{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 3492,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0057273768613974796,
      "grad_norm": 16.379655838012695,
      "learning_rate": 2.9999392970762296e-06,
      "loss": 1.5958,
      "step": 10
    },
    {
      "epoch": 0.011454753722794959,
      "grad_norm": 13.56667423248291,
      "learning_rate": 2.999757193218044e-06,
      "loss": 1.3928,
      "step": 20
    },
    {
      "epoch": 0.01718213058419244,
      "grad_norm": 10.130936622619629,
      "learning_rate": 2.999453703164426e-06,
      "loss": 1.2605,
      "step": 30
    },
    {
      "epoch": 0.022909507445589918,
      "grad_norm": 10.301068305969238,
      "learning_rate": 2.9990288514790206e-06,
      "loss": 1.3193,
      "step": 40
    },
    {
      "epoch": 0.0286368843069874,
      "grad_norm": 8.71345043182373,
      "learning_rate": 2.9984826725481468e-06,
      "loss": 1.1754,
      "step": 50
    },
    {
      "epoch": 0.03436426116838488,
      "grad_norm": 12.313104629516602,
      "learning_rate": 2.9978152105780155e-06,
      "loss": 1.3135,
      "step": 60
    },
    {
      "epoch": 0.04009163802978236,
      "grad_norm": 12.800464630126953,
      "learning_rate": 2.9970265195911502e-06,
      "loss": 1.2516,
      "step": 70
    },
    {
      "epoch": 0.045819014891179836,
      "grad_norm": 10.99730396270752,
      "learning_rate": 2.9961166634220168e-06,
      "loss": 1.203,
      "step": 80
    },
    {
      "epoch": 0.05154639175257732,
      "grad_norm": 13.54122257232666,
      "learning_rate": 2.9950857157118544e-06,
      "loss": 1.2237,
      "step": 90
    },
    {
      "epoch": 0.0572737686139748,
      "grad_norm": 12.151561737060547,
      "learning_rate": 2.9939337599027165e-06,
      "loss": 1.2715,
      "step": 100
    },
    {
      "epoch": 0.06300114547537228,
      "grad_norm": 19.66217803955078,
      "learning_rate": 2.992660889230718e-06,
      "loss": 1.2601,
      "step": 110
    },
    {
      "epoch": 0.06872852233676977,
      "grad_norm": 9.985322952270508,
      "learning_rate": 2.9912672067184863e-06,
      "loss": 1.1591,
      "step": 120
    },
    {
      "epoch": 0.07445589919816724,
      "grad_norm": 10.843229293823242,
      "learning_rate": 2.9897528251668274e-06,
      "loss": 1.1064,
      "step": 130
    },
    {
      "epoch": 0.08018327605956473,
      "grad_norm": 13.866814613342285,
      "learning_rate": 2.9881178671455905e-06,
      "loss": 1.1918,
      "step": 140
    },
    {
      "epoch": 0.0859106529209622,
      "grad_norm": 13.748754501342773,
      "learning_rate": 2.9863624649837522e-06,
      "loss": 1.2726,
      "step": 150
    },
    {
      "epoch": 0.09163802978235967,
      "grad_norm": 9.494117736816406,
      "learning_rate": 2.9844867607587037e-06,
      "loss": 1.1832,
      "step": 160
    },
    {
      "epoch": 0.09736540664375716,
      "grad_norm": 9.150392532348633,
      "learning_rate": 2.982490906284753e-06,
      "loss": 1.1274,
      "step": 170
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 11.971685409545898,
      "learning_rate": 2.980375063100836e-06,
      "loss": 1.1399,
      "step": 180
    },
    {
      "epoch": 0.10882016036655212,
      "grad_norm": 8.963068962097168,
      "learning_rate": 2.9781394024574426e-06,
      "loss": 1.1726,
      "step": 190
    },
    {
      "epoch": 0.1145475372279496,
      "grad_norm": 9.128679275512695,
      "learning_rate": 2.975784105302756e-06,
      "loss": 1.1972,
      "step": 200
    },
    {
      "epoch": 0.12027491408934708,
      "grad_norm": 12.733792304992676,
      "learning_rate": 2.973309362268008e-06,
      "loss": 1.1732,
      "step": 210
    },
    {
      "epoch": 0.12600229095074456,
      "grad_norm": 10.524323463439941,
      "learning_rate": 2.970715373652049e-06,
      "loss": 1.2613,
      "step": 220
    },
    {
      "epoch": 0.13172966781214204,
      "grad_norm": 7.896759510040283,
      "learning_rate": 2.9680023494051364e-06,
      "loss": 1.1437,
      "step": 230
    },
    {
      "epoch": 0.13745704467353953,
      "grad_norm": 7.099399089813232,
      "learning_rate": 2.9651705091119422e-06,
      "loss": 1.0787,
      "step": 240
    },
    {
      "epoch": 0.143184421534937,
      "grad_norm": 9.507390975952148,
      "learning_rate": 2.962220081973781e-06,
      "loss": 1.0828,
      "step": 250
    },
    {
      "epoch": 0.14891179839633448,
      "grad_norm": 10.189469337463379,
      "learning_rate": 2.959151306790057e-06,
      "loss": 1.1324,
      "step": 260
    },
    {
      "epoch": 0.15463917525773196,
      "grad_norm": 10.902261734008789,
      "learning_rate": 2.955964431938939e-06,
      "loss": 1.2347,
      "step": 270
    },
    {
      "epoch": 0.16036655211912945,
      "grad_norm": 10.873947143554688,
      "learning_rate": 2.952659715357254e-06,
      "loss": 1.1749,
      "step": 280
    },
    {
      "epoch": 0.1660939289805269,
      "grad_norm": 8.8741455078125,
      "learning_rate": 2.9492374245196148e-06,
      "loss": 1.2453,
      "step": 290
    },
    {
      "epoch": 0.1718213058419244,
      "grad_norm": 9.256343841552734,
      "learning_rate": 2.9456978364167667e-06,
      "loss": 1.2202,
      "step": 300
    },
    {
      "epoch": 0.1775486827033219,
      "grad_norm": 13.664665222167969,
      "learning_rate": 2.9420412375331737e-06,
      "loss": 1.1165,
      "step": 310
    },
    {
      "epoch": 0.18327605956471935,
      "grad_norm": 12.664402961730957,
      "learning_rate": 2.938267923823825e-06,
      "loss": 1.1933,
      "step": 320
    },
    {
      "epoch": 0.18900343642611683,
      "grad_norm": 10.518999099731445,
      "learning_rate": 2.934378200690288e-06,
      "loss": 1.1306,
      "step": 330
    },
    {
      "epoch": 0.19473081328751432,
      "grad_norm": 11.800782203674316,
      "learning_rate": 2.9303723829559846e-06,
      "loss": 1.1013,
      "step": 340
    },
    {
      "epoch": 0.2004581901489118,
      "grad_norm": 13.071691513061523,
      "learning_rate": 2.9262507948407135e-06,
      "loss": 1.2304,
      "step": 350
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 9.5667142868042,
      "learning_rate": 2.9220137699344057e-06,
      "loss": 1.1543,
      "step": 360
    },
    {
      "epoch": 0.21191294387170675,
      "grad_norm": 11.27755069732666,
      "learning_rate": 2.917661651170129e-06,
      "loss": 1.1647,
      "step": 370
    },
    {
      "epoch": 0.21764032073310424,
      "grad_norm": 12.141744613647461,
      "learning_rate": 2.913194790796327e-06,
      "loss": 1.1378,
      "step": 380
    },
    {
      "epoch": 0.22336769759450173,
      "grad_norm": 14.818400382995605,
      "learning_rate": 2.9086135503483137e-06,
      "loss": 1.1353,
      "step": 390
    },
    {
      "epoch": 0.2290950744558992,
      "grad_norm": 10.42490005493164,
      "learning_rate": 2.903918300619008e-06,
      "loss": 1.2457,
      "step": 400
    },
    {
      "epoch": 0.23482245131729668,
      "grad_norm": 11.577737808227539,
      "learning_rate": 2.8991094216289254e-06,
      "loss": 1.1824,
      "step": 410
    },
    {
      "epoch": 0.24054982817869416,
      "grad_norm": 9.93144416809082,
      "learning_rate": 2.8941873025954193e-06,
      "loss": 1.2134,
      "step": 420
    },
    {
      "epoch": 0.24627720504009165,
      "grad_norm": 8.588950157165527,
      "learning_rate": 2.8891523419011774e-06,
      "loss": 1.2043,
      "step": 430
    },
    {
      "epoch": 0.2520045819014891,
      "grad_norm": 9.073149681091309,
      "learning_rate": 2.884004947061981e-06,
      "loss": 1.2291,
      "step": 440
    },
    {
      "epoch": 0.25773195876288657,
      "grad_norm": 7.688851833343506,
