{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 50000,
  "global_step": 2364,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01692047377326565,
      "grad_norm": 4.428616046905518,
      "learning_rate": 7.042253521126762e-07,
      "loss": 0.6706,
      "step": 10
    },
    {
      "epoch": 0.0338409475465313,
      "grad_norm": 6.486809730529785,
      "learning_rate": 1.4084507042253523e-06,
      "loss": 0.6483,
      "step": 20
    },
    {
      "epoch": 0.050761421319796954,
      "grad_norm": 3.8806605339050293,
      "learning_rate": 2.1126760563380285e-06,
      "loss": 0.5231,
      "step": 30
    },
    {
      "epoch": 0.0676818950930626,
      "grad_norm": 2.814377784729004,
      "learning_rate": 2.8169014084507046e-06,
      "loss": 0.3641,
      "step": 40
    },
    {
      "epoch": 0.08460236886632826,
      "grad_norm": 2.3236453533172607,
      "learning_rate": 3.5211267605633804e-06,
      "loss": 0.2923,
      "step": 50
    },
    {
      "epoch": 0.10152284263959391,
      "grad_norm": 1.732696294784546,
      "learning_rate": 4.225352112676057e-06,
      "loss": 0.2632,
      "step": 60
    },
    {
      "epoch": 0.11844331641285956,
      "grad_norm": 1.71094810962677,
      "learning_rate": 4.929577464788733e-06,
      "loss": 0.2672,
      "step": 70
    },
    {
      "epoch": 0.1353637901861252,
      "grad_norm": 2.0413691997528076,
      "learning_rate": 4.999809944183634e-06,
      "loss": 0.2208,
      "step": 80
    },
    {
      "epoch": 0.15228426395939088,
      "grad_norm": 2.169725179672241,
      "learning_rate": 4.999152998213973e-06,
      "loss": 0.252,
      "step": 90
    },
    {
      "epoch": 0.1692047377326565,
      "grad_norm": 1.3290756940841675,
      "learning_rate": 4.998026939008151e-06,
      "loss": 0.2468,
      "step": 100
    },
    {
      "epoch": 0.18612521150592218,
      "grad_norm": 2.0767831802368164,
      "learning_rate": 4.996431977937491e-06,
      "loss": 0.2468,
      "step": 110
    },
    {
      "epoch": 0.20304568527918782,
      "grad_norm": 1.3648345470428467,
      "learning_rate": 4.994368414390368e-06,
      "loss": 0.2068,
      "step": 120
    },
    {
      "epoch": 0.21996615905245348,
      "grad_norm": 1.9443793296813965,
      "learning_rate": 4.991836635716003e-06,
      "loss": 0.233,
      "step": 130
    },
    {
      "epoch": 0.23688663282571912,
      "grad_norm": 1.4086923599243164,
      "learning_rate": 4.988837117151762e-06,
      "loss": 0.2153,
      "step": 140
    },
    {
      "epoch": 0.25380710659898476,
      "grad_norm": 1.4785760641098022,
      "learning_rate": 4.985370421733948e-06,
      "loss": 0.2316,
      "step": 150
    },
    {
      "epoch": 0.2707275803722504,
      "grad_norm": 1.52463960647583,
      "learning_rate": 4.981437200192111e-06,
      "loss": 0.215,
      "step": 160
    },
    {
      "epoch": 0.2876480541455161,
      "grad_norm": 1.5714839696884155,
      "learning_rate": 4.9770381908269005e-06,
      "loss": 0.2366,
      "step": 170
    },
    {
      "epoch": 0.30456852791878175,
      "grad_norm": 1.8277614116668701,
      "learning_rate": 4.972174219371484e-06,
      "loss": 0.2443,
      "step": 180
    },
    {
      "epoch": 0.32148900169204736,
      "grad_norm": 1.175654411315918,
      "learning_rate": 4.966846198836546e-06,
      "loss": 0.2226,
      "step": 190
    },
    {
      "epoch": 0.338409475465313,
      "grad_norm": 1.5629938840866089,
      "learning_rate": 4.96105512933891e-06,
      "loss": 0.2231,
      "step": 200
    },
    {
      "epoch": 0.3553299492385787,
      "grad_norm": 1.6523264646530151,
      "learning_rate": 4.954802097913804e-06,
      "loss": 0.2193,
      "step": 210
    },
    {
      "epoch": 0.37225042301184436,
      "grad_norm": 1.222960114479065,
