
      "epoch": 1.8233071432855517,
      "eval_loss": 0.9455747008323669,
      "eval_runtime": 175.6671,
      "eval_samples_per_second": 42.182,
      "eval_steps_per_second": 10.548,
      "step": 3800
    },
    {
      "epoch": 1.8257062316319799,
      "grad_norm": 3.03125,
      "learning_rate": 5.924351002418489e-06,
      "loss": 0.9247,
      "step": 3805
    },
    {
      "epoch": 1.828105319978408,
      "grad_norm": 3.265625,
      "learning_rate": 5.90670659765755e-06,
      "loss": 0.9389,
      "step": 3810
    },
    {
      "epoch": 1.8305044083248365,
      "grad_norm": 2.984375,
      "learning_rate": 5.889050509975692e-06,
      "loss": 0.8573,
      "step": 3815
    },
    {
      "epoch": 1.832903496671265,
      "grad_norm": 3.265625,
      "learning_rate": 5.8713829668717295e-06,
      "loss": 0.8662,
      "step": 3820
    },
    {
      "epoch": 1.8353025850176934,
      "grad_norm": 4.3125,
      "learning_rate": 5.853704195992082e-06,
      "loss": 0.9179,
      "step": 3825
    },
    {
      "epoch": 1.8377016733641216,
      "grad_norm": 2.5,
      "learning_rate": 5.836014425127835e-06,
      "loss": 0.7961,
      "step": 3830
    },
    {
      "epoch": 1.84010076171055,
      "grad_norm": 3.59375,
      "learning_rate": 5.8183138822118125e-06,
      "loss": 0.8975,
      "step": 3835
    },
    {
      "epoch": 1.8424998500569783,
      "grad_norm": 3.453125,
      "learning_rate": 5.800602795315633e-06,
      "loss": 0.8728,
      "step": 3840
    },
    {
      "epoch": 1.8448989384034067,
      "grad_norm": 3.453125,
      "learning_rate": 5.7828813926467795e-06,
      "loss": 1.0257,
      "step": 3845
    },
    {
      "epoch": 1.8472980267498351,
      "grad_norm": 3.15625,
      "learning_rate": 5.765149902545649e-06,
      "loss": 0.8953,
      "step": 3850
    },
    {
      "epoch": 1.8496971150962636,
      "grad_norm": 3.859375,
      "learning_rate": 5.747408553482616e-06,
      "loss": 0.8241,
      "step": 3855
    },
    {
      "epoch": 1.8520962034426918,
      "grad_norm": 3.0,
      "learning_rate": 5.729657574055089e-06,
      "loss": 0.9077,
      "step": 3860
    },
    {
      "epoch": 1.85449529178912,
      "grad_norm": 3.875,
      "learning_rate": 5.711897192984567e-06,
      "loss": 0.9028,
      "step": 3865
    },
    {
      "epoch": 1.8568943801355484,
      "grad_norm": 4.03125,
      "learning_rate": 5.694127639113679e-06,
      "loss": 0.8912,
      "step": 3870
    },
    {
      "epoch": 1.8592934684819769,
      "grad_norm": 3.03125,
      "learning_rate": 5.676349141403257e-06,
      "loss": 0.8446,
      "step": 3875
    },
    {
      "epoch": 1.8616925568284053,
      "grad_norm": 2.6875,
      "learning_rate": 5.658561928929368e-06,
      "loss": 0.7482,
      "step": 3880
    },
    {
      "epoch": 1.8640916451748337,
      "grad_norm": 2.703125,
      "learning_rate": 5.6407662308803704e-06,
      "loss": 0.9638,
      "step": 3885
    },
    {
      "epoch": 1.866490733521262,
      "grad_norm": 2.953125,
      "learning_rate": 5.62296227655396e-06,
      "loss": 0.8948,
      "step": 3890
    },
    {
      "epoch": 1.8688898218676901,
      "grad_norm": 2.75,
      "learning_rate": 5.605150295354214e-06,
      "loss": 0.8241,
      "step": 3895
    },
    {
      "epoch": 1.8712889102141186,
      "grad_norm": 2.90625,
      "learning_rate": 5.5873305167886334e-06,
      "loss": 0.9893,
      "step": 3900
    },
    {
      "epoch": 1.8712889102141186,
      "eval_loss": 0.944765031337738,
      "eval_runtime": 175.8075,
      "eval_samples_per_second": 42.148,
      "eval_steps_per_second": 10.54,
      "step": 3900
    },
    {
      "epoch": 1.873687998560547,
      "grad_norm": 3.546875,
      "learning_rate": 5.569503170465196e-06,
      "loss": 0.9387,
      "step": 3905
    },
    {
      "epoch": 1.8760870869069755,
      "grad_norm": 3.40625,
      "learning_rate": 5.55166848608938e-06,
      "loss": 0.8305,
      "step": 3910
    },
    {
      "epoch": 1.8784861752534037,
      "grad_norm": 3.34375,
      "learning_rate": 5.533826693461224e-06,
      "loss": 0.8884,
      "step": 3915
    },
    {
      "epoch": 1.880885263599832,
      "grad_norm": 3.359375,
      "learning_rate": 5.515978022472349e-06,
      "loss": 0.8486,
      "step": 3920
    },
    {
      "epoch": 1.8832843519462603,
      "grad_norm": 3.65625,
      "learning_rate": 5.498122703103009e-06,
      "loss": 0.9519,
      "step": 3925
    },
    {
      "epoch": 1.8856834402926888,
      "grad_norm": 2.9375,
      "learning_rate": 5.48026096541912e-06,
      "loss": 0.9153,
      "step": 3930
    },
    {
      "epoch": 1.8880825286391172,
      "grad_norm": 3.296875,
      "learning_rate": 5.462393039569296e-06,
      "loss": 0.7888,
      "step": 3935
    },
    {
      "epoch": 1.8904816169855456,
      "grad_norm": 2.6875,
      "learning_rate": 5.44451915578189e-06,
      "loss": 0.8447,
      "step": 3940
    },
    {
      "epoch": 1.8928807053319738,
      "grad_norm": 2.6875,
      "learning_rate": 5.42663954436202e-06,
      "loss": 0.9121,
      "step": 3945
    },
    {
      "epoch": 1.895279793678402,
      "grad_norm": 3.53125,
      "learning_rate": 5.408754435688605e-06,
      "loss": 0.9036,
      "step": 3950
    },
    {
      "epoch": 1.8976788820248305,
      "grad_norm": 2.78125,
      "learning_rate": 5.390864060211399e-06,
      "loss": 0.8647,
      "step": 3955
    },
    {
      "epoch": 1.900077970371259,
      "grad_norm": 3.046875,
      "learning_rate": 5.372968648448015e-06,
      "loss": 0.8347,
      "step": 3960
    },
    {
      "epoch": 1.9024770587176874,
      "grad_norm": 3.9375,
      "learning_rate": 5.35506843098096e-06,
      "loss": 0.8615,
      "step": 3965
    },
    {
      "epoch": 1.9048761470641158,
      "grad_norm": 3.34375,
      "learning_rate": 5.337163638454661e-06,
      "loss": 0.815,
      "step": 3970
    },
    {
      "epoch": 1.907275235410544,
      "grad_norm": 3.328125,
      "learning_rate": 5.3192545015724995e-06,
      "loss": 0.9842,
      "step": 3975
    },
    {
      "epoch": 1.9096743237569722,
      "grad_norm": 2.6875,
      "learning_rate": 5.301341251093828e-06,
      "loss": 1.0199,
      "step": 3980
    },
    {
      "epoch": 1.9120734121034006,
      "grad_norm": 3.984375,
      "learning_rate": 5.2834241178310065e-06,
      "loss": 0.9157,
      "step": 3985
    },
    {
      "epoch": 1.914472500449829,
      "grad_norm": 3.359375,
      "learning_rate": 5.265503332646425e-06,
      "loss": 0.9769,
      "step": 3990
    },
    {
      "epoch": 1.9168715887962575,
      "grad_norm": 3.03125,
      "learning_rate": 5.247579126449525e-06,
      "loss": 0.8583,
      "step": 3995
    },
    {
      "epoch": 1.9192706771426857,
      "grad_norm": 3.203125,
      "learning_rate": 5.22965173019383e-06,
      "loss": 0.8915,
      "step": 4000
    },
    {
      "epoch": 1.9192706771426857,
      "eval_loss": 0.9442155361175537,
      "eval_runtime": 177.0197,
      "eval_samples_per_second": 41.86,
      "eval_steps_per_second": 10.468,
      "step": 4000
    },
    {
      "epoch": 1.9216697654891142,
      "grad_norm": 3.21875,
      "learning_rate": 5.211721374873969e-06,
      "loss": 0.8223,
      "step": 4005
    },
    {
      "epoch": 1.9240688538355424,
      "grad_norm": 2.734375,
      "learning_rate": 5.193788291522698e-06,
      "loss": 0.9664,
      "step": 4010
    },
    {
      "epoch": 1.9264679421819708,
      "grad_norm": 3.625,
      "learning_rate": 5.1758527112079194e-06,
      "loss": 0.958,
      "step": 4015
    },
    {
      "epoch": 1.9288670305283993,
      "grad_norm": 2.484375,
      "learning_rate": 5.157914865029715e-06,
      "loss": 0.8964,
      "step": 4020
    },
    {
      "epoch": 1.9312661188748277,
      "grad_norm": 3.46875,
      "learning_rate": 5.13997498411736e-06,
      "loss": 1.014,
      "step": 4025
    },
    {
      "epoch": 1.933665207221256,
      "grad_norm": 2.796875,
      "learning_rate": 5.122033299626344e-06,
      "loss": 0.8947,
      "step": 4030
    },
    {
      "epoch": 1.9360642955676843,
      "grad_norm": 3.609375,
      "learning_rate": 5.104090042735399e-06,
      "loss": 0.8083,
      "step": 4035
    },
    {
      "epoch": 1.9384633839141125,
      "grad_norm": 3.21875,
      "learning_rate": 5.08614544464352e-06,
      "loss": 0.8151,
      "step": 4040
    },
    {
      "epoch": 1.940862472260541,
      "grad_norm": 4.25,
      "learning_rate": 5.068199736566976e-06,
      "loss": 1.0128,
      "step": 4045
    },
    {
      "epoch": 1.9432615606069694,
      "grad_norm": 4.25,
      "learning_rate": 5.0502531497363435e-06,
      "loss": 0.9116,
      "step": 4050
    },
    {
      "epoch": 1.9456606489533979,
      "grad_norm": 4.46875,
      "learning_rate": 5.0323059153935235e-06,
      "loss": 0.9195,
      "step": 4055
    },
    {
      "epoch": 1.948059737299826,
      "grad_norm": 3.09375,
      "learning_rate": 5.014358264788755e-06,
      "loss": 0.8837,
      "step": 4060
    },
    {
      "epoch": 1.9504588256462543,
      "grad_norm": 3.15625,
      "learning_rate": 4.996410429177645e-06,
      "loss": 0.9059,
      "step": 4065
    },
    {
      "epoch": 1.9528579139926827,
      "grad_norm": 3.234375,
      "learning_rate": 4.9784626398181775e-06,
      "loss": 0.9118,
      "step": 4070
    },
    {
      "epoch": 1.9552570023391111,
      "grad_norm": 3.234375,
      "learning_rate": 4.96051512796775e-06,
      "loss": 0.8492,
      "step": 4075
    },
    {
      "epoch": 1.9576560906855396,
      "grad_norm": 3.171875,
      "learning_rate": 4.9425681248801756e-06,
      "loss": 0.9743,
      "step": 4080
    },
    {
      "epoch": 1.9600551790319678,
      "grad_norm": 3.15625,
      "learning_rate": 4.924621861802721e-06,
      "loss": 0.8697,
      "step": 4085
    },
    {
      "epoch": 1.9624542673783962,
      "grad_norm": 2.78125,
      "learning_rate": 4.906676569973107e-06,
      "loss": 0.8178,
      "step": 4090
    },
    {
      "epoch": 1.9648533557248244,
      "grad_norm": 3.203125,
      "learning_rate": 4.88873248061655e-06,
      "loss": 0.8874,
      "step": 4095
    },
    {
      "epoch": 1.9672524440712529,
      "grad_norm": 3.59375,
      "learning_rate": 4.870789824942766e-06,
      "loss": 0.8854,
      "step": 4100
    },
    {
      "epoch": 1.9672524440712529,
      "eval_loss": 0.9435391426086426,
      "eval_runtime": 175.8544,
      "eval_samples_per_second": 42.137,
      "eval_steps_per_second": 10.537,
      "step": 4100
    },
    {
      "epoch": 1.9696515324176813,
      "grad_norm": 3.296875,