      "learning_rate": 2.8787455346937186e-06,
      "loss": 1.2401,
      "step": 450
    },
    {
      "epoch": 0.2634593356242841,
      "grad_norm": 10.9544677734375,
      "learning_rate": 2.873374530478667e-06,
      "loss": 1.2467,
      "step": 460
    },
    {
      "epoch": 0.26918671248568155,
      "grad_norm": 9.605713844299316,
      "learning_rate": 2.8678923691310392e-06,
      "loss": 1.1678,
      "step": 470
    },
    {
      "epoch": 0.27491408934707906,
      "grad_norm": 11.065582275390625,
      "learning_rate": 2.8622994943617984e-06,
      "loss": 1.1364,
      "step": 480
    },
    {
      "epoch": 0.2806414662084765,
      "grad_norm": 9.69209098815918,
      "learning_rate": 2.8565963588427457e-06,
      "loss": 1.0762,
      "step": 490
    },
    {
      "epoch": 0.286368843069874,
      "grad_norm": 10.874582290649414,
      "learning_rate": 2.8507834241698817e-06,
      "loss": 1.1281,
      "step": 500
    },
    {
      "epoch": 0.286368843069874,
      "eval_loss": 1.1230082511901855,
      "eval_runtime": 5.3115,
      "eval_samples_per_second": 67.778,
      "eval_steps_per_second": 16.944,
      "step": 500
    },
    {
      "epoch": 0.2920962199312715,
      "grad_norm": 16.209104537963867,
      "learning_rate": 2.844861160826047e-06,
      "loss": 1.158,
      "step": 510
    },
    {
      "epoch": 0.29782359679266895,
      "grad_norm": 11.166873931884766,
      "learning_rate": 2.8388300481428423e-06,
      "loss": 1.0916,
      "step": 520
    },
    {
      "epoch": 0.3035509736540664,
      "grad_norm": 11.725543022155762,
      "learning_rate": 2.8326905742618318e-06,
      "loss": 1.2,
      "step": 530
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 10.414682388305664,
      "learning_rate": 2.8264432360950353e-06,
      "loss": 1.1615,
      "step": 540
    },
    {
      "epoch": 0.3150057273768614,
      "grad_norm": 9.764941215515137,
      "learning_rate": 2.8200885392847106e-06,
      "loss": 1.175,
      "step": 550
    },
    {
      "epoch": 0.3207331042382589,
      "grad_norm": 16.335607528686523,
      "learning_rate": 2.8136269981624244e-06,
      "loss": 1.0646,
      "step": 560
    },
    {
      "epoch": 0.32646048109965636,
      "grad_norm": 8.465163230895996,
      "learning_rate": 2.8070591357074282e-06,
      "loss": 1.2205,
      "step": 570
    },
    {
      "epoch": 0.3321878579610538,
      "grad_norm": 11.596179008483887,
      "learning_rate": 2.800385483504327e-06,
      "loss": 1.2421,
      "step": 580
    },
    {
      "epoch": 0.33791523482245134,
      "grad_norm": 11.226552963256836,
      "learning_rate": 2.7936065817000565e-06,
      "loss": 1.1326,
      "step": 590
    },
    {
      "epoch": 0.3436426116838488,
      "grad_norm": 11.108230590820312,
      "learning_rate": 2.7867229789601615e-06,
      "loss": 1.2161,
      "step": 600
    },
    {
      "epoch": 0.34936998854524626,
      "grad_norm": 12.34344482421875,
      "learning_rate": 2.7797352324243925e-06,
      "loss": 1.149,
      "step": 610
    },
    {
      "epoch": 0.3550973654066438,
      "grad_norm": 7.934173583984375,
      "learning_rate": 2.772643907661609e-06,
      "loss": 1.1019,
      "step": 620
    },
    {
      "epoch": 0.36082474226804123,
      "grad_norm": 7.608425140380859,
      "learning_rate": 2.765449578624007e-06,
      "loss": 1.0918,
      "step": 630
    },
    {
      "epoch": 0.3665521191294387,
      "grad_norm": 11.046185493469238,
      "learning_rate": 2.7581528276006626e-06,
      "loss": 1.0596,
      "step": 640
    },
    {
      "epoch": 0.3722794959908362,
      "grad_norm": 12.330936431884766,
      "learning_rate": 2.7507542451704043e-06,
      "loss": 1.282,
      "step": 650
    },
    {
      "epoch": 0.37800687285223367,
      "grad_norm": 14.420117378234863,
      "learning_rate": 2.743254430154012e-06,
      "loss": 1.215,
      "step": 660
    },
    {
      "epoch": 0.3837342497136312,
      "grad_norm": 10.501232147216797,
      "learning_rate": 2.7356539895657513e-06,
      "loss": 1.1786,
      "step": 670
    },
    {
      "epoch": 0.38946162657502864,
      "grad_norm": 10.232950210571289,
      "learning_rate": 2.727953538564244e-06,
      "loss": 1.1201,
      "step": 680
    },
    {
      "epoch": 0.3951890034364261,
      "grad_norm": 13.221607208251953,
      "learning_rate": 2.7201537004026756e-06,
      "loss": 1.1669,
      "step": 690
    },
    {
      "epoch": 0.4009163802978236,
      "grad_norm": 9.83636474609375,
      "learning_rate": 2.7122551063783555e-06,
      "loss": 1.2779,
      "step": 700
    },
    {
      "epoch": 0.4066437571592211,
      "grad_norm": 10.758344650268555,
      "learning_rate": 2.704258395781618e-06,
      "loss": 1.1545,
      "step": 710
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 15.233297348022461,
      "learning_rate": 2.696164215844081e-06,
      "loss": 1.0843,
      "step": 720
    },
    {
      "epoch": 0.41809851088201605,
      "grad_norm": 9.102065086364746,
      "learning_rate": 2.687973221686263e-06,
      "loss": 1.0776,
      "step": 730
    },
    {
      "epoch": 0.4238258877434135,
      "grad_norm": 9.260092735290527,
      "learning_rate": 2.679686076264554e-06,
      "loss": 1.1087,
      "step": 740
    },
    {
      "epoch": 0.42955326460481097,
      "grad_norm": 19.753076553344727,
      "learning_rate": 2.6713034503175643e-06,
      "loss": 1.1055,
      "step": 750
    },
    {
      "epoch": 0.4352806414662085,
      "grad_norm": 7.540209770202637,
      "learning_rate": 2.662826022311832e-06,
      "loss": 1.1342,
      "step": 760
    },
    {
      "epoch": 0.44100801832760594,
      "grad_norm": 12.86109733581543,
      "learning_rate": 2.654254478386912e-06,
      "loss": 1.186,
      "step": 770
    },
    {
      "epoch": 0.44673539518900346,
      "grad_norm": 9.657477378845215,
      "learning_rate": 2.6455895122998405e-06,
      "loss": 1.1337,
      "step": 780
    },
    {
      "epoch": 0.4524627720504009,
      "grad_norm": 11.697619438171387,
      "learning_rate": 2.636831825368986e-06,
      "loss": 1.1123,
      "step": 790
    },
    {
      "epoch": 0.4581901489117984,
      "grad_norm": 9.250412940979004,
      "learning_rate": 2.627982126417284e-06,
      "loss": 1.0457,
      "step": 800
    },
    {
      "epoch": 0.4639175257731959,
      "grad_norm": 12.023123741149902,
      "learning_rate": 2.619041131714869e-06,
      "loss": 1.0837,
      "step": 810
    },
    {
      "epoch": 0.46964490263459335,
      "grad_norm": 8.030861854553223,
      "learning_rate": 2.6100095649211017e-06,
      "loss": 1.1827,
      "step": 820
    },
    {
      "epoch": 0.4753722794959908,
      "grad_norm": 10.623430252075195,
      "learning_rate": 2.600888157025995e-06,
      "loss": 1.0982,
      "step": 830
    },
    {
      "epoch": 0.48109965635738833,
      "grad_norm": 8.662152290344238,
      "learning_rate": 2.591677646291054e-06,
      "loss": 1.2227,
      "step": 840
    },
    {
      "epoch": 0.4868270332187858,
      "grad_norm": 11.325638771057129,
      "learning_rate": 2.5823787781895197e-06,
      "loss": 0.9943,
      "step": 850
    },
    {
      "epoch": 0.4925544100801833,
      "grad_norm": 9.88836669921875,
      "learning_rate": 2.5729923053460347e-06,
      "loss": 1.2253,
      "step": 860
    },
    {
      "epoch": 0.49828178694158076,
      "grad_norm": 11.135693550109863,
      "learning_rate": 2.5635189874757255e-06,
      "loss": 1.2906,
      "step": 870
    },
    {
      "epoch": 0.5040091638029782,
      "grad_norm": 8.991135597229004,
      "learning_rate": 2.553959591322716e-06,
      "loss": 1.0484,
      "step": 880
    },
    {
      "epoch": 0.5097365406643757,