      "learning_rate": 4.94808827831082e-06,
      "loss": 0.2534,
      "step": 220
    },
    {
      "epoch": 0.38917089678510997,
      "grad_norm": 1.26102614402771,
      "learning_rate": 4.9409149307735835e-06,
      "loss": 0.2283,
      "step": 230
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 1.4353593587875366,
      "learning_rate": 4.9332834018032025e-06,
      "loss": 0.2594,
      "step": 240
    },
    {
      "epoch": 0.4230118443316413,
      "grad_norm": 1.1220771074295044,
      "learning_rate": 4.925195123905513e-06,
      "loss": 0.2262,
      "step": 250
    },
    {
      "epoch": 0.43993231810490696,
      "grad_norm": 1.7100379467010498,
      "learning_rate": 4.916651615322182e-06,
      "loss": 0.2093,
      "step": 260
    },
    {
      "epoch": 0.45685279187817257,
      "grad_norm": 1.975794792175293,
      "learning_rate": 4.907654479745726e-06,
      "loss": 0.2101,
      "step": 270
    },
    {
      "epoch": 0.47377326565143824,
      "grad_norm": 1.3200404644012451,
      "learning_rate": 4.89820540601848e-06,
      "loss": 0.2462,
      "step": 280
    },
    {
      "epoch": 0.4906937394247039,
      "grad_norm": 1.254960298538208,
      "learning_rate": 4.8883061678155905e-06,
      "loss": 0.2051,
      "step": 290
    },
    {
      "epoch": 0.5076142131979695,
      "grad_norm": 1.7000998258590698,
      "learning_rate": 4.877958623312076e-06,
      "loss": 0.2531,
      "step": 300
    },
    {
      "epoch": 0.5245346869712352,
      "grad_norm": 1.2153050899505615,
      "learning_rate": 4.8671647148340304e-06,
      "loss": 0.2484,
      "step": 310
    },
    {
      "epoch": 0.5414551607445008,
      "grad_norm": 1.4881746768951416,
      "learning_rate": 4.855926468494039e-06,
      "loss": 0.2121,
      "step": 320
    },
    {
      "epoch": 0.5583756345177665,
      "grad_norm": 1.3982009887695312,
      "learning_rate": 4.844245993810852e-06,
      "loss": 0.237,
      "step": 330
    },
    {
      "epoch": 0.5752961082910322,
      "grad_norm": 1.4024406671524048,
      "learning_rate": 4.832125483313411e-06,
      "loss": 0.2126,
      "step": 340
    },
    {
      "epoch": 0.5922165820642978,
      "grad_norm": 1.4790270328521729,
      "learning_rate": 4.819567212129292e-06,
      "loss": 0.2491,
      "step": 350
    },
    {
      "epoch": 0.6091370558375635,
      "grad_norm": 1.5148818492889404,
      "learning_rate": 4.806573537557643e-06,
      "loss": 0.2359,
      "step": 360
    },
    {
      "epoch": 0.626057529610829,
      "grad_norm": 1.7280242443084717,
      "learning_rate": 4.7931468986267014e-06,
      "loss": 0.2363,
      "step": 370
    },
    {
      "epoch": 0.6429780033840947,
      "grad_norm": 1.5051395893096924,
      "learning_rate": 4.779289815635961e-06,
      "loss": 0.2132,
      "step": 380
    },
    {
      "epoch": 0.6598984771573604,
      "grad_norm": 2.0735015869140625,
      "learning_rate": 4.765004889683096e-06,
      "loss": 0.2307,
      "step": 390
    },
    {
      "epoch": 0.676818950930626,
      "grad_norm": 1.4238276481628418,
      "learning_rate": 4.750294802175703e-06,
      "loss": 0.2237,
      "step": 400
    },
    {
      "epoch": 0.6937394247038917,
      "grad_norm": 1.283726453781128,
      "learning_rate": 4.735162314327987e-06,
      "loss": 0.2299,
      "step": 410
    },
    {
      "epoch": 0.7106598984771574,
      "grad_norm": 1.8388890027999878,
      "learning_rate": 4.71961026664245e-06,
      "loss": 0.23,
      "step": 420
    },
    {
      "epoch": 0.727580372250423,
      "grad_norm": 1.456938624382019,
      "learning_rate": 4.703641578376706e-06,
      "loss": 0.1979,
      "step": 430
    },
    {
      "epoch": 0.7445008460236887,
      "grad_norm": 1.3884906768798828,