      "learning_rate": 4.852848834143002e-06,
      "loss": 0.9725,
      "step": 4105
    },
    {
      "epoch": 1.9720506207641098,
      "grad_norm": 2.828125,
      "learning_rate": 4.834909739387048e-06,
      "loss": 0.8971,
      "step": 4110
    },
    {
      "epoch": 1.974449709110538,
      "grad_norm": 3.265625,
      "learning_rate": 4.8169727718202695e-06,
      "loss": 0.7883,
      "step": 4115
    },
    {
      "epoch": 1.9768487974569664,
      "grad_norm": 2.78125,
      "learning_rate": 4.799038162560619e-06,
      "loss": 0.9476,
      "step": 4120
    },
    {
      "epoch": 1.9792478858033946,
      "grad_norm": 3.484375,
      "learning_rate": 4.781106142695664e-06,
      "loss": 0.9393,
      "step": 4125
    },
    {
      "epoch": 1.981646974149823,
      "grad_norm": 3.078125,
      "learning_rate": 4.763176943279608e-06,
      "loss": 0.8465,
      "step": 4130
    },
    {
      "epoch": 1.9840460624962515,
      "grad_norm": 3.546875,
      "learning_rate": 4.745250795330311e-06,
      "loss": 0.8959,
      "step": 4135
    },
    {
      "epoch": 1.98644515084268,
      "grad_norm": 3.25,
      "learning_rate": 4.727327929826318e-06,
      "loss": 0.8703,
      "step": 4140
    },
    {
      "epoch": 1.9888442391891081,
      "grad_norm": 3.40625,
      "learning_rate": 4.709408577703875e-06,
      "loss": 0.9591,
      "step": 4145
    },
    {
      "epoch": 1.9912433275355363,
      "grad_norm": 3.015625,
      "learning_rate": 4.691492969853963e-06,
      "loss": 0.8204,
      "step": 4150
    },
    {
      "epoch": 1.9936424158819648,
      "grad_norm": 2.9375,
      "learning_rate": 4.673581337119313e-06,
      "loss": 0.9007,
      "step": 4155
    },
    {
      "epoch": 1.9960415042283932,
      "grad_norm": 2.9375,
      "learning_rate": 4.655673910291442e-06,
      "loss": 0.8965,
      "step": 4160
    },
    {
      "epoch": 1.9984405925748217,
      "grad_norm": 3.375,
      "learning_rate": 4.637770920107669e-06,
      "loss": 0.88,
      "step": 4165
    },
    {
      "epoch": 2.00083968092125,
      "grad_norm": 3.390625,
      "learning_rate": 4.619872597248153e-06,
      "loss": 0.9135,
      "step": 4170
    },
    {
      "epoch": 2.003238769267678,
      "grad_norm": 2.8125,
      "learning_rate": 4.6019791723329055e-06,
      "loss": 0.8235,
      "step": 4175
    },
    {
      "epoch": 2.0056378576141065,
      "grad_norm": 2.96875,
      "learning_rate": 4.584090875918837e-06,
      "loss": 0.8495,
      "step": 4180
    },
    {
      "epoch": 2.008036945960535,
      "grad_norm": 2.4375,
      "learning_rate": 4.56620793849677e-06,
      "loss": 0.8506,
      "step": 4185
    },
    {
      "epoch": 2.0104360343069634,
      "grad_norm": 3.296875,
      "learning_rate": 4.5483305904884826e-06,
      "loss": 0.8355,
      "step": 4190
    },
    {
      "epoch": 2.012835122653392,
      "grad_norm": 3.234375,
      "learning_rate": 4.530459062243726e-06,
      "loss": 0.9317,
      "step": 4195
    },
    {
      "epoch": 2.0152342109998203,
      "grad_norm": 2.96875,
      "learning_rate": 4.512593584037274e-06,
      "loss": 0.7608,
      "step": 4200
    },
    {
      "epoch": 2.0152342109998203,
      "eval_loss": 0.9447470307350159,
      "eval_runtime": 175.7692,
      "eval_samples_per_second": 42.158,
      "eval_steps_per_second": 10.542,
      "step": 4200
    },
    {
      "epoch": 2.0176332993462482,
      "grad_norm": 2.71875,
      "learning_rate": 4.494734386065933e-06,
      "loss": 0.8606,
      "step": 4205
    },
    {
      "epoch": 2.0200323876926767,
      "grad_norm": 3.40625,
      "learning_rate": 4.476881698445601e-06,
      "loss": 0.9198,
      "step": 4210
    },
    {
      "epoch": 2.022431476039105,
      "grad_norm": 2.59375,
      "learning_rate": 4.45903575120828e-06,
      "loss": 0.8639,
      "step": 4215
    },
    {
      "epoch": 2.0248305643855335,
      "grad_norm": 2.96875,
      "learning_rate": 4.441196774299129e-06,
      "loss": 0.7841,
      "step": 4220
    },
    {
      "epoch": 2.027229652731962,
      "grad_norm": 3.09375,
      "learning_rate": 4.423364997573489e-06,
      "loss": 0.8384,
      "step": 4225
    },
    {
      "epoch": 2.0296287410783904,
      "grad_norm": 2.75,
      "learning_rate": 4.405540650793931e-06,
      "loss": 0.8485,
      "step": 4230
    },
    {
      "epoch": 2.0320278294248184,
      "grad_norm": 2.703125,
      "learning_rate": 4.387723963627288e-06,
      "loss": 0.785,
      "step": 4235
    },
    {
      "epoch": 2.034426917771247,
      "grad_norm": 3.234375,
      "learning_rate": 4.369915165641701e-06,
      "loss": 0.7982,
      "step": 4240
    },
    {
      "epoch": 2.0368260061176753,
      "grad_norm": 2.828125,
      "learning_rate": 4.352114486303657e-06,
      "loss": 0.8574,
      "step": 4245
    },
    {
      "epoch": 2.0392250944641037,
      "grad_norm": 2.640625,
      "learning_rate": 4.334322154975037e-06,
      "loss": 0.8414,
      "step": 4250
    },
    {
      "epoch": 2.041624182810532,
      "grad_norm": 3.484375,
      "learning_rate": 4.3165384009101535e-06,
      "loss": 0.8034,
      "step": 4255
    },
    {
      "epoch": 2.04402327115696,
      "grad_norm": 2.8125,
      "learning_rate": 4.298763453252805e-06,
      "loss": 0.8283,
      "step": 4260
    },
    {
      "epoch": 2.0464223595033886,
      "grad_norm": 3.0,
      "learning_rate": 4.280997541033315e-06,
      "loss": 0.6966,
      "step": 4265
    },
    {
      "epoch": 2.048821447849817,
      "grad_norm": 3.125,
      "learning_rate": 4.263240893165592e-06,
      "loss": 0.8893,
      "step": 4270
    },
    {
      "epoch": 2.0512205361962454,
      "grad_norm": 3.21875,
      "learning_rate": 4.2454937384441665e-06,
      "loss": 0.8797,
      "step": 4275
    },
    {
      "epoch": 2.053619624542674,
      "grad_norm": 3.890625,
      "learning_rate": 4.227756305541253e-06,
      "loss": 0.8327,
      "step": 4280
    },
    {
      "epoch": 2.0560187128891023,
      "grad_norm": 2.75,
      "learning_rate": 4.210028823003802e-06,
      "loss": 0.8423,
      "step": 4285
    },
    {
      "epoch": 2.0584178012355303,
      "grad_norm": 2.6875,
      "learning_rate": 4.192311519250548e-06,
      "loss": 0.8146,
      "step": 4290
    },
    {
      "epoch": 2.0608168895819587,
      "grad_norm": 4.0625,
      "learning_rate": 4.174604622569076e-06,
      "loss": 0.8386,
      "step": 4295
    },
    {
      "epoch": 2.063215977928387,
      "grad_norm": 2.625,
      "learning_rate": 4.156908361112876e-06,
      "loss": 0.796,
      "step": 4300
    },
    {
      "epoch": 2.063215977928387,
      "eval_loss": 0.9463717937469482,
      "eval_runtime": 175.7455,
      "eval_samples_per_second": 42.163,
      "eval_steps_per_second": 10.544,
      "step": 4300
    },
    {
      "epoch": 2.0656150662748156,
      "grad_norm": 3.046875,
      "learning_rate": 4.139222962898401e-06,
      "loss": 0.7608,
      "step": 4305
    },
    {
      "epoch": 2.068014154621244,
      "grad_norm": 2.421875,
      "learning_rate": 4.121548655802132e-06,
      "loss": 0.8912,
      "step": 4310
    },
    {
      "epoch": 2.0704132429676725,
      "grad_norm": 3.015625,
      "learning_rate": 4.103885667557642e-06,
      "loss": 0.7804,
      "step": 4315
    },
    {
      "epoch": 2.0728123313141005,
      "grad_norm": 2.796875,
      "learning_rate": 4.086234225752657e-06,
      "loss": 0.876,
      "step": 4320
    },
    {
      "epoch": 2.075211419660529,
      "grad_norm": 2.890625,
      "learning_rate": 4.068594557826132e-06,
      "loss": 0.7589,
      "step": 4325
    },
    {
      "epoch": 2.0776105080069573,
      "grad_norm": 3.46875,
      "learning_rate": 4.0509668910653114e-06,
      "loss": 0.8852,
      "step": 4330
    },
    {
      "epoch": 2.080009596353386,
      "grad_norm": 2.984375,
      "learning_rate": 4.033351452602807e-06,
      "loss": 0.9057,
      "step": 4335
    },
    {
      "epoch": 2.082408684699814,
      "grad_norm": 2.515625,
      "learning_rate": 4.0157484694136645e-06,
      "loss": 0.8005,
      "step": 4340
    },
    {
      "epoch": 2.0848077730462427,
      "grad_norm": 2.9375,
      "learning_rate": 3.998158168312453e-06,
      "loss": 0.8047,
      "step": 4345
    },
    {
      "epoch": 2.0872068613926706,
      "grad_norm": 3.09375,
      "learning_rate": 3.98058077595032e-06,
      "loss": 0.7325,
      "step": 4350
    },
    {
      "epoch": 2.089605949739099,
      "grad_norm": 2.53125,
      "learning_rate": 3.9630165188120945e-06,
      "loss": 0.7172,
      "step": 4355
    },
    {
      "epoch": 2.0920050380855275,
      "grad_norm": 2.78125,
      "learning_rate": 3.945465623213352e-06,
      "loss": 0.776,
      "step": 4360
    },
    {
      "epoch": 2.094404126431956,
      "grad_norm": 3.28125,
      "learning_rate": 3.927928315297508e-06,
      "loss": 0.8868,
      "step": 4365
    },
    {
      "epoch": 2.0968032147783844,
      "grad_norm": 2.8125,
      "learning_rate": 3.9104048210328965e-06,
      "loss": 0.9293,
      "step": 4370
    },
    {
      "epoch": 2.0992023031248124,
      "grad_norm": 2.765625,
      "learning_rate": 3.892895366209867e-06,
      "loss": 0.7971,
      "step": 4375
    },
    {
      "epoch": 2.101601391471241,
      "grad_norm": 2.734375,
      "learning_rate": 3.875400176437867e-06,
      "loss": 0.861,
      "step": 4380
    },
    {
      "epoch": 2.1040004798176692,
      "grad_norm": 3.203125,
      "learning_rate": 3.8579194771425414e-06,
      "loss": 0.867,
      "step": 4385
    },
    {
      "epoch": 2.1063995681640977,
      "grad_norm": 2.984375,
      "learning_rate": 3.840453493562823e-06,
      "loss": 0.8466,
      "step": 4390
    },
    {
      "epoch": 2.108798656510526,
      "grad_norm": 2.859375,
      "learning_rate": 3.8230024507480375e-06,
      "loss": 0.7525,
      "step": 4395
    },
    {
      "epoch": 2.1111977448569546,
      "grad_norm": 3.3125,
      "learning_rate": 3.80556657355499e-06,
      "loss": 0.9225,
      "step": 4400
    },
    {
      "epoch": 2.1111977448569546,
      "eval_loss": 0.9466894865036011,
      "eval_runtime": 175.7759,
      "eval_samples_per_second": 42.156,
      "eval_steps_per_second": 10.542,
      "step": 4400
    },
    {
      "epoch": 2.1135968332033825,
      "grad_norm": 2.9375,
      "learning_rate": 3.788146086645084e-06,
      "loss": 0.7987,
      "step": 4405
    },
    {
      "epoch": 2.115995921549811,
      "grad_norm": 3.15625,
      "learning_rate": 3.7707412144814154e-06,
      "loss": 0.7869,
      "step": 4410
    },
    {
      "epoch": 2.1183950098962394,