      "grad_norm": 8.861202239990234,
      "learning_rate": 2.544314890598067e-06,
      "loss": 1.0301,
      "step": 890
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 13.207406997680664,
      "learning_rate": 2.5345856659171565e-06,
      "loss": 1.2542,
      "step": 900
    },
    {
      "epoch": 0.5211912943871707,
      "grad_norm": 9.178120613098145,
      "learning_rate": 2.5247727047364956e-06,
      "loss": 1.085,
      "step": 910
    },
    {
      "epoch": 0.5269186712485682,
      "grad_norm": 10.854491233825684,
      "learning_rate": 2.5148768012899977e-06,
      "loss": 1.2201,
      "step": 920
    },
    {
      "epoch": 0.5326460481099656,
      "grad_norm": 10.438577651977539,
      "learning_rate": 2.5048987565246934e-06,
      "loss": 1.1778,
      "step": 930
    },
    {
      "epoch": 0.5383734249713631,
      "grad_norm": 10.084646224975586,
      "learning_rate": 2.4948393780359025e-06,
      "loss": 1.1588,
      "step": 940
    },
    {
      "epoch": 0.5441008018327605,
      "grad_norm": 10.475250244140625,
      "learning_rate": 2.484699480001873e-06,
      "loss": 1.138,
      "step": 950
    },
    {
      "epoch": 0.5498281786941581,
      "grad_norm": 11.831788063049316,
      "learning_rate": 2.4744798831178817e-06,
      "loss": 1.1661,
      "step": 960
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 11.067156791687012,
      "learning_rate": 2.464181414529809e-06,
      "loss": 1.047,
      "step": 970
    },
    {
      "epoch": 0.561282932416953,
      "grad_norm": 12.043890953063965,
      "learning_rate": 2.4538049077671936e-06,
      "loss": 1.2494,
      "step": 980
    },
    {
      "epoch": 0.5670103092783505,
      "grad_norm": 12.70505142211914,
      "learning_rate": 2.4433512026757667e-06,
      "loss": 1.1073,
      "step": 990
    },
    {
      "epoch": 0.572737686139748,
      "grad_norm": 12.398516654968262,
      "learning_rate": 2.4328211453494805e-06,
      "loss": 1.183,
      "step": 1000
    },
    {
      "epoch": 0.572737686139748,
      "eval_loss": 1.1016738414764404,
      "eval_runtime": 5.3796,
      "eval_samples_per_second": 66.92,
      "eval_steps_per_second": 16.73,
      "step": 1000
    },
    {
      "epoch": 0.5784650630011455,
      "grad_norm": 12.058919906616211,
      "learning_rate": 2.4222155880620232e-06,
      "loss": 1.1006,
      "step": 1010
    },
    {
      "epoch": 0.584192439862543,
      "grad_norm": 9.806292533874512,
      "learning_rate": 2.4115353891978432e-06,
      "loss": 1.1667,
      "step": 1020
    },
    {
      "epoch": 0.5899198167239404,
      "grad_norm": 8.650660514831543,
      "learning_rate": 2.4007814131826705e-06,
      "loss": 1.1308,
      "step": 1030
    },
    {
      "epoch": 0.5956471935853379,
      "grad_norm": 11.571444511413574,
      "learning_rate": 2.3899545304135533e-06,
      "loss": 1.2669,
      "step": 1040
    },
    {
      "epoch": 0.6013745704467354,
      "grad_norm": 10.114933013916016,
      "learning_rate": 2.3790556171884107e-06,
      "loss": 1.1049,
      "step": 1050
    },
    {
      "epoch": 0.6071019473081328,
      "grad_norm": 11.771845817565918,
      "learning_rate": 2.3680855556351075e-06,
      "loss": 1.1428,
      "step": 1060
    },
    {
      "epoch": 0.6128293241695304,
      "grad_norm": 10.653964042663574,
      "learning_rate": 2.357045233640057e-06,
      "loss": 1.1645,
      "step": 1070
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 7.707538604736328,
      "learning_rate": 2.34593554477636e-06,
      "loss": 1.0534,
      "step": 1080
    },
    {
      "epoch": 0.6242840778923253,
      "grad_norm": 13.67249870300293,
      "learning_rate": 2.3347573882314764e-06,
      "loss": 1.1389,
      "step": 1090
    },
    {
      "epoch": 0.6300114547537228,
      "grad_norm": 11.033611297607422,
      "learning_rate": 2.323511668734453e-06,
      "loss": 1.2522,
      "step": 1100
    },
    {
      "epoch": 0.6357388316151202,
      "grad_norm": 10.89904499053955,
      "learning_rate": 2.312199296482695e-06,
      "loss": 1.2145,
      "step": 1110
    },
    {
      "epoch": 0.6414662084765178,
      "grad_norm": 15.723979949951172,
      "learning_rate": 2.3008211870682958e-06,
      "loss": 1.1635,
      "step": 1120
    },
    {
      "epoch": 0.6471935853379153,
      "grad_norm": 9.854010581970215,
      "learning_rate": 2.2893782614039325e-06,
      "loss": 1.1548,
      "step": 1130
    },
    {
      "epoch": 0.6529209621993127,
      "grad_norm": 7.649772644042969,
      "learning_rate": 2.277871445648332e-06,
      "loss": 1.1,
      "step": 1140
    },
    {
      "epoch": 0.6586483390607102,
      "grad_norm": 10.99400806427002,
      "learning_rate": 2.266301671131307e-06,
      "loss": 1.0816,
      "step": 1150
    },
    {
      "epoch": 0.6643757159221076,
      "grad_norm": 10.330283164978027,
      "learning_rate": 2.2546698742783778e-06,
      "loss": 1.0854,
      "step": 1160
    },
    {
      "epoch": 0.6701030927835051,
      "grad_norm": 10.694668769836426,
      "learning_rate": 2.2429769965349817e-06,
      "loss": 1.1252,
      "step": 1170
    },
    {
      "epoch": 0.6758304696449027,
      "grad_norm": 8.67434310913086,
      "learning_rate": 2.231223984290273e-06,
      "loss": 1.2776,
      "step": 1180
    },
    {
      "epoch": 0.6815578465063001,
      "grad_norm": 16.908462524414062,
      "learning_rate": 2.219411788800528e-06,
      "loss": 1.1514,
      "step": 1190
    },
    {
      "epoch": 0.6872852233676976,
      "grad_norm": 9.732044219970703,
      "learning_rate": 2.2075413661121492e-06,
      "loss": 1.1616,
      "step": 1200
    },
    {
      "epoch": 0.693012600229095,
      "grad_norm": 12.07868766784668,
      "learning_rate": 2.1956136769842885e-06,
      "loss": 1.0862,
      "step": 1210
    },
    {
      "epoch": 0.6987399770904925,
      "grad_norm": 15.745189666748047,
      "learning_rate": 2.1836296868110836e-06,
      "loss": 1.172,
      "step": 1220
    },
    {
      "epoch": 0.7044673539518901,
      "grad_norm": 7.780303478240967,
      "learning_rate": 2.1715903655435243e-06,
      "loss": 1.1839,
      "step": 1230
    },
    {
      "epoch": 0.7101947308132875,
      "grad_norm": 10.862656593322754,
      "learning_rate": 2.1594966876109456e-06,
      "loss": 1.2164,
      "step": 1240
    },
    {
      "epoch": 0.715922107674685,
      "grad_norm": 11.16341495513916,
      "learning_rate": 2.1473496318421603e-06,
      "loss": 1.1294,
      "step": 1250
    },
    {
      "epoch": 0.7216494845360825,
      "grad_norm": 6.848244667053223,
      "learning_rate": 2.135150181386236e-06,
      "loss": 1.1519,
      "step": 1260
    },
    {
      "epoch": 0.7273768613974799,
      "grad_norm": 11.13603401184082,
      "learning_rate": 2.122899323632919e-06,
      "loss": 1.1782,
      "step": 1270
    },
    {
      "epoch": 0.7331042382588774,
      "grad_norm": 7.1757402420043945,
      "learning_rate": 2.1105980501327234e-06,
      "loss": 1.0794,
      "step": 1280
    },
    {
      "epoch": 0.738831615120275,
      "grad_norm": 8.15458869934082,
      "learning_rate": 2.0982473565166727e-06,
      "loss": 1.0031,
      "step": 1290
    },
    {
      "epoch": 0.7445589919816724,
      "grad_norm": 11.527698516845703,
      "learning_rate": 2.0858482424157164e-06,
      "loss": 1.1078,
      "step": 1300
    },
    {
      "epoch": 0.7502863688430699,
      "grad_norm": 7.272953510284424,
      "learning_rate": 2.073401711379827e-06,
      "loss": 1.2212,
      "step": 1310
    },
    {
      "epoch": 0.7560137457044673,
      "grad_norm": 8.701218605041504,
      "learning_rate": 2.060908770796769e-06,
      "loss": 1.1787,
      "step": 1320
    },