      "learning_rate": 4.687259246995512e-06,
      "loss": 0.2358,
      "step": 440
    },
    {
      "epoch": 0.7614213197969543,
      "grad_norm": 1.8143761157989502,
      "learning_rate": 4.670466347608109e-06,
      "loss": 0.2382,
      "step": 450
    },
    {
      "epoch": 0.7783417935702199,
      "grad_norm": 1.3720914125442505,
      "learning_rate": 4.65326603239101e-06,
      "loss": 0.1786,
      "step": 460
    },
    {
      "epoch": 0.7952622673434856,
      "grad_norm": 1.378342866897583,
      "learning_rate": 4.6356615299963e-06,
      "loss": 0.2109,
      "step": 470
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 1.6520991325378418,
      "learning_rate": 4.617656144945584e-06,
      "loss": 0.2296,
      "step": 480
    },
    {
      "epoch": 0.8291032148900169,
      "grad_norm": 1.617244839668274,
      "learning_rate": 4.599253257009716e-06,
      "loss": 0.21,
      "step": 490
    },
    {
      "epoch": 0.8460236886632826,
      "grad_norm": 1.8561798334121704,
      "learning_rate": 4.580456320574367e-06,
      "loss": 0.1935,
      "step": 500
    },
    {
      "epoch": 0.8629441624365483,
      "grad_norm": 2.1917858123779297,
      "learning_rate": 4.561268863991618e-06,
      "loss": 0.2246,
      "step": 510
    },
    {
      "epoch": 0.8798646362098139,
      "grad_norm": 1.4701884984970093,
      "learning_rate": 4.541694488917654e-06,
      "loss": 0.2234,
      "step": 520
    },
    {
      "epoch": 0.8967851099830795,
      "grad_norm": 1.3482675552368164,
      "learning_rate": 4.521736869636699e-06,
      "loss": 0.2423,
      "step": 530
    },
    {
      "epoch": 0.9137055837563451,
      "grad_norm": 1.618618130683899,
      "learning_rate": 4.501399752371323e-06,
      "loss": 0.2384,
      "step": 540
    },
    {
      "epoch": 0.9306260575296108,
      "grad_norm": 1.7225276231765747,
      "learning_rate": 4.480686954579241e-06,
      "loss": 0.2208,
      "step": 550
    },
    {
      "epoch": 0.9475465313028765,
      "grad_norm": 1.5978412628173828,
      "learning_rate": 4.459602364236743e-06,
      "loss": 0.2421,
      "step": 560
    },
    {
      "epoch": 0.9644670050761421,
      "grad_norm": 1.34050452709198,
      "learning_rate": 4.438149939108887e-06,
      "loss": 0.2117,
      "step": 570
    },
    {
      "epoch": 0.9813874788494078,
      "grad_norm": 1.7396200895309448,
      "learning_rate": 4.4163337060065895e-06,
      "loss": 0.2395,
      "step": 580
    },
    {
      "epoch": 0.9983079526226735,
      "grad_norm": 1.5188558101654053,
      "learning_rate": 4.394157760030756e-06,
      "loss": 0.2414,
      "step": 590
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 1.3064625263214111,
      "learning_rate": 4.371626263803601e-06,
      "loss": 0.1639,
      "step": 600
    },
    {
      "epoch": 1.0321489001692048,
      "grad_norm": 1.2497655153274536,
      "learning_rate": 4.348743446687279e-06,
      "loss": 0.1606,
      "step": 610
    },
    {
      "epoch": 1.0490693739424704,
      "grad_norm": 1.3260619640350342,
      "learning_rate": 4.325513603989998e-06,
      "loss": 0.1373,
      "step": 620
    },
    {
      "epoch": 1.0659898477157361,
      "grad_norm": 1.7090791463851929,
      "learning_rate": 4.301941096159757e-06,
      "loss": 0.151,
      "step": 630
    },
    {
      "epoch": 1.0829103214890017,
      "grad_norm": 1.4625272750854492,
      "learning_rate": 4.278030347965842e-06,
      "loss": 0.1567,
      "step": 640
    },
    {
      "epoch": 1.0998307952622675,
      "grad_norm": 1.374337911605835,
      "learning_rate": 4.253785847668264e-06,
      "loss": 0.1412,
      "step": 650
    },
    {
      "epoch": 1.116751269035533,
      "grad_norm": 1.326503038406372,
      "learning_rate": 4.229212146175273e-06,
      "loss": 0.1249,