      "grad_norm": 2.671875,
      "learning_rate": 3.7533521813258845e-06,
      "loss": 0.9135,
      "step": 4415
    },
    {
      "epoch": 2.120794098242668,
      "grad_norm": 2.65625,
      "learning_rate": 3.735979211236309e-06,
      "loss": 0.8726,
      "step": 4420
    },
    {
      "epoch": 2.1231931865890963,
      "grad_norm": 2.890625,
      "learning_rate": 3.7186225280635286e-06,
      "loss": 0.8516,
      "step": 4425
    },
    {
      "epoch": 2.1255922749355243,
      "grad_norm": 2.625,
      "learning_rate": 3.701282355448531e-06,
      "loss": 0.7885,
      "step": 4430
    },
    {
      "epoch": 2.1279913632819527,
      "grad_norm": 3.015625,
      "learning_rate": 3.6839589168195605e-06,
      "loss": 0.8337,
      "step": 4435
    },
    {
      "epoch": 2.130390451628381,
      "grad_norm": 2.90625,
      "learning_rate": 3.666652435389248e-06,
      "loss": 0.9509,
      "step": 4440
    },
    {
      "epoch": 2.1327895399748096,
      "grad_norm": 3.09375,
      "learning_rate": 3.6493631341517274e-06,
      "loss": 0.8803,
      "step": 4445
    },
    {
      "epoch": 2.135188628321238,
      "grad_norm": 3.34375,
      "learning_rate": 3.632091235879769e-06,
      "loss": 0.8565,
      "step": 4450
    },
    {
      "epoch": 2.1375877166676664,
      "grad_norm": 3.15625,
      "learning_rate": 3.614836963121902e-06,
      "loss": 0.865,
      "step": 4455
    },
    {
      "epoch": 2.1399868050140944,
      "grad_norm": 2.515625,
      "learning_rate": 3.5976005381995573e-06,
      "loss": 0.8987,
      "step": 4460
    },
    {
      "epoch": 2.142385893360523,
      "grad_norm": 2.828125,
      "learning_rate": 3.5803821832041857e-06,
      "loss": 0.7778,
      "step": 4465
    },
    {
      "epoch": 2.1447849817069513,
      "grad_norm": 2.875,
      "learning_rate": 3.563182119994417e-06,
      "loss": 0.7757,
      "step": 4470
    },
    {
      "epoch": 2.1471840700533797,
      "grad_norm": 3.0625,
      "learning_rate": 3.5460005701931864e-06,
      "loss": 0.7319,
      "step": 4475
    },
    {
      "epoch": 2.149583158399808,
      "grad_norm": 3.125,
      "learning_rate": 3.5288377551848855e-06,
      "loss": 0.9632,
      "step": 4480
    },
    {
      "epoch": 2.1519822467462366,
      "grad_norm": 3.5,
      "learning_rate": 3.511693896112503e-06,
      "loss": 0.8698,
      "step": 4485
    },
    {
      "epoch": 2.1543813350926646,
      "grad_norm": 2.921875,
      "learning_rate": 3.4945692138747898e-06,
      "loss": 0.8772,
      "step": 4490
    },
    {
      "epoch": 2.156780423439093,
      "grad_norm": 3.40625,
      "learning_rate": 3.4774639291233937e-06,
      "loss": 0.8039,
      "step": 4495
    },
    {
      "epoch": 2.1591795117855215,
      "grad_norm": 3.109375,
      "learning_rate": 3.4603782622600307e-06,
      "loss": 0.9901,
      "step": 4500
    },
    {
      "epoch": 2.1591795117855215,
      "eval_loss": 0.9466681480407715,
      "eval_runtime": 177.7999,
      "eval_samples_per_second": 41.676,
      "eval_steps_per_second": 10.422,
      "step": 4500
    },
    {
      "epoch": 2.16157860013195,
      "grad_norm": 3.390625,
      "learning_rate": 3.4433124334336383e-06,
      "loss": 0.8059,
      "step": 4505
    },
    {
      "epoch": 2.1639776884783783,
      "grad_norm": 3.03125,
      "learning_rate": 3.426266662537544e-06,
      "loss": 0.9415,
      "step": 4510
    },
    {
      "epoch": 2.166376776824807,
      "grad_norm": 2.78125,
      "learning_rate": 3.409241169206623e-06,
      "loss": 0.8322,
      "step": 4515
    },
    {
      "epoch": 2.1687758651712348,
      "grad_norm": 2.8125,
      "learning_rate": 3.3922361728144804e-06,
      "loss": 0.7832,
      "step": 4520
    },
    {
      "epoch": 2.171174953517663,
      "grad_norm": 2.78125,
      "learning_rate": 3.375251892470611e-06,
      "loss": 0.7673,
      "step": 4525
    },
    {
      "epoch": 2.1735740418640916,
      "grad_norm": 3.390625,
      "learning_rate": 3.358288547017591e-06,
      "loss": 0.8652,
      "step": 4530
    },
    {
      "epoch": 2.17597313021052,
      "grad_norm": 3.265625,
      "learning_rate": 3.3413463550282437e-06,
      "loss": 0.8218,
      "step": 4535
    },
    {
      "epoch": 2.1783722185569485,
      "grad_norm": 2.96875,
      "learning_rate": 3.324425534802835e-06,
      "loss": 0.8349,
      "step": 4540
    },
    {
      "epoch": 2.1807713069033765,
      "grad_norm": 2.640625,
      "learning_rate": 3.307526304366251e-06,
      "loss": 0.8862,
      "step": 4545
    },
    {
      "epoch": 2.183170395249805,
      "grad_norm": 2.953125,
      "learning_rate": 3.290648881465196e-06,
      "loss": 0.7924,
      "step": 4550
    },
    {
      "epoch": 2.1855694835962334,
      "grad_norm": 3.484375,
      "learning_rate": 3.2737934835653827e-06,
      "loss": 0.8539,
      "step": 4555
    },
    {
      "epoch": 2.187968571942662,
      "grad_norm": 2.703125,
      "learning_rate": 3.2569603278487335e-06,
      "loss": 0.7829,
      "step": 4560
    },
    {
      "epoch": 2.1903676602890902,
      "grad_norm": 2.578125,
      "learning_rate": 3.2401496312105786e-06,
      "loss": 0.8173,
      "step": 4565
    },
    {
      "epoch": 2.1927667486355187,
      "grad_norm": 3.421875,
      "learning_rate": 3.223361610256861e-06,
      "loss": 0.9411,
      "step": 4570
    },
    {
      "epoch": 2.1951658369819467,
      "grad_norm": 3.90625,
      "learning_rate": 3.2065964813013533e-06,
      "loss": 0.9485,
      "step": 4575
    },
    {
      "epoch": 2.197564925328375,
      "grad_norm": 2.6875,
      "learning_rate": 3.1898544603628563e-06,
      "loss": 0.8482,
      "step": 4580
    },
    {
      "epoch": 2.1999640136748035,
      "grad_norm": 2.609375,
      "learning_rate": 3.1731357631624304e-06,
      "loss": 0.7985,
      "step": 4585
    },
    {
      "epoch": 2.202363102021232,
      "grad_norm": 3.4375,
      "learning_rate": 3.1564406051206063e-06,
      "loss": 0.8566,
      "step": 4590
    },
    {
      "epoch": 2.2047621903676604,
      "grad_norm": 26.625,
      "learning_rate": 3.1397692013546154e-06,
      "loss": 0.8479,
      "step": 4595
    },
    {
      "epoch": 2.207161278714089,
      "grad_norm": 2.65625,
      "learning_rate": 3.1231217666756085e-06,
      "loss": 0.9263,
      "step": 4600
    },
    {
      "epoch": 2.207161278714089,
      "eval_loss": 0.9467650651931763,
      "eval_runtime": 175.6879,
      "eval_samples_per_second": 42.177,
      "eval_steps_per_second": 10.547,
      "step": 4600
    },
    {
      "epoch": 2.209560367060517,
      "grad_norm": 3.3125,
      "learning_rate": 3.106498515585905e-06,
      "loss": 0.8065,
      "step": 4605
    },
    {
      "epoch": 2.2119594554069453,
      "grad_norm": 2.796875,
      "learning_rate": 3.089899662276208e-06,
      "loss": 0.7885,
      "step": 4610
    },
    {
      "epoch": 2.2143585437533737,
      "grad_norm": 3.171875,
      "learning_rate": 3.073325420622862e-06,
      "loss": 0.9224,
      "step": 4615
    },
    {
      "epoch": 2.216757632099802,
      "grad_norm": 3.578125,
      "learning_rate": 3.056776004185086e-06,
      "loss": 0.8299,
      "step": 4620
    },
    {
      "epoch": 2.2191567204462306,
      "grad_norm": 2.875,
      "learning_rate": 3.0402516262022312e-06,
      "loss": 0.698,
      "step": 4625
    },
    {
      "epoch": 2.221555808792659,
      "grad_norm": 3.171875,
      "learning_rate": 3.0237524995910205e-06,
      "loss": 0.8001,
      "step": 4630
    },
    {
      "epoch": 2.223954897139087,
      "grad_norm": 3.5625,
      "learning_rate": 3.0072788369428195e-06,
      "loss": 0.7655,
      "step": 4635
    },
    {
      "epoch": 2.2263539854855154,
      "grad_norm": 2.484375,
      "learning_rate": 2.9908308505208864e-06,
      "loss": 0.8143,
      "step": 4640
    },
    {
      "epoch": 2.228753073831944,
      "grad_norm": 4.125,
      "learning_rate": 2.9744087522576444e-06,
      "loss": 0.8349,
      "step": 4645
    },
    {
      "epoch": 2.2311521621783723,
      "grad_norm": 3.28125,
      "learning_rate": 2.9580127537519432e-06,
      "loss": 0.8516,
      "step": 4650
    },
    {
      "epoch": 2.2335512505248007,
      "grad_norm": 3.78125,
      "learning_rate": 2.9416430662663432e-06,
      "loss": 0.8004,
      "step": 4655
    },
    {
      "epoch": 2.2359503388712287,
      "grad_norm": 2.78125,
      "learning_rate": 2.9252999007243786e-06,
      "loss": 0.8981,
      "step": 4660
    },
    {
      "epoch": 2.238349427217657,
      "grad_norm": 2.953125,
      "learning_rate": 2.908983467707856e-06,
      "loss": 0.7687,
      "step": 4665
    },
    {
      "epoch": 2.2407485155640856,
      "grad_norm": 2.984375,
      "learning_rate": 2.8926939774541273e-06,
      "loss": 0.7922,
      "step": 4670
    },
    {
      "epoch": 2.243147603910514,
      "grad_norm": 2.84375,
      "learning_rate": 2.876431639853396e-06,
      "loss": 0.7277,
      "step": 4675
    },
    {
      "epoch": 2.2455466922569425,
      "grad_norm": 2.640625,
      "learning_rate": 2.860196664445988e-06,
      "loss": 0.7337,
      "step": 4680
    },
    {
      "epoch": 2.247945780603371,
      "grad_norm": 3.265625,
      "learning_rate": 2.8439892604196834e-06,
      "loss": 0.9654,
      "step": 4685
    },
    {
      "epoch": 2.250344868949799,
      "grad_norm": 2.890625,
      "learning_rate": 2.8278096366069945e-06,
      "loss": 0.875,
      "step": 4690
    },
    {
      "epoch": 2.2527439572962273,
      "grad_norm": 3.375,
      "learning_rate": 2.811658001482489e-06,
      "loss": 0.8906,
      "step": 4695
    },
    {
      "epoch": 2.2551430456426558,
      "grad_norm": 3.28125,
      "learning_rate": 2.7955345631600993e-06,
      "loss": 0.7735,
      "step": 4700
    },
    {
      "epoch": 2.2551430456426558,
      "eval_loss": 0.9466607570648193,
      "eval_runtime": 179.2943,
      "eval_samples_per_second": 41.329,
      "eval_steps_per_second": 10.335,
      "step": 4700
    },
    {
      "epoch": 2.257542133989084,
      "grad_norm": 3.03125,
      "learning_rate": 2.779439529390441e-06,
      "loss": 0.8957,
      "step": 4705
    },
    {
      "epoch": 2.2599412223355126,
      "grad_norm": 3.03125,
      "learning_rate": 2.7633731075581406e-06,
      "loss": 0.8791,
      "step": 4710
    },
    {
      "epoch": 2.2623403106819406,
      "grad_norm": 3.125,
      "learning_rate": 2.747335504679156e-06,
      "loss": 0.8097,
      "step": 4715
    },
    {
      "epoch": 2.264739399028369,
      "grad_norm": 2.59375,
      "learning_rate": 2.7313269273981135e-06,
      "loss": 0.8695,
      "step": 4720