    {
      "epoch": 0.7617411225658648,
      "grad_norm": 8.603687286376953,
      "learning_rate": 2.0483704318105704e-06,
      "loss": 1.0969,
      "step": 1330
    },
    {
      "epoch": 0.7674684994272624,
      "grad_norm": 8.430140495300293,
      "learning_rate": 2.0357877092396775e-06,
      "loss": 1.1568,
      "step": 1340
    },
    {
      "epoch": 0.7731958762886598,
      "grad_norm": 10.301761627197266,
      "learning_rate": 2.023161621494823e-06,
      "loss": 1.1787,
      "step": 1350
    },
    {
      "epoch": 0.7789232531500573,
      "grad_norm": 10.218490600585938,
      "learning_rate": 2.0104931904965965e-06,
      "loss": 1.0406,
      "step": 1360
    },
    {
      "epoch": 0.7846506300114547,
      "grad_norm": 11.949338912963867,
      "learning_rate": 1.9977834415927326e-06,
      "loss": 1.1459,
      "step": 1370
    },
    {
      "epoch": 0.7903780068728522,
      "grad_norm": 9.224528312683105,
      "learning_rate": 1.985033403475123e-06,
      "loss": 1.1283,
      "step": 1380
    },
    {
      "epoch": 0.7961053837342497,
      "grad_norm": 9.380345344543457,
      "learning_rate": 1.9722441080965565e-06,
      "loss": 1.0648,
      "step": 1390
    },
    {
      "epoch": 0.8018327605956472,
      "grad_norm": 12.056816101074219,
      "learning_rate": 1.9594165905871968e-06,
      "loss": 1.1021,
      "step": 1400
    },
    {
      "epoch": 0.8075601374570447,
      "grad_norm": 11.585986137390137,
      "learning_rate": 1.946551889170801e-06,
      "loss": 1.0863,
      "step": 1410
    },
    {
      "epoch": 0.8132875143184422,
      "grad_norm": 10.934039115905762,
      "learning_rate": 1.9336510450806875e-06,
      "loss": 1.1676,
      "step": 1420
    },
    {
      "epoch": 0.8190148911798396,
      "grad_norm": 8.694523811340332,
      "learning_rate": 1.920715102475464e-06,
      "loss": 1.033,
      "step": 1430
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 9.625450134277344,
      "learning_rate": 1.9077451083545143e-06,
      "loss": 1.1275,
      "step": 1440
    },
    {
      "epoch": 0.8304696449026346,
      "grad_norm": 7.567831516265869,
      "learning_rate": 1.8947421124732576e-06,
      "loss": 1.2054,
      "step": 1450
    },
    {
      "epoch": 0.8361970217640321,
      "grad_norm": 10.554939270019531,
      "learning_rate": 1.8817071672581839e-06,
      "loss": 1.1276,
      "step": 1460
    },
    {
      "epoch": 0.8419243986254296,
      "grad_norm": 10.357826232910156,
      "learning_rate": 1.868641327721675e-06,
      "loss": 1.2319,
      "step": 1470
    },
    {
      "epoch": 0.847651775486827,
      "grad_norm": 16.142457962036133,
      "learning_rate": 1.8555456513766118e-06,
      "loss": 1.1066,
      "step": 1480
    },
    {
      "epoch": 0.8533791523482245,
      "grad_norm": 9.302567481994629,
      "learning_rate": 1.8424211981507858e-06,
      "loss": 0.9964,
      "step": 1490
    },
    {
      "epoch": 0.8591065292096219,
      "grad_norm": 13.468149185180664,
      "learning_rate": 1.8292690303011076e-06,
      "loss": 1.1009,
      "step": 1500
    },
    {
      "epoch": 0.8591065292096219,
      "eval_loss": 1.086885690689087,
      "eval_runtime": 5.2843,
      "eval_samples_per_second": 68.127,
      "eval_steps_per_second": 17.032,
      "step": 1500
    },
    {
      "epoch": 0.8648339060710195,
      "grad_norm": 9.433229446411133,
      "learning_rate": 1.816090212327634e-06,
      "loss": 1.0945,
      "step": 1510
    },
    {
      "epoch": 0.870561282932417,
      "grad_norm": 12.731361389160156,
      "learning_rate": 1.8028858108874094e-06,
      "loss": 1.0626,
      "step": 1520
    },
    {
      "epoch": 0.8762886597938144,
      "grad_norm": 10.284618377685547,
      "learning_rate": 1.7896568947081322e-06,
      "loss": 1.1405,
      "step": 1530
    },
    {
      "epoch": 0.8820160366552119,
      "grad_norm": 11.09489631652832,
      "learning_rate": 1.776404534501656e-06,
      "loss": 1.1558,
      "step": 1540
    },
    {
      "epoch": 0.8877434135166093,
      "grad_norm": 11.1835298538208,
      "learning_rate": 1.7631298028773296e-06,
      "loss": 1.2162,
      "step": 1550
    },
    {
      "epoch": 0.8934707903780069,
      "grad_norm": 8.714800834655762,
      "learning_rate": 1.7498337742551817e-06,
      "loss": 1.0309,
      "step": 1560
    },
    {
      "epoch": 0.8991981672394044,
      "grad_norm": 13.366388320922852,
      "learning_rate": 1.7365175247789623e-06,
      "loss": 1.0123,
      "step": 1570
    },
    {
      "epoch": 0.9049255441008018,
      "grad_norm": 8.921965599060059,
      "learning_rate": 1.7231821322290399e-06,
      "loss": 1.1375,
      "step": 1580
    },
    {
      "epoch": 0.9106529209621993,
      "grad_norm": 11.560654640197754,
      "learning_rate": 1.7098286759351717e-06,
      "loss": 1.1892,
      "step": 1590
    },
    {
      "epoch": 0.9163802978235968,
      "grad_norm": 7.3394670486450195,
      "learning_rate": 1.6964582366891427e-06,
      "loss": 1.2526,
      "step": 1600
    },
    {
      "epoch": 0.9221076746849943,
      "grad_norm": 8.516718864440918,
      "learning_rate": 1.6830718966572928e-06,
      "loss": 1.0736,
      "step": 1610
    },
    {
      "epoch": 0.9278350515463918,
      "grad_norm": 10.884456634521484,
      "learning_rate": 1.6696707392929268e-06,
      "loss": 1.1471,
      "step": 1620
    },
    {
      "epoch": 0.9335624284077892,
      "grad_norm": 11.866485595703125,
      "learning_rate": 1.6562558492486226e-06,
      "loss": 1.0814,
      "step": 1630
    },
    {
      "epoch": 0.9392898052691867,
      "grad_norm": 10.002748489379883,
      "learning_rate": 1.6428283122884437e-06,
      "loss": 1.0393,
      "step": 1640
    },
    {
      "epoch": 0.9450171821305842,
      "grad_norm": 10.888050079345703,
      "learning_rate": 1.6293892152000613e-06,
      "loss": 1.1509,
      "step": 1650
    },
    {
      "epoch": 0.9507445589919816,
      "grad_norm": 10.640623092651367,
      "learning_rate": 1.6159396457067895e-06,
      "loss": 1.2364,
      "step": 1660
    },
    {
      "epoch": 0.9564719358533792,
      "grad_norm": 10.190733909606934,
      "learning_rate": 1.60248069237955e-06,
      "loss": 1.203,
      "step": 1670
    },
    {
      "epoch": 0.9621993127147767,
      "grad_norm": 9.43203353881836,
      "learning_rate": 1.5890134445487679e-06,
      "loss": 1.1801,
      "step": 1680
    },
    {
      "epoch": 0.9679266895761741,
      "grad_norm": 11.540404319763184,
      "learning_rate": 1.5755389922162003e-06,
      "loss": 1.097,
      "step": 1690
    },
    {
      "epoch": 0.9736540664375716,
      "grad_norm": 10.006051063537598,
      "learning_rate": 1.5620584259667175e-06,
      "loss": 1.1071,
      "step": 1700
    },
    {
      "epoch": 0.979381443298969,
      "grad_norm": 11.146818161010742,
      "learning_rate": 1.548572836880034e-06,
      "loss": 1.297,
      "step": 1710
    },
    {
      "epoch": 0.9851088201603666,
      "grad_norm": 15.32827091217041,
      "learning_rate": 1.5350833164423974e-06,
      "loss": 1.1533,
      "step": 1720
    },
    {
      "epoch": 0.9908361970217641,
      "grad_norm": 13.463982582092285,
      "learning_rate": 1.5215909564582499e-06,
      "loss": 1.1084,
      "step": 1730
    },
    {
      "epoch": 0.9965635738831615,
      "grad_norm": 10.163274765014648,
      "learning_rate": 1.5080968489618567e-06,
      "loss": 1.0932,
      "step": 1740
    },
    {
      "epoch": 1.002290950744559,
      "grad_norm": 8.775249481201172,
      "learning_rate": 1.494602086128923e-06,
      "loss": 0.9301,
      "step": 1750
    },
    {
      "epoch": 1.0080183276059564,
      "grad_norm": 7.848943710327148,