      "step": 660
    },
    {
      "epoch": 1.1336717428087986,
      "grad_norm": 1.2745444774627686,
      "learning_rate": 4.2043138561891086e-06,
      "loss": 0.1697,
      "step": 670
    },
    {
      "epoch": 1.1505922165820643,
      "grad_norm": 1.3895198106765747,
      "learning_rate": 4.179095651340155e-06,
      "loss": 0.1429,
      "step": 680
    },
    {
      "epoch": 1.16751269035533,
      "grad_norm": 1.331308126449585,
      "learning_rate": 4.153562265309662e-06,
      "loss": 0.1196,
      "step": 690
    },
    {
      "epoch": 1.1844331641285957,
      "grad_norm": 1.7868833541870117,
      "learning_rate": 4.1277184909411885e-06,
      "loss": 0.1512,
      "step": 700
    },
    {
      "epoch": 1.2013536379018612,
      "grad_norm": 1.2682002782821655,
      "learning_rate": 4.101569179340946e-06,
      "loss": 0.1795,
      "step": 710
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 1.1160831451416016,
      "learning_rate": 4.075119238967196e-06,
      "loss": 0.1353,
      "step": 720
    },
    {
      "epoch": 1.2351945854483926,
      "grad_norm": 1.2749055624008179,
      "learning_rate": 4.048373634708899e-06,
      "loss": 0.1452,
      "step": 730
    },
    {
      "epoch": 1.252115059221658,
      "grad_norm": 1.6326930522918701,
      "learning_rate": 4.0213373869537504e-06,
      "loss": 0.1498,
      "step": 740
    },
    {
      "epoch": 1.2690355329949239,
      "grad_norm": 1.2723865509033203,
      "learning_rate": 3.994015570645818e-06,
      "loss": 0.1221,
      "step": 750
    },
    {
      "epoch": 1.2859560067681894,
      "grad_norm": 1.4743432998657227,
      "learning_rate": 3.966413314332924e-06,
      "loss": 0.1436,
      "step": 760
    },
    {
      "epoch": 1.3028764805414552,
      "grad_norm": 1.3993955850601196,
      "learning_rate": 3.938535799203971e-06,
      "loss": 0.1456,
      "step": 770
    },
    {
      "epoch": 1.3197969543147208,
      "grad_norm": 1.7814866304397583,
      "learning_rate": 3.91038825811639e-06,
      "loss": 0.1329,
      "step": 780
    },
    {
      "epoch": 1.3367174280879865,
      "grad_norm": 1.551604151725769,
      "learning_rate": 3.881975974613888e-06,
      "loss": 0.1636,
      "step": 790
    },
    {
      "epoch": 1.353637901861252,
      "grad_norm": 1.244645118713379,
      "learning_rate": 3.85330428193467e-06,
      "loss": 0.1291,
      "step": 800
    },
    {
      "epoch": 1.3705583756345177,
      "grad_norm": 1.368085503578186,
      "learning_rate": 3.824378562010358e-06,
      "loss": 0.1292,
      "step": 810
    },
    {
      "epoch": 1.3874788494077834,
      "grad_norm": 1.6551761627197266,
      "learning_rate": 3.7952042444557396e-06,
      "loss": 0.158,
      "step": 820
    },
    {
      "epoch": 1.404399323181049,
      "grad_norm": 1.4006320238113403,
      "learning_rate": 3.7657868055495917e-06,
      "loss": 0.146,
      "step": 830
    },
    {
      "epoch": 1.4213197969543148,
      "grad_norm": 1.8334702253341675,
      "learning_rate": 3.736131767206727e-06,
      "loss": 0.1572,
      "step": 840
    },
    {
      "epoch": 1.4382402707275803,
      "grad_norm": 1.3605976104736328,
      "learning_rate": 3.706244695941489e-06,
      "loss": 0.1274,
      "step": 850
    },
    {
      "epoch": 1.455160744500846,
      "grad_norm": 1.306801199913025,
      "learning_rate": 3.6761312018228597e-06,
      "loss": 0.1582,
      "step": 860
    },
    {
      "epoch": 1.4720812182741116,
      "grad_norm": 1.4271026849746704,
      "learning_rate": 3.645796937421406e-06,
      "loss": 0.1197,
      "step": 870
    },
    {
      "epoch": 1.4890016920473772,
      "grad_norm": 1.3257379531860352,
      "learning_rate": 3.6152475967482393e-06,
      "loss": 0.1392,
      "step": 880