    },
    {
      "epoch": 2.2671384873747975,
      "grad_norm": 2.984375,
      "learning_rate": 2.7153475819856425e-06,
      "loss": 0.8596,
      "step": 4725
    },
    {
      "epoch": 2.269537575721226,
      "grad_norm": 4.40625,
      "learning_rate": 2.6993976743357264e-06,
      "loss": 0.8044,
      "step": 4730
    },
    {
      "epoch": 2.2719366640676544,
      "grad_norm": 3.1875,
      "learning_rate": 2.6834774099630323e-06,
      "loss": 0.8534,
      "step": 4735
    },
    {
      "epoch": 2.274335752414083,
      "grad_norm": 3.328125,
      "learning_rate": 2.667586994000283e-06,
      "loss": 0.8286,
      "step": 4740
    },
    {
      "epoch": 2.2767348407605112,
      "grad_norm": 3.40625,
      "learning_rate": 2.651726631195599e-06,
      "loss": 0.871,
      "step": 4745
    },
    {
      "epoch": 2.2791339291069392,
      "grad_norm": 2.734375,
      "learning_rate": 2.635896525909868e-06,
      "loss": 0.8352,
      "step": 4750
    },
    {
      "epoch": 2.2815330174533677,
      "grad_norm": 3.03125,
      "learning_rate": 2.620096882114106e-06,
      "loss": 0.7521,
      "step": 4755
    },
    {
      "epoch": 2.283932105799796,
      "grad_norm": 3.1875,
      "learning_rate": 2.6043279033868407e-06,
      "loss": 0.9201,
      "step": 4760
    },
    {
      "epoch": 2.2863311941462245,
      "grad_norm": 3.125,
      "learning_rate": 2.5885897929114662e-06,
      "loss": 0.8577,
      "step": 4765
    },
    {
      "epoch": 2.288730282492653,
      "grad_norm": 2.953125,
      "learning_rate": 2.572882753473654e-06,
      "loss": 0.8168,
      "step": 4770
    },
    {
      "epoch": 2.291129370839081,
      "grad_norm": 2.953125,
      "learning_rate": 2.5572069874587157e-06,
      "loss": 0.8826,
      "step": 4775
    },
    {
      "epoch": 2.2935284591855094,
      "grad_norm": 3.21875,
      "learning_rate": 2.5415626968490075e-06,
      "loss": 0.8244,
      "step": 4780
    },
    {
      "epoch": 2.295927547531938,
      "grad_norm": 3.046875,
      "learning_rate": 2.5259500832213224e-06,
      "loss": 0.7957,
      "step": 4785
    },
    {
      "epoch": 2.2983266358783663,
      "grad_norm": 2.953125,
      "learning_rate": 2.510369347744303e-06,
      "loss": 0.8546,
      "step": 4790
    },
    {
      "epoch": 2.3007257242247947,
      "grad_norm": 2.8125,
      "learning_rate": 2.4948206911758284e-06,
      "loss": 0.8521,
      "step": 4795
    },
    {
      "epoch": 2.303124812571223,
      "grad_norm": 3.234375,
      "learning_rate": 2.4793043138604546e-06,
      "loss": 0.8454,
      "step": 4800
    },
    {
      "epoch": 2.303124812571223,
      "eval_loss": 0.94642174243927,
      "eval_runtime": 193.1371,
      "eval_samples_per_second": 38.367,
      "eval_steps_per_second": 9.594,
      "step": 4800
    },
    {
      "epoch": 2.305523900917651,
      "grad_norm": 3.5625,
      "learning_rate": 2.46382041572681e-06,
      "loss": 0.7795,
      "step": 4805
    },
    {
      "epoch": 2.3079229892640796,
      "grad_norm": 3.125,
      "learning_rate": 2.4483691962850327e-06,
      "loss": 0.9321,
      "step": 4810
    },
    {
      "epoch": 2.310322077610508,
      "grad_norm": 3.421875,
      "learning_rate": 2.432950854624193e-06,
      "loss": 0.8356,
      "step": 4815
    },
    {
      "epoch": 2.3127211659569364,
      "grad_norm": 3.34375,
      "learning_rate": 2.4175655894097335e-06,
      "loss": 0.7706,
      "step": 4820
    },
    {
      "epoch": 2.315120254303365,
      "grad_norm": 2.734375,
      "learning_rate": 2.4022135988809025e-06,
      "loss": 0.791,
      "step": 4825
    },
    {
      "epoch": 2.317519342649793,
      "grad_norm": 3.265625,
      "learning_rate": 2.3868950808482107e-06,
      "loss": 0.7859,
      "step": 4830
    },
    {
      "epoch": 2.3199184309962213,
      "grad_norm": 2.953125,
      "learning_rate": 2.371610232690869e-06,
      "loss": 0.8915,
      "step": 4835
    },
    {
      "epoch": 2.3223175193426497,
      "grad_norm": 3.0625,
      "learning_rate": 2.3563592513542543e-06,
      "loss": 0.9431,
      "step": 4840
    },
    {
      "epoch": 2.324716607689078,
      "grad_norm": 2.9375,
      "learning_rate": 2.3411423333473683e-06,
      "loss": 0.8549,
      "step": 4845
    },
    {
      "epoch": 2.3271156960355066,
      "grad_norm": 3.328125,
      "learning_rate": 2.325959674740306e-06,
      "loss": 0.8688,
      "step": 4850
    },
    {
      "epoch": 2.329514784381935,
      "grad_norm": 3.703125,
      "learning_rate": 2.3108114711617335e-06,
      "loss": 0.9216,
      "step": 4855
    },
    {
      "epoch": 2.331913872728363,
      "grad_norm": 2.71875,
      "learning_rate": 2.29569791779636e-06,
      "loss": 0.8822,
      "step": 4860
    },
    {
      "epoch": 2.3343129610747915,
      "grad_norm": 3.0,
      "learning_rate": 2.2806192093824277e-06,
      "loss": 0.837,
      "step": 4865
    },
    {
      "epoch": 2.33671204942122,
      "grad_norm": 3.109375,
      "learning_rate": 2.265575540209198e-06,
      "loss": 0.8704,
      "step": 4870
    },
    {
      "epoch": 2.3391111377676483,
      "grad_norm": 2.75,
      "learning_rate": 2.250567104114461e-06,
      "loss": 0.7557,
      "step": 4875
    },
    {
      "epoch": 2.3415102261140768,
      "grad_norm": 3.015625,
      "learning_rate": 2.235594094482014e-06,
      "loss": 0.756,
      "step": 4880
    },
    {
      "epoch": 2.3439093144605048,
      "grad_norm": 3.03125,
      "learning_rate": 2.220656704239198e-06,
      "loss": 0.887,
      "step": 4885
    },
    {
      "epoch": 2.346308402806933,
      "grad_norm": 3.390625,
      "learning_rate": 2.2057551258543893e-06,
      "loss": 0.799,
      "step": 4890
    },
    {
      "epoch": 2.3487074911533616,
      "grad_norm": 3.921875,
      "learning_rate": 2.1908895513345314e-06,
      "loss": 0.9533,
      "step": 4895
    },
    {
      "epoch": 2.35110657949979,
      "grad_norm": 3.328125,
      "learning_rate": 2.176060172222654e-06,
      "loss": 0.8562,
      "step": 4900
    },
    {
      "epoch": 2.35110657949979,
      "eval_loss": 0.9465652108192444,
      "eval_runtime": 187.1163,
      "eval_samples_per_second": 39.601,
      "eval_steps_per_second": 9.903,
      "step": 4900
    },
    {
      "epoch": 2.3535056678462185,
      "grad_norm": 3.03125,
      "learning_rate": 2.1612671795954193e-06,
      "loss": 0.9049,
      "step": 4905
    },
    {
      "epoch": 2.355904756192647,
      "grad_norm": 2.75,
      "learning_rate": 2.146510764060633e-06,
      "loss": 0.7734,
      "step": 4910
    },
    {
      "epoch": 2.3583038445390754,
      "grad_norm": 2.953125,
      "learning_rate": 2.1317911157548187e-06,
      "loss": 0.8603,
      "step": 4915
    },
    {
      "epoch": 2.3607029328855034,
      "grad_norm": 3.203125,
      "learning_rate": 2.1171084243407487e-06,
      "loss": 0.877,
      "step": 4920
    },
    {
      "epoch": 2.363102021231932,
      "grad_norm": 2.875,
      "learning_rate": 2.1024628790050038e-06,
      "loss": 0.9646,
      "step": 4925
    },
    {
      "epoch": 2.3655011095783602,
      "grad_norm": 2.546875,
      "learning_rate": 2.0878546684555384e-06,
      "loss": 0.9592,
      "step": 4930
    },
    {
      "epoch": 2.3679001979247887,
      "grad_norm": 3.6875,
      "learning_rate": 2.073283980919254e-06,
      "loss": 0.7567,
      "step": 4935
    },
    {
      "epoch": 2.370299286271217,
      "grad_norm": 2.65625,
      "learning_rate": 2.0587510041395553e-06,
      "loss": 0.7229,
      "step": 4940
    },
    {
      "epoch": 2.372698374617645,
      "grad_norm": 3.234375,
      "learning_rate": 2.044255925373956e-06,
      "loss": 0.8459,
      "step": 4945
    },
    {
      "epoch": 2.3750974629640735,
      "grad_norm": 3.296875,
      "learning_rate": 2.029798931391646e-06,
      "loss": 0.8717,
      "step": 4950
    },
    {
      "epoch": 2.377496551310502,
      "grad_norm": 3.125,
      "learning_rate": 2.015380208471096e-06,
      "loss": 0.8185,
      "step": 4955
    },
    {
      "epoch": 2.3798956396569304,
      "grad_norm": 3.09375,
      "learning_rate": 2.0009999423976527e-06,
      "loss": 0.8071,
      "step": 4960
    },
    {
      "epoch": 2.382294728003359,
      "grad_norm": 2.734375,
      "learning_rate": 1.9866583184611452e-06,
      "loss": 0.8988,
      "step": 4965
    },
    {
      "epoch": 2.3846938163497873,
      "grad_norm": 2.78125,
      "learning_rate": 1.9723555214534983e-06,
      "loss": 0.8504,
      "step": 4970
    },
    {
      "epoch": 2.3870929046962153,
      "grad_norm": 3.078125,
      "learning_rate": 1.958091735666356e-06,
      "loss": 0.8153,
      "step": 4975
    },
    {
      "epoch": 2.3894919930426437,
      "grad_norm": 2.453125,
      "learning_rate": 1.9438671448886963e-06,
      "loss": 0.8636,
      "step": 4980
    },
    {
      "epoch": 2.391891081389072,
      "grad_norm": 3.125,
      "learning_rate": 1.929681932404473e-06,
      "loss": 0.7966,
      "step": 4985
    },
    {
      "epoch": 2.3942901697355006,
      "grad_norm": 3.859375,
      "learning_rate": 1.915536280990249e-06,
      "loss": 0.7249,
      "step": 4990
    },
    {
      "epoch": 2.396689258081929,
      "grad_norm": 3.546875,
      "learning_rate": 1.9014303729128402e-06,
      "loss": 0.8304,
      "step": 4995
    },
    {
      "epoch": 2.399088346428357,
      "grad_norm": 3.015625,
      "learning_rate": 1.8873643899269761e-06,
      "loss": 0.8923,
      "step": 5000
    },
    {
      "epoch": 2.399088346428357,
      "eval_loss": 0.946355402469635,
      "eval_runtime": 175.772,
      "eval_samples_per_second": 42.157,
      "eval_steps_per_second": 10.542,
      "step": 5000
    },
    {
      "epoch": 2.4014874347747854,
      "grad_norm": 3.203125,
      "learning_rate": 1.8733385132729453e-06,
      "loss": 0.9251,
      "step": 5005
    },
    {
      "epoch": 2.403886523121214,
      "grad_norm": 4.28125,
      "learning_rate": 1.859352923674267e-06,
      "loss": 0.9606,
      "step": 5010
    },
    {
      "epoch": 2.4062856114676423,
      "grad_norm": 2.8125,
      "learning_rate": 1.8454078013353604e-06,
      "loss": 0.9109,
      "step": 5015
    },
    {
      "epoch": 2.4086846998140707,
      "grad_norm": 3.265625,
      "learning_rate": 1.8315033259392313e-06,
      "loss": 0.7988,
      "step": 5020
    },
    {
      "epoch": 2.411083788160499,
      "grad_norm": 3.0625,
      "learning_rate": 1.8176396766451353e-06,
      "loss": 0.8556,
      "step": 5025
    },
    {
      "epoch": 2.4134828765069276,
      "grad_norm": 3.4375,
      "learning_rate": 1.803817032086298e-06,