      "learning_rate": 1.4811077601881953e-06,
      "loss": 0.6965,
      "step": 1760
    },
    {
      "epoch": 1.013745704467354,
      "grad_norm": 13.538870811462402,
      "learning_rate": 1.4676149633330583e-06,
      "loss": 0.6828,
      "step": 1770
    },
    {
      "epoch": 1.0194730813287514,
      "grad_norm": 10.99616527557373,
      "learning_rate": 1.4541247876331374e-06,
      "loss": 0.6853,
      "step": 1780
    },
    {
      "epoch": 1.0252004581901488,
      "grad_norm": 8.934946060180664,
      "learning_rate": 1.440638324945909e-06,
      "loss": 0.759,
      "step": 1790
    },
    {
      "epoch": 1.0309278350515463,
      "grad_norm": 8.413115501403809,
      "learning_rate": 1.4271566668283281e-06,
      "loss": 0.8167,
      "step": 1800
    },
    {
      "epoch": 1.036655211912944,
      "grad_norm": 12.693419456481934,
      "learning_rate": 1.413680904448481e-06,
      "loss": 0.7765,
      "step": 1810
    },
    {
      "epoch": 1.0423825887743414,
      "grad_norm": 18.639657974243164,
      "learning_rate": 1.4002121284972705e-06,
      "loss": 0.6784,
      "step": 1820
    },
    {
      "epoch": 1.0481099656357389,
      "grad_norm": 9.834324836730957,
      "learning_rate": 1.3867514291001363e-06,
      "loss": 0.6933,
      "step": 1830
    },
    {
      "epoch": 1.0538373424971363,
      "grad_norm": 10.812657356262207,
      "learning_rate": 1.373299895728824e-06,
      "loss": 0.7267,
      "step": 1840
    },
    {
      "epoch": 1.0595647193585338,
      "grad_norm": 10.644402503967285,
      "learning_rate": 1.3598586171132068e-06,
      "loss": 0.7609,
      "step": 1850
    },
    {
      "epoch": 1.0652920962199313,
      "grad_norm": 10.04351806640625,
      "learning_rate": 1.3464286811531663e-06,
      "loss": 0.6473,
      "step": 1860
    },
    {
      "epoch": 1.0710194730813287,
      "grad_norm": 8.527778625488281,
      "learning_rate": 1.3330111748305409e-06,
      "loss": 0.6419,
      "step": 1870
    },
    {
      "epoch": 1.0767468499427262,
      "grad_norm": 10.766478538513184,
      "learning_rate": 1.3196071841211486e-06,
      "loss": 0.707,
      "step": 1880
    },
    {
      "epoch": 1.0824742268041236,
      "grad_norm": 9.123445510864258,
      "learning_rate": 1.3062177939068908e-06,
      "loss": 0.7146,
      "step": 1890
    },
    {
      "epoch": 1.088201603665521,
      "grad_norm": 13.148359298706055,
      "learning_rate": 1.292844087887946e-06,
      "loss": 0.7023,
      "step": 1900
    },
    {
      "epoch": 1.0939289805269188,
      "grad_norm": 9.834714889526367,
      "learning_rate": 1.279487148495056e-06,
      "loss": 0.6432,
      "step": 1910
    },
    {
      "epoch": 1.0996563573883162,
      "grad_norm": 8.148127555847168,
      "learning_rate": 1.2661480568019203e-06,
      "loss": 0.6387,
      "step": 1920
    },
    {
      "epoch": 1.1053837342497137,
      "grad_norm": 14.112127304077148,
      "learning_rate": 1.2528278924376933e-06,
      "loss": 0.7409,
      "step": 1930
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 12.995244979858398,
      "learning_rate": 1.2395277334996047e-06,
      "loss": 0.7186,
      "step": 1940
    },
    {
      "epoch": 1.1168384879725086,
      "grad_norm": 11.058789253234863,
      "learning_rate": 1.2262486564656995e-06,
      "loss": 0.7758,
      "step": 1950
    },
    {
      "epoch": 1.122565864833906,
      "grad_norm": 11.76444149017334,
      "learning_rate": 1.2129917361077125e-06,
      "loss": 0.6512,
      "step": 1960
    },
    {
      "epoch": 1.1282932416953035,
      "grad_norm": 12.445648193359375,
      "learning_rate": 1.1997580454040784e-06,
      "loss": 0.6401,
      "step": 1970
    },
    {
      "epoch": 1.134020618556701,
      "grad_norm": 12.5171537399292,
      "learning_rate": 1.1865486554530874e-06,
      "loss": 0.7075,
      "step": 1980
    },
    {
      "epoch": 1.1397479954180985,
      "grad_norm": 11.082114219665527,
      "learning_rate": 1.1733646353861951e-06,
      "loss": 0.7184,
      "step": 1990
    },
    {
      "epoch": 1.145475372279496,
      "grad_norm": 8.820939064025879,
      "learning_rate": 1.1602070522814882e-06,
      "loss": 0.6901,
      "step": 2000
    },
    {
      "epoch": 1.145475372279496,
      "eval_loss": 1.1485834121704102,
      "eval_runtime": 5.46,
      "eval_samples_per_second": 65.934,
      "eval_steps_per_second": 16.484,
      "step": 2000
    },
    {
      "epoch": 1.1512027491408934,
      "grad_norm": 12.333765983581543,
      "learning_rate": 1.1470769710773189e-06,
      "loss": 0.7145,
      "step": 2010
    },
    {
      "epoch": 1.1569301260022908,
      "grad_norm": 9.632255554199219,
      "learning_rate": 1.1339754544861111e-06,
      "loss": 0.6381,
      "step": 2020
    },
    {
      "epoch": 1.1626575028636885,
      "grad_norm": 10.103231430053711,
      "learning_rate": 1.1209035629083498e-06,
      "loss": 0.6353,
      "step": 2030
    },
    {
      "epoch": 1.168384879725086,
      "grad_norm": 11.000112533569336,
      "learning_rate": 1.1078623543467518e-06,
      "loss": 0.676,
      "step": 2040
    },
    {
      "epoch": 1.1741122565864834,
      "grad_norm": 11.119033813476562,
      "learning_rate": 1.0948528843206357e-06,
      "loss": 0.6746,
      "step": 2050
    },
    {
      "epoch": 1.179839633447881,
      "grad_norm": 11.437407493591309,
      "learning_rate": 1.081876205780492e-06,
      "loss": 0.6734,
      "step": 2060
    },
    {
      "epoch": 1.1855670103092784,
      "grad_norm": 8.372086524963379,
      "learning_rate": 1.068933369022758e-06,
      "loss": 0.7758,
      "step": 2070
    },
    {
      "epoch": 1.1912943871706758,
      "grad_norm": 7.501805305480957,
      "learning_rate": 1.0560254216048105e-06,
      "loss": 0.8655,
      "step": 2080
    },
    {
      "epoch": 1.1970217640320733,
      "grad_norm": 11.55257797241211,
      "learning_rate": 1.0431534082601814e-06,
      "loss": 0.7701,
      "step": 2090
    },
    {
      "epoch": 1.2027491408934707,
      "grad_norm": 13.704492568969727,
      "learning_rate": 1.0303183708139966e-06,
      "loss": 0.7325,
      "step": 2100
    },
    {
      "epoch": 1.2084765177548682,
      "grad_norm": 9.895234107971191,
      "learning_rate": 1.0175213480986553e-06,
      "loss": 0.7678,
      "step": 2110
    },
    {
      "epoch": 1.2142038946162657,
      "grad_norm": 11.093770027160645,
      "learning_rate": 1.0047633758697507e-06,
      "loss": 0.7074,
      "step": 2120
    },
    {
      "epoch": 1.2199312714776633,
      "grad_norm": 7.835501194000244,
      "learning_rate": 9.920454867222362e-07,
      "loss": 0.6939,
      "step": 2130
    },
    {
      "epoch": 1.2256586483390608,
      "grad_norm": 12.765395164489746,
      "learning_rate": 9.793687100068537e-07,
      "loss": 0.6601,
      "step": 2140
    },
    {
      "epoch": 1.2313860252004583,
      "grad_norm": 8.465944290161133,
      "learning_rate": 9.667340717468165e-07,
      "loss": 0.6601,
      "step": 2150
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 8.568673133850098,
      "learning_rate": 9.54142594554769e-07,
      "loss": 0.6536,
      "step": 2160
    },
    {
      "epoch": 1.2428407789232532,
      "grad_norm": 11.03416633605957,
      "learning_rate": 9.415952975500177e-07,
      "loss": 0.7267,
      "step": 2170
    },
    {
      "epoch": 1.2485681557846506,
      "grad_norm": 12.343921661376953,
      "learning_rate": 9.290931962760477e-07,
      "loss": 0.7921,
      "step": 2180
    },
    {
      "epoch": 1.254295532646048,
      "grad_norm": 12.319411277770996,
      "learning_rate": 9.166373026183269e-07,
      "loss": 0.6843,
      "step": 2190
    },
    {
      "epoch": 1.2600229095074456,
      "grad_norm": 11.100428581237793,
      "learning_rate": 9.042286247224054e-07,
      "loss": 0.7535,
      "step": 2200
    },
    {
      "epoch": 1.265750286368843,
      "grad_norm": 10.030117988586426,
      "learning_rate": 8.918681669123218e-07,
      "loss": 0.6571,
      "step": 2210
    },
    {
      "epoch": 1.2714776632302405,
      "grad_norm": 10.269767761230469,
      "learning_rate": 8.795569296093133e-07,
      "loss": 0.7255,
      "step": 2220
    },
    {
      "epoch": 1.277205040091638,
      "grad_norm": 11.503005027770996,
      "learning_rate": 8.67295909250846e-07,
      "loss": 0.7021,
      "step": 2230
    },
    {
      "epoch": 1.2829324169530354,
      "grad_norm": 19.0956974029541,
      "learning_rate": 8.55086098209965e-07,
      "loss": 0.644,
      "step": 2240
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 11.60425090789795,
      "learning_rate": 8.429284847149765e-07,
      "loss": 0.7289,
      "step": 2250
    },
    {
      "epoch": 1.2943871706758305,
      "grad_norm": 11.15966510772705,
      "learning_rate": 8.308240527694602e-07,
      "loss": 0.6789,
      "step": 2260
    },
    {
      "epoch": 1.300114547537228,
      "grad_norm": 11.775574684143066,
      "learning_rate": 8.187737820726293e-07,
      "loss": 0.7226,
      "step": 2270
    },
    {
      "epoch": 1.3058419243986255,
      "grad_norm": 10.303918838500977,
      "learning_rate": 8.067786479400346e-07,
      "loss": 0.6405,
      "step": 2280
    },
    {
      "epoch": 1.311569301260023,
      "grad_norm": 11.979586601257324,
      "learning_rate": 7.948396212246273e-07,
      "loss": 0.6847,
      "step": 2290
    },
    {
      "epoch": 1.3172966781214204,
      "grad_norm": 8.767427444458008,
      "learning_rate": 7.829576682381782e-07,
      "loss": 0.681,
      "step": 2300
    },
    {
      "epoch": 1.3230240549828178,
      "grad_norm": 12.378206253051758,
      "learning_rate": 7.711337506730695e-07,
      "loss": 0.7108,
      "step": 2310
    },
    {
      "epoch": 1.3287514318442153,
      "grad_norm": 9.478968620300293,
      "learning_rate": 7.593688255244566e-07,
      "loss": 0.6728,
      "step": 2320
    },
    {
      "epoch": 1.3344788087056128,
      "grad_norm": 13.136493682861328,
      "learning_rate": 7.476638450128123e-07,
      "loss": 0.6783,
      "step": 2330
    },
    {
      "epoch": 1.3402061855670104,
      "grad_norm": 12.096550941467285,
      "learning_rate": 7.360197565068561e-07,
      "loss": 0.6398,
      "step": 2340
    },
    {
      "epoch": 1.345933562428408,
      "grad_norm": 9.796936988830566,
      "learning_rate": 7.244375024468765e-07,
      "loss": 0.659,
      "step": 2350
    },
    {
      "epoch": 1.3516609392898054,
      "grad_norm": 12.897764205932617,
      "learning_rate": 7.129180202684554e-07,
      "loss": 0.6226,
      "step": 2360
    },
    {
      "epoch": 1.3573883161512028,
      "grad_norm": 13.480414390563965,
      "learning_rate": 7.0146224232659e-07,
      "loss": 0.7311,
      "step": 2370
    },
    {
      "epoch": 1.3631156930126003,
      "grad_norm": 9.178197860717773,
      "learning_rate": 6.900710958202341e-07,
      "loss": 0.7167,
      "step": 2380
    },
    {
      "epoch": 1.3688430698739977,
      "grad_norm": 7.676974773406982,
      "learning_rate": 6.787455027172515e-07,
      "loss": 0.7085,
      "step": 2390
    },
    {
      "epoch": 1.3745704467353952,
      "grad_norm": 19.0358829498291,
      "learning_rate": 6.674863796797954e-07,
      "loss": 0.6221,
      "step": 2400
    },
    {
      "epoch": 1.3802978235967927,
      "grad_norm": 10.022867202758789,
      "learning_rate": 6.562946379901158e-07,
      "loss": 0.7337,
      "step": 2410
    },
    {
      "epoch": 1.38602520045819,
      "grad_norm": 11.5038480758667,
      "learning_rate": 6.451711834768026e-07,
      "loss": 0.6849,
      "step": 2420
    },
    {
      "epoch": 1.3917525773195876,
      "grad_norm": 10.91701889038086,
      "learning_rate": 6.341169164414712e-07,
      "loss": 0.6061,
      "step": 2430
    },
    {
      "epoch": 1.397479954180985,
      "grad_norm": 11.305731773376465,
      "learning_rate": 6.23132731585894e-07,
      "loss": 0.6072,
      "step": 2440
    },
    {
      "epoch": 1.4032073310423825,
      "grad_norm": 11.2603178024292,
      "learning_rate": 6.122195179395854e-07,
      "loss": 0.7589,
      "step": 2450
    },
    {
      "epoch": 1.40893470790378,
      "grad_norm": 10.891215324401855,
      "learning_rate": 6.013781587878464e-07,
      "loss": 0.6253,
      "step": 2460
    },
    {
      "epoch": 1.4146620847651776,
      "grad_norm": 13.905660629272461,
      "learning_rate": 5.906095316002755e-07,
      "loss": 0.7618,
      "step": 2470
    },
    {
      "epoch": 1.420389461626575,
      "grad_norm": 8.74496078491211,
      "learning_rate": 5.799145079597458e-07,
      "loss": 0.6866,
      "step": 2480
    },
    {
      "epoch": 1.4261168384879725,
      "grad_norm": 10.08084487915039,
      "learning_rate": 5.69293953491864e-07,
      "loss": 0.6584,
      "step": 2490
    },
    {
      "epoch": 1.43184421534937,
      "grad_norm": 9.269981384277344,
      "learning_rate": 5.587487277949075e-07,
      "loss": 0.7377,
      "step": 2500
    },
    {
      "epoch": 1.43184421534937,
      "eval_loss": 1.1487523317337036,
      "eval_runtime": 5.3152,
      "eval_samples_per_second": 67.73,
      "eval_steps_per_second": 16.933,
      "step": 2500
    },
    {
      "epoch": 1.4375715922107675,
      "grad_norm": 8.877992630004883,
      "learning_rate": 5.48279684370252e-07,
      "loss": 0.7111,
      "step": 2510
    },
    {
      "epoch": 1.443298969072165,
      "grad_norm": 24.23348045349121,
      "learning_rate": 5.378876705532904e-07,
      "loss": 0.62,
      "step": 2520
    },
    {
      "epoch": 1.4490263459335624,
      "grad_norm": 9.67536735534668,
      "learning_rate": 5.275735274448529e-07,
      "loss": 0.7702,
      "step": 2530
    },
    {
      "epoch": 1.4547537227949598,
      "grad_norm": 10.30777359008789,
      "learning_rate": 5.173380898431305e-07,
      "loss": 0.6682,
      "step": 2540
    },
    {
      "epoch": 1.4604810996563573,
      "grad_norm": 12.1168212890625,
      "learning_rate": 5.071821861761074e-07,
      "loss": 0.6713,
      "step": 2550
    },
    {
      "epoch": 1.466208476517755,
      "grad_norm": 7.113183975219727,
      "learning_rate": 4.97106638434512e-07,
      "loss": 0.7582,
      "step": 2560
    },
    {
      "epoch": 1.4719358533791524,
      "grad_norm": 13.592216491699219,
      "learning_rate": 4.87112262105286e-07,
      "loss": 0.6917,
      "step": 2570
    },
    {
      "epoch": 1.47766323024055,
      "grad_norm": 10.175348281860352,
      "learning_rate": 4.771998661055823e-07,
      "loss": 0.7341,
      "step": 2580
    },
    {
      "epoch": 1.4833906071019474,
      "grad_norm": 10.051101684570312,
      "learning_rate": 4.673702527172919e-07,
      "loss": 0.5795,
      "step": 2590
    },
    {
      "epoch": 1.4891179839633448,
      "grad_norm": 9.961878776550293,
      "learning_rate": 4.576242175221121e-07,
      "loss": 0.6895,
      "step": 2600
    },
    {
      "epoch": 1.4948453608247423,
      "grad_norm": 10.214471817016602,
      "learning_rate": 4.479625493371512e-07,
      "loss": 0.6552,
      "step": 2610
    },
    {
      "epoch": 1.5005727376861397,
      "grad_norm": 16.537036895751953,
      "learning_rate": 4.3838603015108537e-07,
      "loss": 0.6821,
      "step": 2620
    },
    {
      "epoch": 1.5063001145475372,
      "grad_norm": 8.615083694458008,
      "learning_rate": 4.288954350608672e-07,
      "loss": 0.6581,
      "step": 2630