      "loss": 0.8578,
      "step": 5030
    },
    {
      "epoch": 2.4158819648533556,
      "grad_norm": 3.9375,
      "learning_rate": 1.7900355703675893e-06,
      "loss": 0.8618,
      "step": 5035
    },
    {
      "epoch": 2.418281053199784,
      "grad_norm": 3.34375,
      "learning_rate": 1.7762954690632416e-06,
      "loss": 0.8395,
      "step": 5040
    },
    {
      "epoch": 2.4206801415462125,
      "grad_norm": 4.09375,
      "learning_rate": 1.7625969052145557e-06,
      "loss": 0.885,
      "step": 5045
    },
    {
      "epoch": 2.423079229892641,
      "grad_norm": 3.296875,
      "learning_rate": 1.7489400553276281e-06,
      "loss": 0.7281,
      "step": 5050
    },
    {
      "epoch": 2.4254783182390693,
      "grad_norm": 2.71875,
      "learning_rate": 1.73532509537106e-06,
      "loss": 0.8529,
      "step": 5055
    },
    {
      "epoch": 2.4278774065854973,
      "grad_norm": 3.34375,
      "learning_rate": 1.7217522007737108e-06,
      "loss": 0.9102,
      "step": 5060
    },
    {
      "epoch": 2.4302764949319258,
      "grad_norm": 2.921875,
      "learning_rate": 1.7082215464224228e-06,
      "loss": 0.8904,
      "step": 5065
    },
    {
      "epoch": 2.432675583278354,
      "grad_norm": 3.0625,
      "learning_rate": 1.6947333066597721e-06,
      "loss": 0.883,
      "step": 5070
    },
    {
      "epoch": 2.4350746716247826,
      "grad_norm": 3.359375,
      "learning_rate": 1.6812876552818236e-06,
      "loss": 0.9472,
      "step": 5075
    },
    {
      "epoch": 2.437473759971211,
      "grad_norm": 3.09375,
      "learning_rate": 1.6678847655358899e-06,
      "loss": 0.9375,
      "step": 5080
    },
    {
      "epoch": 2.4398728483176395,
      "grad_norm": 2.796875,
      "learning_rate": 1.6545248101182992e-06,
      "loss": 0.8775,
      "step": 5085
    },
    {
      "epoch": 2.4422719366640675,
      "grad_norm": 4.96875,
      "learning_rate": 1.641207961172175e-06,
      "loss": 0.8681,
      "step": 5090
    },
    {
      "epoch": 2.444671025010496,
      "grad_norm": 2.40625,
      "learning_rate": 1.627934390285207e-06,
      "loss": 0.82,
      "step": 5095
    },
    {
      "epoch": 2.4470701133569244,
      "grad_norm": 3.5625,
      "learning_rate": 1.614704268487451e-06,
      "loss": 0.7529,
      "step": 5100
    },
    {
      "epoch": 2.4470701133569244,
      "eval_loss": 0.9462727904319763,
      "eval_runtime": 175.7988,
      "eval_samples_per_second": 42.15,
      "eval_steps_per_second": 10.54,
      "step": 5100
    },
    {
      "epoch": 2.449469201703353,
      "grad_norm": 3.4375,
      "learning_rate": 1.60151776624912e-06,
      "loss": 0.7723,
      "step": 5105
    },
    {
      "epoch": 2.4518682900497812,
      "grad_norm": 2.84375,
      "learning_rate": 1.5883750534783876e-06,
      "loss": 0.7418,
      "step": 5110
    },
    {
      "epoch": 2.4542673783962092,
      "grad_norm": 2.71875,
      "learning_rate": 1.5752762995192e-06,
      "loss": 0.823,
      "step": 5115
    },
    {
      "epoch": 2.4566664667426377,
      "grad_norm": 3.125,
      "learning_rate": 1.5622216731490975e-06,
      "loss": 0.889,
      "step": 5120
    },
    {
      "epoch": 2.459065555089066,
      "grad_norm": 3.265625,
      "learning_rate": 1.549211342577031e-06,
      "loss": 0.7486,
      "step": 5125
    },
    {
      "epoch": 2.4614646434354945,
      "grad_norm": 3.140625,
      "learning_rate": 1.536245475441201e-06,
      "loss": 0.8351,
      "step": 5130
    },
    {
      "epoch": 2.463863731781923,
      "grad_norm": 2.875,
      "learning_rate": 1.523324238806902e-06,
      "loss": 0.7765,
      "step": 5135
    },
    {
      "epoch": 2.4662628201283514,
      "grad_norm": 4.03125,
      "learning_rate": 1.5104477991643517e-06,
      "loss": 0.9074,
      "step": 5140
    },
    {
      "epoch": 2.4686619084747794,
      "grad_norm": 2.921875,
      "learning_rate": 1.4976163224265728e-06,
      "loss": 0.8319,
      "step": 5145
    },
    {
      "epoch": 2.471060996821208,
      "grad_norm": 3.484375,
      "learning_rate": 1.4848299739272304e-06,
      "loss": 0.7772,
      "step": 5150
    },
    {
      "epoch": 2.4734600851676363,
      "grad_norm": 2.515625,
      "learning_rate": 1.4720889184185155e-06,
      "loss": 0.8531,
      "step": 5155
    },
    {
      "epoch": 2.4758591735140647,
      "grad_norm": 3.796875,
      "learning_rate": 1.459393320069018e-06,
      "loss": 0.7806,
      "step": 5160
    },
    {
      "epoch": 2.478258261860493,
      "grad_norm": 3.53125,
      "learning_rate": 1.4467433424616155e-06,
      "loss": 1.0109,
      "step": 5165
    },
    {
      "epoch": 2.480657350206921,
      "grad_norm": 3.0,
      "learning_rate": 1.4341391485913536e-06,
      "loss": 0.8122,
      "step": 5170
    },
    {
      "epoch": 2.4830564385533496,
      "grad_norm": 2.921875,
      "learning_rate": 1.4215809008633636e-06,
      "loss": 0.8517,
      "step": 5175
    },
    {
      "epoch": 2.485455526899778,
      "grad_norm": 3.15625,
      "learning_rate": 1.409068761090755e-06,
      "loss": 0.7956,
      "step": 5180
    },
    {
      "epoch": 2.4878546152462064,
      "grad_norm": 3.59375,
      "learning_rate": 1.3966028904925372e-06,
      "loss": 0.7672,
      "step": 5185
    },
    {
      "epoch": 2.490253703592635,
      "grad_norm": 2.609375,
      "learning_rate": 1.384183449691539e-06,
      "loss": 0.7573,
      "step": 5190
    },
    {
      "epoch": 2.4926527919390633,
      "grad_norm": 3.625,
      "learning_rate": 1.3718105987123482e-06,
      "loss": 0.8696,
      "step": 5195
    },
    {
      "epoch": 2.4950518802854917,
      "grad_norm": 2.703125,
      "learning_rate": 1.3594844969792304e-06,
      "loss": 0.8421,
      "step": 5200
    },
    {
      "epoch": 2.4950518802854917,
      "eval_loss": 0.9463452100753784,
      "eval_runtime": 175.6343,
      "eval_samples_per_second": 42.19,
      "eval_steps_per_second": 10.55,
      "step": 5200
    },
    {
      "epoch": 2.4974509686319197,
      "grad_norm": 3.546875,
      "learning_rate": 1.347205303314098e-06,
      "loss": 0.9172,
      "step": 5205
    },
    {
      "epoch": 2.499850056978348,
      "grad_norm": 3.28125,
      "learning_rate": 1.3349731759344469e-06,
      "loss": 0.8079,
      "step": 5210
    },
    {
      "epoch": 2.5022491453247766,
      "grad_norm": 3.03125,
      "learning_rate": 1.3227882724513253e-06,
      "loss": 0.752,
      "step": 5215
    },
    {
      "epoch": 2.504648233671205,
      "grad_norm": 2.96875,
      "learning_rate": 1.3106507498672999e-06,
      "loss": 0.8044,
      "step": 5220
    },
    {
      "epoch": 2.507047322017633,
      "grad_norm": 3.6875,
      "learning_rate": 1.2985607645744352e-06,
      "loss": 0.8345,
      "step": 5225
    },
    {
      "epoch": 2.5094464103640615,
      "grad_norm": 2.984375,
      "learning_rate": 1.286518472352276e-06,
      "loss": 0.7665,
      "step": 5230
    },
    {
      "epoch": 2.51184549871049,
      "grad_norm": 2.71875,
      "learning_rate": 1.2745240283658456e-06,
      "loss": 0.8592,
      "step": 5235
    },
    {
      "epoch": 2.5142445870569183,
      "grad_norm": 2.734375,
      "learning_rate": 1.2625775871636376e-06,
      "loss": 0.7824,
      "step": 5240
    },
    {
      "epoch": 2.5166436754033468,
      "grad_norm": 3.59375,
      "learning_rate": 1.2506793026756314e-06,
      "loss": 0.827,
      "step": 5245
    },
    {
      "epoch": 2.519042763749775,
      "grad_norm": 3.296875,
      "learning_rate": 1.2388293282113067e-06,
      "loss": 0.8529,
      "step": 5250
    },
    {
      "epoch": 2.5214418520962036,
      "grad_norm": 3.53125,
      "learning_rate": 1.2270278164576688e-06,
      "loss": 0.8162,
      "step": 5255
    },
    {
      "epoch": 2.523840940442632,
      "grad_norm": 2.84375,
      "learning_rate": 1.2152749194772783e-06,
      "loss": 0.8861,
      "step": 5260
    },
    {
      "epoch": 2.52624002878906,
      "grad_norm": 2.6875,
      "learning_rate": 1.2035707887062981e-06,
      "loss": 0.7473,
      "step": 5265
    },
    {
      "epoch": 2.5286391171354885,
      "grad_norm": 2.78125,
      "learning_rate": 1.1919155749525357e-06,
      "loss": 0.7434,
      "step": 5270
    },
    {
      "epoch": 2.531038205481917,
      "grad_norm": 2.890625,
      "learning_rate": 1.1803094283935002e-06,
      "loss": 0.835,
      "step": 5275
    },
    {
      "epoch": 2.5334372938283454,
      "grad_norm": 3.125,
      "learning_rate": 1.1687524985744764e-06,
      "loss": 0.86,
      "step": 5280
    },
    {
      "epoch": 2.5358363821747734,
      "grad_norm": 3.078125,
      "learning_rate": 1.1572449344065816e-06,
      "loss": 0.84,
      "step": 5285
    },
    {
      "epoch": 2.538235470521202,
      "grad_norm": 3.6875,
      "learning_rate": 1.1457868841648656e-06,
      "loss": 0.8313,
      "step": 5290
    },
    {
      "epoch": 2.5406345588676302,
      "grad_norm": 2.75,
      "learning_rate": 1.1343784954863847e-06,
      "loss": 0.794,
      "step": 5295
    },
    {
      "epoch": 2.5430336472140587,
      "grad_norm": 2.8125,
      "learning_rate": 1.123019915368308e-06,
      "loss": 0.8578,
      "step": 5300
    },
    {
      "epoch": 2.5430336472140587,
      "eval_loss": 0.9463096261024475,
      "eval_runtime": 175.7279,
      "eval_samples_per_second": 42.167,
      "eval_steps_per_second": 10.545,
      "step": 5300
    },
    {
      "epoch": 2.545432735560487,
      "grad_norm": 2.5625,
      "learning_rate": 1.1117112901660193e-06,
      "loss": 0.7754,
      "step": 5305
    },
    {
      "epoch": 2.5478318239069155,
      "grad_norm": 3.96875,
      "learning_rate": 1.1004527655912383e-06,
      "loss": 0.9777,
      "step": 5310
    },
    {
      "epoch": 2.550230912253344,
      "grad_norm": 2.953125,
      "learning_rate": 1.0892444867101288e-06,
      "loss": 0.8786,
      "step": 5315
    },
    {
      "epoch": 2.552630000599772,
      "grad_norm": 3.1875,
      "learning_rate": 1.0780865979414463e-06,
      "loss": 0.9006,
      "step": 5320
    },
    {
      "epoch": 2.5550290889462004,
      "grad_norm": 2.9375,
      "learning_rate": 1.0669792430546655e-06,
      "loss": 0.7912,
      "step": 5325
    },
    {
      "epoch": 2.557428177292629,
      "grad_norm": 2.78125,
      "learning_rate": 1.0559225651681332e-06,
      "loss": 0.8563,
      "step": 5330
    },
    {
      "epoch": 2.5598272656390573,
      "grad_norm": 2.765625,
      "learning_rate": 1.0449167067472205e-06,
      "loss": 0.8267,
      "step": 5335
    },
    {
      "epoch": 2.5622263539854853,
      "grad_norm": 3.140625,