    },
    {
      "epoch": 1.5120274914089347,
      "grad_norm": 12.914610862731934,
      "learning_rate": 4.1949153220898987e-07,
      "loss": 0.7688,
      "step": 2640
    },
    {
      "epoch": 1.5177548682703321,
      "grad_norm": 10.495301246643066,
      "learning_rate": 4.1017508272131794e-07,
      "loss": 0.6831,
      "step": 2650
    },
    {
      "epoch": 1.5234822451317296,
      "grad_norm": 11.30919075012207,
      "learning_rate": 4.009468406454818e-07,
      "loss": 0.796,
      "step": 2660
    },
    {
      "epoch": 1.529209621993127,
      "grad_norm": 9.941025733947754,
      "learning_rate": 3.9180755288984837e-07,
      "loss": 0.7486,
      "step": 2670
    },
    {
      "epoch": 1.5349369988545245,
      "grad_norm": 11.294404983520508,
      "learning_rate": 3.8275795916306827e-07,
      "loss": 0.6736,
      "step": 2680
    },
    {
      "epoch": 1.540664375715922,
      "grad_norm": 12.294686317443848,
      "learning_rate": 3.737987919142061e-07,
      "loss": 0.6431,
      "step": 2690
    },
    {
      "epoch": 1.5463917525773194,
      "grad_norm": 11.265865325927734,
      "learning_rate": 3.64930776273457e-07,
      "loss": 0.8467,
      "step": 2700
    },
    {
      "epoch": 1.552119129438717,
      "grad_norm": 7.701663970947266,
      "learning_rate": 3.561546299934577e-07,
      "loss": 0.7307,
      "step": 2710
    },
    {
      "epoch": 1.5578465063001146,
      "grad_norm": 14.973838806152344,
      "learning_rate": 3.4747106339119276e-07,
      "loss": 0.7451,
      "step": 2720
    },
    {
      "epoch": 1.563573883161512,
      "grad_norm": 23.487585067749023,
      "learning_rate": 3.388807792905045e-07,
      "loss": 0.6253,
      "step": 2730
    },
    {
      "epoch": 1.5693012600229095,
      "grad_norm": 10.245492935180664,
      "learning_rate": 3.303844729652073e-07,
      "loss": 0.7833,
      "step": 2740
    },
    {
      "epoch": 1.575028636884307,
      "grad_norm": 10.44719123840332,
      "learning_rate": 3.219828320828148e-07,
      "loss": 0.6858,
      "step": 2750
    },
    {
      "epoch": 1.5807560137457046,
      "grad_norm": 8.665092468261719,
      "learning_rate": 3.1367653664888173e-07,
      "loss": 0.6299,
      "step": 2760
    },
    {
      "epoch": 1.586483390607102,
      "grad_norm": 11.613511085510254,
      "learning_rate": 3.054662589519661e-07,
      "loss": 0.8032,
      "step": 2770
    },
    {
      "epoch": 1.5922107674684995,
      "grad_norm": 7.786969184875488,
      "learning_rate": 2.973526635092163e-07,
      "loss": 0.7187,
      "step": 2780
    },
    {
      "epoch": 1.597938144329897,
      "grad_norm": 13.719048500061035,
      "learning_rate": 2.8933640701258597e-07,
      "loss": 0.7349,
      "step": 2790
    },
    {
      "epoch": 1.6036655211912945,
      "grad_norm": 11.509296417236328,
      "learning_rate": 2.8141813827568525e-07,
      "loss": 0.6635,
      "step": 2800
    },
    {
      "epoch": 1.609392898052692,
      "grad_norm": 10.817230224609375,
      "learning_rate": 2.7359849818126533e-07,
      "loss": 0.7161,
      "step": 2810
    },
    {
      "epoch": 1.6151202749140894,
      "grad_norm": 13.700976371765137,
      "learning_rate": 2.6587811962934823e-07,
      "loss": 0.7205,
      "step": 2820
    },
    {
      "epoch": 1.6208476517754868,
      "grad_norm": 10.548551559448242,
      "learning_rate": 2.582576274860015e-07,
      "loss": 0.6782,
      "step": 2830
    },
    {
      "epoch": 1.6265750286368843,
      "grad_norm": 9.971668243408203,
      "learning_rate": 2.507376385327634e-07,
      "loss": 0.7375,
      "step": 2840
    },
    {
      "epoch": 1.6323024054982818,
      "grad_norm": 11.734025001525879,
      "learning_rate": 2.4331876141672204e-07,
      "loss": 0.6543,
      "step": 2850
    },
    {
      "epoch": 1.6380297823596792,
      "grad_norm": 10.10330581665039,
      "learning_rate": 2.360015966012537e-07,
      "loss": 0.7008,
      "step": 2860
    },
    {
      "epoch": 1.6437571592210767,
      "grad_norm": 8.163260459899902,
      "learning_rate": 2.2878673631742236e-07,
      "loss": 0.758,
      "step": 2870
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 9.96818733215332,
      "learning_rate": 2.2167476451604624e-07,
      "loss": 0.7625,
      "step": 2880
    },
    {
      "epoch": 1.6552119129438716,
      "grad_norm": 11.303291320800781,
      "learning_rate": 2.1466625682043516e-07,
      "loss": 0.6977,
      "step": 2890
    },
    {
      "epoch": 1.660939289805269,
      "grad_norm": 11.4495210647583,
      "learning_rate": 2.0776178047979978e-07,
      "loss": 0.7007,
      "step": 2900
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 13.028532981872559,
      "learning_rate": 2.0096189432334195e-07,
      "loss": 0.6004,
      "step": 2910
    },
    {
      "epoch": 1.6723940435280642,
      "grad_norm": 10.361539840698242,
      "learning_rate": 1.9426714871502292e-07,
      "loss": 0.7078,
      "step": 2920
    },
    {
      "epoch": 1.6781214203894617,
      "grad_norm": 10.60468864440918,
      "learning_rate": 1.8767808550901911e-07,
      "loss": 0.7244,
      "step": 2930
    },
    {
      "epoch": 1.6838487972508591,
      "grad_norm": 9.203190803527832,
      "learning_rate": 1.811952380058657e-07,
      "loss": 0.636,
      "step": 2940
    },
    {
      "epoch": 1.6895761741122566,
      "grad_norm": 11.939068794250488,
      "learning_rate": 1.7481913090929314e-07,
      "loss": 0.7098,
      "step": 2950
    },
    {
      "epoch": 1.695303550973654,
      "grad_norm": 9.831012725830078,
      "learning_rate": 1.6855028028375897e-07,
      "loss": 0.6872,
      "step": 2960
    },
    {
      "epoch": 1.7010309278350515,
      "grad_norm": 18.38079261779785,
      "learning_rate": 1.623891935126786e-07,
      "loss": 0.6811,
      "step": 2970
    },
    {
      "epoch": 1.7067583046964492,
      "grad_norm": 9.422789573669434,
      "learning_rate": 1.5633636925735933e-07,
      "loss": 0.6251,
      "step": 2980
    },
    {
      "epoch": 1.7124856815578466,
      "grad_norm": 9.562675476074219,
      "learning_rate": 1.503922974166406e-07,
      "loss": 0.7189,
      "step": 2990
    },
    {
      "epoch": 1.718213058419244,
      "grad_norm": 9.7070894241333,
      "learning_rate": 1.4455745908724226e-07,
      "loss": 0.6947,
      "step": 3000
    },
    {
      "epoch": 1.718213058419244,
      "eval_loss": 1.150073528289795,
      "eval_runtime": 5.3064,
      "eval_samples_per_second": 67.842,
      "eval_steps_per_second": 16.961,
      "step": 3000
    },
    {
      "epoch": 1.7239404352806416,
      "grad_norm": 12.27785873413086,
      "learning_rate": 1.3883232652482553e-07,
      "loss": 0.8112,
      "step": 3010
    },
    {
      "epoch": 1.729667812142039,
      "grad_norm": 12.872515678405762,
      "learning_rate": 1.3321736310577164e-07,
      "loss": 0.7575,
      "step": 3020
    },
    {
      "epoch": 1.7353951890034365,
      "grad_norm": 11.842426300048828,
      "learning_rate": 1.2771302328967577e-07,
      "loss": 0.7448,
      "step": 3030
    },
    {
      "epoch": 1.741122565864834,
      "grad_norm": 11.257376670837402,
      "learning_rate": 1.2231975258256462e-07,
      "loss": 0.6947,
      "step": 3040
    },
    {
      "epoch": 1.7468499427262314,
      "grad_norm": 12.237911224365234,
      "learning_rate": 1.1703798750083938e-07,
      "loss": 0.684,
      "step": 3050
    },
    {
      "epoch": 1.7525773195876289,
      "grad_norm": 11.303350448608398,
      "learning_rate": 1.1186815553594382e-07,
      "loss": 0.7047,
      "step": 3060
    },
    {
      "epoch": 1.7583046964490263,
      "grad_norm": 10.056689262390137,
      "learning_rate": 1.068106751197659e-07,