      "learning_rate": 1.0339618096024946e-06,
      "loss": 0.7603,
      "step": 5340
    },
    {
      "epoch": 2.5646254423319137,
      "grad_norm": 3.171875,
      "learning_rate": 1.0230580148878777e-06,
      "loss": 0.9592,
      "step": 5345
    },
    {
      "epoch": 2.567024530678342,
      "grad_norm": 2.5625,
      "learning_rate": 1.0122054630988454e-06,
      "loss": 0.8984,
      "step": 5350
    },
    {
      "epoch": 2.5694236190247706,
      "grad_norm": 2.828125,
      "learning_rate": 1.0014042940706031e-06,
      "loss": 0.8046,
      "step": 5355
    },
    {
      "epoch": 2.571822707371199,
      "grad_norm": 2.65625,
      "learning_rate": 9.9065464697629e-07,
      "loss": 0.727,
      "step": 5360
    },
    {
      "epoch": 2.5742217957176274,
      "grad_norm": 2.90625,
      "learning_rate": 9.799566603251847e-07,
      "loss": 0.7937,
      "step": 5365
    },
    {
      "epoch": 2.576620884064056,
      "grad_norm": 2.59375,
      "learning_rate": 9.693104719609213e-07,
      "loss": 0.7619,
      "step": 5370
    },
    {
      "epoch": 2.579019972410484,
      "grad_norm": 2.9375,
      "learning_rate": 9.587162190597104e-07,
      "loss": 0.8086,
      "step": 5375
    },
    {
      "epoch": 2.5814190607569123,
      "grad_norm": 3.03125,
      "learning_rate": 9.481740381285782e-07,
      "loss": 0.9133,
      "step": 5380
    },
    {
      "epoch": 2.5838181491033407,
      "grad_norm": 2.84375,
      "learning_rate": 9.376840650035995e-07,
      "loss": 0.8183,
      "step": 5385
    },
    {
      "epoch": 2.586217237449769,
      "grad_norm": 3.109375,
      "learning_rate": 9.272464348481513e-07,
      "loss": 0.7881,
      "step": 5390
    },
    {
      "epoch": 2.5886163257961976,
      "grad_norm": 3.328125,
      "learning_rate": 9.168612821511729e-07,
      "loss": 0.7716,
      "step": 5395
    },
    {
      "epoch": 2.5910154141426256,
      "grad_norm": 3.171875,
      "learning_rate": 9.065287407254292e-07,
      "loss": 0.8143,
      "step": 5400
    },
    {
      "epoch": 2.5910154141426256,
      "eval_loss": 0.9463847279548645,
      "eval_runtime": 175.7856,
      "eval_samples_per_second": 42.154,
      "eval_steps_per_second": 10.541,
      "step": 5400
    },
    {
      "epoch": 2.593414502489054,
      "grad_norm": 3.0,
      "learning_rate": 8.962489437057892e-07,
      "loss": 0.8354,
      "step": 5405
    },
    {
      "epoch": 2.5958135908354825,
      "grad_norm": 3.03125,
      "learning_rate": 8.860220235475136e-07,
      "loss": 0.757,
      "step": 5410
    },
    {
      "epoch": 2.598212679181911,
      "grad_norm": 2.78125,
      "learning_rate": 8.758481120245355e-07,
      "loss": 0.8845,
      "step": 5415
    },
    {
      "epoch": 2.6006117675283393,
      "grad_norm": 3.34375,
      "learning_rate": 8.6572734022778e-07,
      "loss": 0.8356,
      "step": 5420
    },
    {
      "epoch": 2.6030108558747678,
      "grad_norm": 3.40625,
      "learning_rate": 8.556598385634645e-07,
      "loss": 0.8755,
      "step": 5425
    },
    {
      "epoch": 2.605409944221196,
      "grad_norm": 3.546875,
      "learning_rate": 8.456457367514154e-07,
      "loss": 0.8882,
      "step": 5430
    },
    {
      "epoch": 2.607809032567624,
      "grad_norm": 3.125,
      "learning_rate": 8.356851638234087e-07,
      "loss": 0.8723,
      "step": 5435
    },
    {
      "epoch": 2.6102081209140526,
      "grad_norm": 2.921875,
      "learning_rate": 8.257782481214954e-07,
      "loss": 0.7859,
      "step": 5440
    },
    {
      "epoch": 2.612607209260481,
      "grad_norm": 3.21875,
      "learning_rate": 8.159251172963545e-07,
      "loss": 0.8549,
      "step": 5445
    },
    {
      "epoch": 2.6150062976069095,
      "grad_norm": 3.390625,
      "learning_rate": 8.061258983056452e-07,
      "loss": 0.9265,
      "step": 5450
    },
    {
      "epoch": 2.6174053859533375,
      "grad_norm": 2.859375,
      "learning_rate": 7.963807174123772e-07,
      "loss": 0.6679,
      "step": 5455
    },
    {
      "epoch": 2.619804474299766,
      "grad_norm": 2.84375,
      "learning_rate": 7.866897001832696e-07,
      "loss": 0.6948,
      "step": 5460
    },
    {
      "epoch": 2.6222035626461944,
      "grad_norm": 2.828125,
      "learning_rate": 7.770529714871527e-07,
      "loss": 0.7636,
      "step": 5465
    },
    {
      "epoch": 2.624602650992623,
      "grad_norm": 2.96875,
      "learning_rate": 7.674706554933414e-07,
      "loss": 0.9437,
      "step": 5470
    },
    {
      "epoch": 2.6270017393390512,
      "grad_norm": 3.609375,
      "learning_rate": 7.579428756700463e-07,
      "loss": 0.8655,
      "step": 5475
    },
    {
      "epoch": 2.6294008276854797,
      "grad_norm": 3.4375,
      "learning_rate": 7.484697547827763e-07,
      "loss": 0.9146,
      "step": 5480
    },
    {
      "epoch": 2.631799916031908,
      "grad_norm": 2.90625,
      "learning_rate": 7.390514148927619e-07,
      "loss": 0.9243,
      "step": 5485
    },
    {
      "epoch": 2.634199004378336,
      "grad_norm": 3.34375,
      "learning_rate": 7.296879773553784e-07,
      "loss": 0.812,
      "step": 5490
    },
    {
      "epoch": 2.6365980927247645,
      "grad_norm": 3.484375,
      "learning_rate": 7.203795628185856e-07,
      "loss": 0.9371,
      "step": 5495
    },
    {
      "epoch": 2.638997181071193,
      "grad_norm": 3.109375,
      "learning_rate": 7.111262912213707e-07,
      "loss": 0.8117,
      "step": 5500
    },
    {
      "epoch": 2.638997181071193,
      "eval_loss": 0.9463163614273071,
      "eval_runtime": 175.7487,
      "eval_samples_per_second": 42.162,
      "eval_steps_per_second": 10.543,
      "step": 5500
    },
    {
      "epoch": 2.6413962694176214,
      "grad_norm": 2.984375,
      "learning_rate": 7.019282817922029e-07,
      "loss": 0.7939,
      "step": 5505
    },
    {
      "epoch": 2.6437953577640494,
      "grad_norm": 3.40625,
      "learning_rate": 6.927856530474985e-07,
      "loss": 0.7907,
      "step": 5510
    },
    {
      "epoch": 2.646194446110478,
      "grad_norm": 4.09375,
      "learning_rate": 6.836985227900944e-07,
      "loss": 0.9429,
      "step": 5515
    },
    {
      "epoch": 2.6485935344569063,
      "grad_norm": 3.734375,
      "learning_rate": 6.746670081077266e-07,
      "loss": 0.8834,
      "step": 5520
    },
    {
      "epoch": 2.6509926228033347,
      "grad_norm": 3.4375,
      "learning_rate": 6.656912253715281e-07,
      "loss": 0.8487,
      "step": 5525
    },
    {
      "epoch": 2.653391711149763,
      "grad_norm": 3.5,
      "learning_rate": 6.567712902345208e-07,
      "loss": 0.9643,
      "step": 5530
    },
    {
      "epoch": 2.6557907994961916,
      "grad_norm": 3.15625,
      "learning_rate": 6.479073176301332e-07,
      "loss": 0.8767,
      "step": 5535
    },
    {
      "epoch": 2.65818988784262,
      "grad_norm": 3.140625,
      "learning_rate": 6.390994217707142e-07,
      "loss": 0.8471,
      "step": 5540
    },
    {
      "epoch": 2.6605889761890484,
      "grad_norm": 2.8125,
      "learning_rate": 6.303477161460647e-07,
      "loss": 0.8318,
      "step": 5545
    },
    {
      "epoch": 2.6629880645354764,
      "grad_norm": 2.453125,
      "learning_rate": 6.216523135219715e-07,
      "loss": 0.8511,
      "step": 5550
    },
    {
      "epoch": 2.665387152881905,
      "grad_norm": 3.578125,
      "learning_rate": 6.130133259387633e-07,
      "loss": 0.8708,
      "step": 5555
    },
    {
      "epoch": 2.6677862412283333,
      "grad_norm": 3.265625,
      "learning_rate": 6.044308647098512e-07,
      "loss": 0.7847,
      "step": 5560
    },
    {
      "epoch": 2.6701853295747617,
      "grad_norm": 3.375,
      "learning_rate": 5.959050404203109e-07,
      "loss": 0.9311,
      "step": 5565
    },
    {
      "epoch": 2.6725844179211897,
      "grad_norm": 2.859375,
      "learning_rate": 5.874359629254511e-07,
      "loss": 0.8072,
      "step": 5570
    },
    {
      "epoch": 2.674983506267618,
      "grad_norm": 2.8125,
      "learning_rate": 5.79023741349391e-07,
      "loss": 0.8887,
      "step": 5575
    },
    {
      "epoch": 2.6773825946140466,
      "grad_norm": 3.171875,
      "learning_rate": 5.706684840836674e-07,
      "loss": 0.8971,
      "step": 5580
    },
    {
      "epoch": 2.679781682960475,
      "grad_norm": 2.640625,
      "learning_rate": 5.623702987858293e-07,
      "loss": 0.7582,
      "step": 5585
    },
    {
      "epoch": 2.6821807713069035,
      "grad_norm": 2.90625,
      "learning_rate": 5.541292923780516e-07,
      "loss": 0.7481,
      "step": 5590
    },
    {
      "epoch": 2.684579859653332,
      "grad_norm": 3.125,
      "learning_rate": 5.459455710457601e-07,
      "loss": 0.8242,
      "step": 5595
    },
    {
      "epoch": 2.6869789479997603,
      "grad_norm": 2.796875,
      "learning_rate": 5.378192402362653e-07,
      "loss": 0.861,
      "step": 5600
    },
    {
      "epoch": 2.6869789479997603,
      "eval_loss": 0.9463757276535034,
      "eval_runtime": 175.8536,
      "eval_samples_per_second": 42.137,
      "eval_steps_per_second": 10.537,
      "step": 5600
    },
    {
      "epoch": 2.6893780363461883,
      "grad_norm": 2.8125,
      "learning_rate": 5.29750404657392e-07,
      "loss": 0.6736,
      "step": 5605
    },
    {
      "epoch": 2.6917771246926168,
      "grad_norm": 2.890625,
      "learning_rate": 5.217391682761469e-07,
      "loss": 0.8122,
      "step": 5610
    },
    {
      "epoch": 2.694176213039045,
      "grad_norm": 2.9375,
      "learning_rate": 5.137856343173675e-07,
      "loss": 0.8597,
      "step": 5615
    },
    {
      "epoch": 2.6965753013854736,
      "grad_norm": 2.515625,
      "learning_rate": 5.058899052623933e-07,
      "loss": 0.6924,
      "step": 5620
    },
    {
      "epoch": 2.6989743897319016,
      "grad_norm": 3.296875,
      "learning_rate": 4.980520828477509e-07,
      "loss": 0.9464,
      "step": 5625
    },
    {
      "epoch": 2.70137347807833,
      "grad_norm": 2.6875,
      "learning_rate": 4.902722680638356e-07,
      "loss": 0.8155,
      "step": 5630
    },
    {
      "epoch": 2.7037725664247585,
      "grad_norm": 3.140625,
      "learning_rate": 4.825505611536163e-07,
      "loss": 0.8895,
      "step": 5635
    },
    {
      "epoch": 2.706171654771187,
      "grad_norm": 3.25,
      "learning_rate": 4.7488706161134266e-07,
      "loss": 0.9353,
      "step": 5640
    },
    {
      "epoch": 2.7085707431176154,
      "grad_norm": 3.125,
      "learning_rate": 4.672818681812591e-07,
      "loss": 0.8699,
      "step": 5645
    },
    {
      "epoch": 2.710969831464044,
      "grad_norm": 3.125,
      "learning_rate": 4.597350788563376e-07,
      "loss": 0.9011,
      "step": 5650
    },
    {
      "epoch": 2.7133689198104722,
      "grad_norm": 2.65625,
      "learning_rate": 4.522467908770106e-07,
      "loss": 0.8674,
      "step": 5655
    },
    {
      "epoch": 2.7157680081569002,
      "grad_norm": 2.46875,
      "learning_rate": 4.448171007299229e-07,
      "loss": 0.7725,
      "step": 5660
    },
    {
      "epoch": 2.7181670965033287,
      "grad_norm": 3.234375,
      "learning_rate": 4.3744610414668265e-07,
      "loss": 0.8388,
      "step": 5665
    },
    {
      "epoch": 2.720566184849757,
      "grad_norm": 2.953125,
      "learning_rate": 4.3013389610263636e-07,
      "loss": 0.8038,
      "step": 5670
    },
    {
      "epoch": 2.7229652731961855,
      "grad_norm": 3.0625,
      "learning_rate": 4.2288057081563247e-07,
      "loss": 0.8106,
      "step": 5675
    },
    {
      "epoch": 2.725364361542614,
      "grad_norm": 3.015625,
      "learning_rate": 4.156862217448215e-07,
      "loss": 0.7503,
      "step": 5680
    },
    {
      "epoch": 2.727763449889042,
      "grad_norm": 3.296875,
      "learning_rate": 4.0855094158944066e-07,
      "loss": 0.8085,
      "step": 5685
    },
    {
      "epoch": 2.7301625382354704,
      "grad_norm": 3.0,
      "learning_rate": 4.014748222876258e-07,
      "loss": 0.781,
      "step": 5690
    },
    {
      "epoch": 2.732561626581899,
      "grad_norm": 2.71875,
      "learning_rate": 3.9445795501522276e-07,
      "loss": 0.8548,
      "step": 5695
    },
    {
      "epoch": 2.7349607149283273,
      "grad_norm": 3.421875,
      "learning_rate": 3.875004301846186e-07,
      "loss": 0.8415,
      "step": 5700
    },
    {
      "epoch": 2.7349607149283273,
      "eval_loss": 0.946337103843689,
      "eval_runtime": 175.7065,
      "eval_samples_per_second": 42.173,
      "eval_steps_per_second": 10.546,
      "step": 5700
    },
    {
      "epoch": 2.7373598032747557,
      "grad_norm": 2.984375,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.8101,
      "step": 5705
    },
    {
      "epoch": 2.739758891621184,
      "grad_norm": 2.9375,
      "learning_rate": 3.737637656740423e-07,
      "loss": 0.8063,
      "step": 5710
    },
    {
      "epoch": 2.7421579799676126,
      "grad_norm": 3.15625,
      "learning_rate": 3.6698480299109273e-07,
      "loss": 0.7506,
      "step": 5715
    },
    {
      "epoch": 2.7445570683140406,
      "grad_norm": 2.84375,
      "learning_rate": 3.602655367416968e-07,
      "loss": 0.8546,
      "step": 5720
    },
    {
      "epoch": 2.746956156660469,
      "grad_norm": 3.09375,
      "learning_rate": 3.5360605350365006e-07,
      "loss": 0.8406,
      "step": 5725
    },
    {
      "epoch": 2.7493552450068974,
      "grad_norm": 3.15625,
      "learning_rate": 3.470064390844402e-07,
      "loss": 0.8724,
      "step": 5730
    },
    {
      "epoch": 2.751754333353326,
      "grad_norm": 2.703125,
      "learning_rate": 3.404667785201454e-07,
      "loss": 0.7694,
      "step": 5735
    },
    {
      "epoch": 2.754153421699754,
      "grad_norm": 2.625,
      "learning_rate": 3.3398715607433794e-07,
      "loss": 0.8349,
      "step": 5740
    },
    {
      "epoch": 2.7565525100461823,
      "grad_norm": 3.171875,
      "learning_rate": 3.2756765523700165e-07,
      "loss": 0.8109,
      "step": 5745
    },
    {
      "epoch": 2.7589515983926107,
      "grad_norm": 3.296875,
      "learning_rate": 3.2120835872344547e-07,
      "loss": 0.7918,
      "step": 5750
    },
    {
      "epoch": 2.761350686739039,
      "grad_norm": 2.671875,
      "learning_rate": 3.1490934847325406e-07,
      "loss": 0.7042,
      "step": 5755
    },
    {
      "epoch": 2.7637497750854676,
      "grad_norm": 3.921875,
      "learning_rate": 3.0867070564921665e-07,
      "loss": 0.8729,
      "step": 5760
    },
    {
      "epoch": 2.766148863431896,
      "grad_norm": 2.765625,
      "learning_rate": 3.0249251063629137e-07,
      "loss": 0.8672,
      "step": 5765
    },
    {
      "epoch": 2.7685479517783245,
      "grad_norm": 2.875,
      "learning_rate": 2.9637484304056387e-07,
      "loss": 0.8593,
      "step": 5770
    },
    {
      "epoch": 2.7709470401247525,
      "grad_norm": 3.578125,
      "learning_rate": 2.9031778168822466e-07,
      "loss": 0.7651,
      "step": 5775
    },
    {
      "epoch": 2.773346128471181,
      "grad_norm": 3.21875,
      "learning_rate": 2.843214046245507e-07,
      "loss": 0.7516,
      "step": 5780
    },
    {
      "epoch": 2.7757452168176093,
      "grad_norm": 3.328125,
      "learning_rate": 2.783857891129055e-07,
      "loss": 0.834,
      "step": 5785
    },
    {
      "epoch": 2.7781443051640378,
      "grad_norm": 3.828125,
      "learning_rate": 2.725110116337354e-07,
      "loss": 0.842,
      "step": 5790
    },
    {
      "epoch": 2.7805433935104658,
      "grad_norm": 2.875,
      "learning_rate": 2.6669714788358946e-07,
      "loss": 0.8287,
      "step": 5795
    },
    {
      "epoch": 2.782942481856894,
      "grad_norm": 3.203125,
      "learning_rate": 2.60944272774144e-07,
      "loss": 0.7846,
      "step": 5800
    },
    {
      "epoch": 2.782942481856894,
      "eval_loss": 0.946326732635498,
      "eval_runtime": 175.7382,
      "eval_samples_per_second": 42.165,
      "eval_steps_per_second": 10.544,
      "step": 5800
    },
    {
      "epoch": 2.7853415702033226,
      "grad_norm": 2.578125,
      "learning_rate": 2.552524604312351e-07,
      "loss": 0.8889,
      "step": 5805
    },
    {
      "epoch": 2.787740658549751,
      "grad_norm": 2.96875,
      "learning_rate": 2.4962178419390357e-07,
      "loss": 0.8157,
      "step": 5810
    },
    {
      "epoch": 2.7901397468961795,
      "grad_norm": 2.78125,
      "learning_rate": 2.440523166134562e-07,
      "loss": 0.7996,
      "step": 5815
    },
    {
      "epoch": 2.792538835242608,
      "grad_norm": 2.59375,
      "learning_rate": 2.3854412945251757e-07,
      "loss": 0.8163,
      "step": 5820
    },
    {
      "epoch": 2.7949379235890364,
      "grad_norm": 2.671875,
      "learning_rate": 2.3309729368412193e-07,
      "loss": 0.7702,
      "step": 5825
    },
    {
      "epoch": 2.7973370119354644,
      "grad_norm": 3.078125,
      "learning_rate": 2.2771187949078455e-07,
      "loss": 0.8732,
      "step": 5830
    },
    {
      "epoch": 2.799736100281893,
      "grad_norm": 2.78125,
      "learning_rate": 2.223879562636061e-07,
      "loss": 0.7786,
      "step": 5835
    },
    {
      "epoch": 2.8021351886283212,
      "grad_norm": 4.15625,
      "learning_rate": 2.1712559260137434e-07,
      "loss": 0.7568,
      "step": 5840
    },
    {
      "epoch": 2.8045342769747497,
      "grad_norm": 2.921875,
      "learning_rate": 2.1192485630968374e-07,
      "loss": 1.008,
      "step": 5845
    },
    {
      "epoch": 2.806933365321178,
      "grad_norm": 3.125,
      "learning_rate": 2.0678581440005617e-07,
      "loss": 0.8665,
      "step": 5850
    },
    {
      "epoch": 2.809332453667606,
      "grad_norm": 3.515625,
      "learning_rate": 2.0170853308908388e-07,
      "loss": 0.8216,
      "step": 5855
    },
    {
      "epoch": 2.8117315420140345,
      "grad_norm": 3.125,
      "learning_rate": 1.966930777975734e-07,
      "loss": 0.7792,
      "step": 5860
    },
    {
      "epoch": 2.814130630360463,
      "grad_norm": 3.140625,
      "learning_rate": 1.9173951314970018e-07,
      "loss": 0.826,
      "step": 5865
    },
    {
      "epoch": 2.8165297187068914,
      "grad_norm": 2.875,
      "learning_rate": 1.8684790297218037e-07,
      "loss": 0.8359,
      "step": 5870
    },
    {
      "epoch": 2.81892880705332,
      "grad_norm": 3.515625,
      "learning_rate": 1.8201831029344585e-07,
      "loss": 0.8419,
      "step": 5875
    },
    {
      "epoch": 2.8213278953997483,
      "grad_norm": 3.28125,
      "learning_rate": 1.7725079734283223e-07,
      "loss": 0.8791,
      "step": 5880
    },
    {
      "epoch": 2.8237269837461767,
      "grad_norm": 2.984375,
      "learning_rate": 1.7254542554977771e-07,
      "loss": 0.8924,
      "step": 5885
    },
    {
      "epoch": 2.8261260720926047,
      "grad_norm": 2.9375,
      "learning_rate": 1.679022555430304e-07,
      "loss": 0.9084,
      "step": 5890
    },
    {
      "epoch": 2.828525160439033,
      "grad_norm": 3.65625,
      "learning_rate": 1.6332134714986848e-07,
      "loss": 0.8278,
      "step": 5895
    },
    {
      "epoch": 2.8309242487854616,
      "grad_norm": 2.484375,
      "learning_rate": 1.5880275939533063e-07,
      "loss": 0.7605,
      "step": 5900
    },
    {
      "epoch": 2.8309242487854616,
      "eval_loss": 0.9463862776756287,
      "eval_runtime": 177.234,
      "eval_samples_per_second": 41.809,
      "eval_steps_per_second": 10.455,
      "step": 5900
    },
    {
      "epoch": 2.83332333713189,
      "grad_norm": 2.625,
      "learning_rate": 1.5434655050145077e-07,
      "loss": 0.9009,
      "step": 5905
    },
    {
      "epoch": 2.835722425478318,
      "grad_norm": 2.8125,
      "learning_rate": 1.499527778865123e-07,
      "loss": 0.8512,
      "step": 5910
    },
    {
      "epoch": 2.8381215138247464,
      "grad_norm": 3.28125,
      "learning_rate": 1.4562149816430616e-07,
      "loss": 0.6809,
      "step": 5915
    },
    {
      "epoch": 2.840520602171175,
      "grad_norm": 2.625,
      "learning_rate": 1.413527671434023e-07,
      "loss": 0.8097,
      "step": 5920
    },
    {
      "epoch": 2.8429196905176033,
      "grad_norm": 2.703125,
      "learning_rate": 1.3714663982642984e-07,
      "loss": 0.8181,
      "step": 5925
    },
    {
      "epoch": 2.8453187788640317,
      "grad_norm": 2.859375,
      "learning_rate": 1.3300317040936927e-07,
      "loss": 0.8673,
      "step": 5930
    },
    {
      "epoch": 2.84771786721046,
      "grad_norm": 3.078125,
      "learning_rate": 1.2892241228085355e-07,
      "loss": 0.8847,
      "step": 5935
    },
    {
      "epoch": 2.8501169555568886,
      "grad_norm": 2.8125,
      "learning_rate": 1.2490441802148036e-07,
      "loss": 0.8172,
      "step": 5940
    },
    {
      "epoch": 2.8525160439033166,
      "grad_norm": 2.78125,
      "learning_rate": 1.2094923940313308e-07,
      "loss": 0.8514,
      "step": 5945
    },
    {
      "epoch": 2.854915132249745,
      "grad_norm": 3.609375,
      "learning_rate": 1.1705692738831654e-07,
      "loss": 0.8143,
      "step": 5950
    },
    {
      "epoch": 2.8573142205961735,
      "grad_norm": 3.203125,
      "learning_rate": 1.1322753212949844e-07,
      "loss": 0.8274,