      "loss": 0.6862,
      "step": 3070
    },
    {
      "epoch": 1.7640320733104238,
      "grad_norm": 9.765776634216309,
      "learning_rate": 1.0186595559076939e-07,
      "loss": 0.7131,
      "step": 3080
    },
    {
      "epoch": 1.7697594501718212,
      "grad_norm": 10.987778663635254,
      "learning_rate": 9.703439716086493e-08,
      "loss": 0.6012,
      "step": 3090
    },
    {
      "epoch": 1.7754868270332187,
      "grad_norm": 14.675382614135742,
      "learning_rate": 9.231639088301635e-08,
      "loss": 0.6843,
      "step": 3100
    },
    {
      "epoch": 1.7812142038946162,
      "grad_norm": 15.27701187133789,
      "learning_rate": 8.771231861959106e-08,
      "loss": 0.7452,
      "step": 3110
    },
    {
      "epoch": 1.7869415807560136,
      "grad_norm": 12.017073631286621,
      "learning_rate": 8.322255301145204e-08,
      "loss": 0.7111,
      "step": 3120
    },
    {
      "epoch": 1.792668957617411,
      "grad_norm": 8.804431915283203,
      "learning_rate": 7.884745744779937e-08,
      "loss": 0.7718,
      "step": 3130
    },
    {
      "epoch": 1.7983963344788088,
      "grad_norm": 10.447242736816406,
      "learning_rate": 7.458738603675602e-08,
      "loss": 0.8338,
      "step": 3140
    },
    {
      "epoch": 1.8041237113402062,
      "grad_norm": 13.058491706848145,
      "learning_rate": 7.044268357670869e-08,
      "loss": 0.5838,
      "step": 3150
    },
    {
      "epoch": 1.8098510882016037,
      "grad_norm": 11.285879135131836,
      "learning_rate": 6.641368552840088e-08,
      "loss": 0.7741,
      "step": 3160
    },
    {
      "epoch": 1.8155784650630011,
      "grad_norm": 8.553770065307617,
      "learning_rate": 6.250071798778117e-08,
      "loss": 0.7114,
      "step": 3170
    },
    {
      "epoch": 1.8213058419243986,
      "grad_norm": 8.297019004821777,
      "learning_rate": 5.870409765960966e-08,
      "loss": 0.761,
      "step": 3180
    },
    {
      "epoch": 1.827033218785796,
      "grad_norm": 9.877187728881836,
      "learning_rate": 5.502413183182592e-08,
      "loss": 0.7156,
      "step": 3190
    },
    {
      "epoch": 1.8327605956471937,
      "grad_norm": 11.764557838439941,
      "learning_rate": 5.146111835067646e-08,
      "loss": 0.6779,
      "step": 3200
    },
    {
      "epoch": 1.8384879725085912,
      "grad_norm": 11.132621765136719,
      "learning_rate": 4.8015345596608805e-08,
      "loss": 0.7472,
      "step": 3210
    },
    {
      "epoch": 1.8442153493699887,
      "grad_norm": 11.750364303588867,
      "learning_rate": 4.468709246093133e-08,
      "loss": 0.6553,
      "step": 3220
    },
    {
      "epoch": 1.8499427262313861,
      "grad_norm": 14.244782447814941,
      "learning_rate": 4.147662832323884e-08,
      "loss": 0.8058,
      "step": 3230
    },
    {
      "epoch": 1.8556701030927836,
      "grad_norm": 10.789249420166016,
      "learning_rate": 3.838421302961098e-08,
      "loss": 0.7955,
      "step": 3240
    },
    {
      "epoch": 1.861397479954181,
      "grad_norm": 9.885890007019043,
      "learning_rate": 3.541009687158109e-08,
      "loss": 0.6453,
      "step": 3250
    },
    {
      "epoch": 1.8671248568155785,
      "grad_norm": 9.926796913146973,
      "learning_rate": 3.255452056587754e-08,
      "loss": 0.7055,
      "step": 3260
    },
    {
      "epoch": 1.872852233676976,
      "grad_norm": 12.368999481201172,
      "learning_rate": 2.98177152349417e-08,
      "loss": 0.6359,
      "step": 3270
    },
    {
      "epoch": 1.8785796105383734,
      "grad_norm": 10.849148750305176,
      "learning_rate": 2.7199902388220343e-08,
      "loss": 0.6735,
      "step": 3280
    },
    {
      "epoch": 1.8843069873997709,
      "grad_norm": 12.963729858398438,
      "learning_rate": 2.47012939042387e-08,
      "loss": 0.6878,
      "step": 3290
    },
    {
      "epoch": 1.8900343642611683,
      "grad_norm": 10.135659217834473,
      "learning_rate": 2.2322092013450313e-08,
      "loss": 0.657,
      "step": 3300
    },
    {
      "epoch": 1.8957617411225658,
      "grad_norm": 16.199338912963867,
      "learning_rate": 2.0062489281870156e-08,
      "loss": 0.8076,
      "step": 3310
    },
    {
      "epoch": 1.9014891179839633,
      "grad_norm": 12.090662002563477,
      "learning_rate": 1.7922668595487744e-08,
      "loss": 0.7334,
      "step": 3320
    },
    {
      "epoch": 1.9072164948453607,
      "grad_norm": 10.486949920654297,
      "learning_rate": 1.5902803145465994e-08,
      "loss": 0.7883,
      "step": 3330
    },
    {
      "epoch": 1.9129438717067582,
      "grad_norm": 8.531800270080566,
      "learning_rate": 1.4003056414122583e-08,
      "loss": 0.6773,
      "step": 3340
    },
    {
      "epoch": 1.9186712485681556,
      "grad_norm": 10.624215126037598,
      "learning_rate": 1.222358216169922e-08,
      "loss": 0.6701,
      "step": 3350
    },
    {
      "epoch": 1.9243986254295533,
      "grad_norm": 12.858296394348145,
      "learning_rate": 1.0564524413915422e-08,
      "loss": 0.6576,
      "step": 3360
    },
    {
      "epoch": 1.9301260022909508,
      "grad_norm": 10.357311248779297,
      "learning_rate": 9.026017450312673e-09,
      "loss": 0.5779,
      "step": 3370
    },
    {
      "epoch": 1.9358533791523482,
      "grad_norm": 8.941463470458984,
      "learning_rate": 7.608185793385458e-09,
      "loss": 0.7369,
      "step": 3380
    },
    {
      "epoch": 1.9415807560137457,
      "grad_norm": 8.527290344238281,
      "learning_rate": 6.311144198503316e-09,
      "loss": 0.6539,
      "step": 3390
    },
    {
      "epoch": 1.9473081328751431,
      "grad_norm": 9.527642250061035,
      "learning_rate": 5.134997644622108e-09,
      "loss": 0.6612,
      "step": 3400
    },
    {
      "epoch": 1.9530355097365406,
      "grad_norm": 9.823323249816895,
      "learning_rate": 4.079841325787981e-09,
      "loss": 0.7576,
      "step": 3410
    },
    {
      "epoch": 1.9587628865979383,
      "grad_norm": 12.560805320739746,
      "learning_rate": 3.145760643432527e-09,
      "loss": 0.6648,
      "step": 3420
    },
    {
      "epoch": 1.9644902634593358,
      "grad_norm": 11.428935050964355,
      "learning_rate": 2.33283119946015e-09,
      "loss": 0.7002,
      "step": 3430
    },
    {
      "epoch": 1.9702176403207332,
      "grad_norm": 9.424127578735352,
      "learning_rate": 1.6411187901297919e-09,
      "loss": 0.6452,
      "step": 3440
    },
    {
      "epoch": 1.9759450171821307,
      "grad_norm": 10.48491096496582,
      "learning_rate": 1.0706794007290243e-09,
      "loss": 0.5779,
      "step": 3450
    },
    {
      "epoch": 1.9816723940435281,
      "grad_norm": 9.892742156982422,
      "learning_rate": 6.215592010426763e-10,
      "loss": 0.7057,
      "step": 3460
    },
    {
      "epoch": 1.9873997709049256,
      "grad_norm": 10.53469181060791,
      "learning_rate": 2.937945416166543e-10,
      "loss": 0.684,
      "step": 3470
    },
    {
      "epoch": 1.993127147766323,
      "grad_norm": 10.819581031799316,
      "learning_rate": 8.741195081479747e-11,
      "loss": 0.6306,
      "step": 3480
    },
    {
      "epoch": 1.9988545246277205,
      "grad_norm": 13.76382064819336,
      "learning_rate": 2.428132672926964e-12,
      "loss": 0.6657,
      "step": 3490
    },
    {
      "epoch": 2.0,
      "step": 3492,
      "total_flos": 8.351047625159475e+16,
      "train_loss": 0.9289274918937465,
      "train_runtime": 3314.9598,
      "train_samples_per_second": 4.212,
      "train_steps_per_second": 1.053
    }
  ],
  "logging_steps": 10,
  "max_steps": 3492,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.351047625159475e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}