      "step": 5955
    },
    {
      "epoch": 2.859713308942602,
      "grad_norm": 3.75,
      "learning_rate": 1.0946110296846447e-07,
      "loss": 0.8353,
      "step": 5960
    },
    {
      "epoch": 2.8621123972890303,
      "grad_norm": 3.15625,
      "learning_rate": 1.057576884356798e-07,
      "loss": 0.8685,
      "step": 5965
    },
    {
      "epoch": 2.8645114856354583,
      "grad_norm": 4.0,
      "learning_rate": 1.0211733624966802e-07,
      "loss": 0.8467,
      "step": 5970
    },
    {
      "epoch": 2.8669105739818868,
      "grad_norm": 3.453125,
      "learning_rate": 9.854009331639214e-08,
      "loss": 0.8935,
      "step": 5975
    },
    {
      "epoch": 2.869309662328315,
      "grad_norm": 3.015625,
      "learning_rate": 9.502600572865284e-08,
      "loss": 0.8584,
      "step": 5980
    },
    {
      "epoch": 2.8717087506747436,
      "grad_norm": 2.796875,
      "learning_rate": 9.157511876549286e-08,
      "loss": 0.7601,
      "step": 5985
    },
    {
      "epoch": 2.874107839021172,
      "grad_norm": 3.359375,
      "learning_rate": 8.818747689161688e-08,
      "loss": 0.8792,
      "step": 5990
    },
    {
      "epoch": 2.8765069273676005,
      "grad_norm": 3.078125,
      "learning_rate": 8.486312375681205e-08,
      "loss": 0.7975,
      "step": 5995
    },
    {
      "epoch": 2.878906015714029,
      "grad_norm": 2.625,
      "learning_rate": 8.160210219539333e-08,
      "loss": 0.8721,
      "step": 6000
    },
    {
      "epoch": 2.878906015714029,
      "eval_loss": 0.9464093446731567,
      "eval_runtime": 183.9687,
      "eval_samples_per_second": 40.279,
      "eval_steps_per_second": 10.072,
      "step": 6000
    },
    {
      "epoch": 2.881305104060457,
      "grad_norm": 3.578125,
      "learning_rate": 7.840445422564735e-08,
      "loss": 0.8902,
      "step": 6005
    },
    {
      "epoch": 2.8837041924068854,
      "grad_norm": 2.84375,
      "learning_rate": 7.527022104928893e-08,
      "loss": 0.8215,
      "step": 6010
    },
    {
      "epoch": 2.886103280753314,
      "grad_norm": 2.734375,
      "learning_rate": 7.219944305093596e-08,
      "loss": 0.8512,
      "step": 6015
    },
    {
      "epoch": 2.8885023690997422,
      "grad_norm": 2.875,
      "learning_rate": 6.919215979758476e-08,
      "loss": 0.8271,
      "step": 6020
    },
    {
      "epoch": 2.89090145744617,
      "grad_norm": 2.796875,
      "learning_rate": 6.624841003810056e-08,
      "loss": 0.7854,
      "step": 6025
    },
    {
      "epoch": 2.8933005457925987,
      "grad_norm": 3.21875,
      "learning_rate": 6.336823170272011e-08,
      "loss": 0.8432,
      "step": 6030
    },
    {
      "epoch": 2.895699634139027,
      "grad_norm": 3.140625,
      "learning_rate": 6.055166190256145e-08,
      "loss": 0.872,
      "step": 6035
    },
    {
      "epoch": 2.8980987224854555,
      "grad_norm": 2.953125,
      "learning_rate": 5.779873692914606e-08,
      "loss": 0.8188,
      "step": 6040
    },
    {
      "epoch": 2.900497810831884,
      "grad_norm": 3.578125,
      "learning_rate": 5.5109492253933025e-08,
      "loss": 0.7619,
      "step": 6045
    },
    {
      "epoch": 2.9028968991783124,
      "grad_norm": 2.59375,
      "learning_rate": 5.2483962527857813e-08,
      "loss": 0.704,
      "step": 6050
    },
    {
      "epoch": 2.905295987524741,
      "grad_norm": 3.296875,
      "learning_rate": 4.992218158088979e-08,
      "loss": 0.7787,
      "step": 6055
    },
    {
      "epoch": 2.907695075871169,
      "grad_norm": 3.0625,
      "learning_rate": 4.7424182421594854e-08,
      "loss": 0.8102,
      "step": 6060
    },
    {
      "epoch": 2.9100941642175973,
      "grad_norm": 3.390625,
      "learning_rate": 4.498999723670905e-08,
      "loss": 0.8945,
      "step": 6065
    },
    {
      "epoch": 2.9124932525640257,
      "grad_norm": 2.609375,
      "learning_rate": 4.2619657390726154e-08,
      "loss": 0.8516,
      "step": 6070
    },
    {
      "epoch": 2.914892340910454,
      "grad_norm": 3.171875,
      "learning_rate": 4.0313193425492446e-08,
      "loss": 0.7551,
      "step": 6075
    },
    {
      "epoch": 2.917291429256882,
      "grad_norm": 3.03125,
      "learning_rate": 3.8070635059811455e-08,
      "loss": 0.8393,
      "step": 6080
    },
    {
      "epoch": 2.9196905176033106,
      "grad_norm": 3.046875,
      "learning_rate": 3.589201118906427e-08,
      "loss": 0.8361,
      "step": 6085
    },
    {
      "epoch": 2.922089605949739,
      "grad_norm": 3.078125,
      "learning_rate": 3.3777349884834275e-08,
      "loss": 0.8572,
      "step": 6090
    },
    {
      "epoch": 2.9244886942961674,
      "grad_norm": 3.15625,
      "learning_rate": 3.1726678394547464e-08,
      "loss": 0.8325,
      "step": 6095
    },
    {
      "epoch": 2.926887782642596,
      "grad_norm": 3.0,
      "learning_rate": 2.9740023141120455e-08,
      "loss": 0.8566,
      "step": 6100
    },
    {
      "epoch": 2.926887782642596,
      "eval_loss": 0.9463550448417664,
      "eval_runtime": 178.3552,
      "eval_samples_per_second": 41.546,
      "eval_steps_per_second": 10.389,
      "step": 6100
    },
    {
      "epoch": 2.9292868709890243,
      "grad_norm": 3.0625,
      "learning_rate": 2.7817409722621368e-08,
      "loss": 0.8511,
      "step": 6105
    },
    {
      "epoch": 2.9316859593354527,
      "grad_norm": 2.6875,
      "learning_rate": 2.5958862911935613e-08,
      "loss": 1.0181,
      "step": 6110
    },
    {
      "epoch": 2.9340850476818807,
      "grad_norm": 2.65625,
      "learning_rate": 2.4164406656453364e-08,
      "loss": 0.7093,
      "step": 6115
    },
    {
      "epoch": 2.936484136028309,
      "grad_norm": 2.96875,
      "learning_rate": 2.2434064077755945e-08,
      "loss": 0.7972,
      "step": 6120
    },
    {
      "epoch": 2.9388832243747376,
      "grad_norm": 3.34375,
      "learning_rate": 2.076785747131993e-08,
      "loss": 0.9193,
      "step": 6125
    },
    {
      "epoch": 2.941282312721166,
      "grad_norm": 3.25,
      "learning_rate": 1.9165808306228496e-08,
      "loss": 0.7502,
      "step": 6130
    },
    {
      "epoch": 2.9436814010675945,
      "grad_norm": 3.0625,
      "learning_rate": 1.7627937224897197e-08,
      "loss": 0.8204,
      "step": 6135
    },
    {
      "epoch": 2.9460804894140225,
      "grad_norm": 3.015625,
      "learning_rate": 1.615426404280529e-08,
      "loss": 0.8157,
      "step": 6140
    },
    {
      "epoch": 2.948479577760451,
      "grad_norm": 3.078125,
      "learning_rate": 1.474480774824205e-08,
      "loss": 0.8643,
      "step": 6145
    },
    {
      "epoch": 2.9508786661068793,
      "grad_norm": 3.109375,
      "learning_rate": 1.3399586502062523e-08,
      "loss": 0.8144,
      "step": 6150
    },
    {
      "epoch": 2.9532777544533078,
      "grad_norm": 2.625,
      "learning_rate": 1.2118617637451035e-08,
      "loss": 0.8754,
      "step": 6155
    },
    {
      "epoch": 2.955676842799736,
      "grad_norm": 4.0,
      "learning_rate": 1.090191765970139e-08,
      "loss": 0.8617,
      "step": 6160
    },
    {
      "epoch": 2.9580759311461646,
      "grad_norm": 3.203125,
      "learning_rate": 9.749502246000352e-09,
      "loss": 0.778,
      "step": 6165
    },
    {
      "epoch": 2.960475019492593,
      "grad_norm": 3.671875,
      "learning_rate": 8.661386245229498e-09,
      "loss": 0.8113,
      "step": 6170
    },
    {
      "epoch": 2.962874107839021,
      "grad_norm": 3.015625,
      "learning_rate": 7.637583677771453e-09,
      "loss": 0.8138,
      "step": 6175
    },
    {
      "epoch": 2.9652731961854495,
      "grad_norm": 2.34375,
      "learning_rate": 6.678107735328398e-09,
      "loss": 0.7571,
      "step": 6180
    },
    {
      "epoch": 2.967672284531878,
      "grad_norm": 2.890625,
      "learning_rate": 5.782970780755515e-09,
      "loss": 0.7821,
      "step": 6185
    },
    {
      "epoch": 2.9700713728783064,
      "grad_norm": 2.90625,
      "learning_rate": 4.952184347898903e-09,
      "loss": 0.8735,
      "step": 6190
    },
    {
      "epoch": 2.9724704612247344,
      "grad_norm": 2.71875,
      "learning_rate": 4.1857591414468106e-09,
      "loss": 0.7953,
      "step": 6195
    },
    {
      "epoch": 2.974869549571163,
      "grad_norm": 2.640625,
      "learning_rate": 3.4837050367936275e-09,
      "loss": 0.7978,
      "step": 6200
    },
    {
      "epoch": 2.974869549571163,
      "eval_loss": 0.946397602558136,
      "eval_runtime": 176.2072,
      "eval_samples_per_second": 42.053,
      "eval_steps_per_second": 10.516,
      "step": 6200
    },
    {
      "epoch": 2.977268637917591,
      "grad_norm": 2.84375,
      "learning_rate": 2.846031079912215e-09,
      "loss": 0.8122,
      "step": 6205
    },
    {
      "epoch": 2.9796677262640197,
      "grad_norm": 3.4375,
      "learning_rate": 2.2727454872351062e-09,
      "loss": 0.8337,
      "step": 6210
    },
    {
      "epoch": 2.982066814610448,
      "grad_norm": 3.296875,
      "learning_rate": 1.7638556455518152e-09,
      "loss": 0.8998,
      "step": 6215
    },
    {
      "epoch": 2.9844659029568765,
      "grad_norm": 3.125,
      "learning_rate": 1.3193681119116897e-09,
      "loss": 0.7873,
      "step": 6220
    },
    {
      "epoch": 2.986864991303305,
      "grad_norm": 3.09375,
      "learning_rate": 9.3928861353898e-10,
      "loss": 0.7989,
      "step": 6225
    },
    {
      "epoch": 2.989264079649733,
      "grad_norm": 3.140625,
      "learning_rate": 6.236220477612298e-10,
      "loss": 0.9234,
      "step": 6230
    },
    {
      "epoch": 2.9916631679961614,
      "grad_norm": 2.75,
      "learning_rate": 3.723724819443275e-10,
      "loss": 0.7815,
      "step": 6235
    },
    {
      "epoch": 2.99406225634259,
      "grad_norm": 2.5625,
      "learning_rate": 1.8554315344088136e-10,
      "loss": 0.6956,
      "step": 6240
    },
    {
      "epoch": 2.9964613446890183,
      "grad_norm": 3.328125,
      "learning_rate": 6.313646954747565e-11,
      "loss": 0.794,
      "step": 6245
    },
    {
      "epoch": 2.9988604330354467,
      "grad_norm": 2.953125,
      "learning_rate": 5.154007475249856e-12,
      "loss": 0.7592,
      "step": 6250
    },
    {
      "epoch": 2.9998200683740177,
      "step": 6252,
      "total_flos": 6.498465993073754e+18,
      "train_loss": 0.937529916860168,
      "train_runtime": 37609.4943,
      "train_samples_per_second": 5.32,
      "train_steps_per_second": 0.166
    }
  ],
  "logging_steps": 5,
  "max_steps": 6252,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.498465993073754e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}