    },
    {
      "epoch": 1.505922165820643,
      "grad_norm": 1.4457297325134277,
      "learning_rate": 3.5844889141861977e-06,
      "loss": 0.1334,
      "step": 890
    },
    {
      "epoch": 1.5228426395939088,
      "grad_norm": 1.5230379104614258,
      "learning_rate": 3.5535266634134556e-06,
      "loss": 0.1325,
      "step": 900
    },
    {
      "epoch": 1.5397631133671743,
      "grad_norm": 1.1343902349472046,
      "learning_rate": 3.522366656319748e-06,
      "loss": 0.1315,
      "step": 910
    },
    {
      "epoch": 1.5566835871404399,
      "grad_norm": 1.58329439163208,
      "learning_rate": 3.4910147419154306e-06,
      "loss": 0.1464,
      "step": 920
    },
    {
      "epoch": 1.5736040609137056,
      "grad_norm": 1.3953927755355835,
      "learning_rate": 3.459476805233567e-06,
      "loss": 0.132,
      "step": 930
    },
    {
      "epoch": 1.5905245346869712,
      "grad_norm": 1.5068899393081665,
      "learning_rate": 3.427758766225257e-06,
      "loss": 0.1348,
      "step": 940
    },
    {
      "epoch": 1.6074450084602367,
      "grad_norm": 1.7132457494735718,
      "learning_rate": 3.3958665786484063e-06,
      "loss": 0.1343,
      "step": 950
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 1.1419628858566284,
      "learning_rate": 3.3638062289501584e-06,
      "loss": 0.1381,
      "step": 960
    },
    {
      "epoch": 1.6412859560067683,
      "grad_norm": 1.8477777242660522,
      "learning_rate": 3.331583735143179e-06,
      "loss": 0.1642,
      "step": 970
    },
    {
      "epoch": 1.6582064297800339,
      "grad_norm": 1.5605595111846924,
      "learning_rate": 3.2992051456760242e-06,
      "loss": 0.1479,
      "step": 980
    },
    {
      "epoch": 1.6751269035532994,
      "grad_norm": 1.643684983253479,
      "learning_rate": 3.266676538297797e-06,
      "loss": 0.1564,
      "step": 990
    },
    {
      "epoch": 1.6920473773265652,
      "grad_norm": 1.6141108274459839,
      "learning_rate": 3.234004018917293e-06,
      "loss": 0.1496,
      "step": 1000
    },
    {
      "epoch": 1.708967851099831,
      "grad_norm": 1.8253343105316162,
      "learning_rate": 3.20119372045687e-06,
      "loss": 0.1535,
      "step": 1010
    },
    {
      "epoch": 1.7258883248730963,
      "grad_norm": 1.3858433961868286,
      "learning_rate": 3.1682518017012452e-06,
      "loss": 0.1273,
      "step": 1020
    },
    {
      "epoch": 1.742808798646362,
      "grad_norm": 1.9302154779434204,
      "learning_rate": 3.1351844461414348e-06,
      "loss": 0.15,
      "step": 1030
    },
    {
      "epoch": 1.7597292724196278,
      "grad_norm": 1.4370014667510986,
      "learning_rate": 3.1019978608140584e-06,
      "loss": 0.1387,
      "step": 1040
    },
    {
      "epoch": 1.7766497461928934,
      "grad_norm": 1.0620018243789673,
      "learning_rate": 3.0686982751362234e-06,
      "loss": 0.1261,
      "step": 1050
    },
    {
      "epoch": 1.793570219966159,
      "grad_norm": 1.401853084564209,
      "learning_rate": 3.0352919397362064e-06,
      "loss": 0.1492,
      "step": 1060
    },
    {
      "epoch": 1.8104906937394247,
      "grad_norm": 1.4132765531539917,
      "learning_rate": 3.0017851252801574e-06,
      "loss": 0.1401,
      "step": 1070
    },
    {
      "epoch": 1.8274111675126905,
      "grad_norm": 1.540830373764038,
      "learning_rate": 2.968184121295038e-06,
      "loss": 0.1327,
      "step": 1080
    },
    {
      "epoch": 1.844331641285956,
      "grad_norm": 1.434097170829773,
      "learning_rate": 2.93449523498802e-06,
      "loss": 0.1501,
      "step": 1090
    },
    {
      "epoch": 1.8612521150592216,
      "grad_norm": 1.495847463607788,
      "learning_rate": 2.900724790062571e-06,
      "loss": 0.1294,
      "step": 1100
    },
    {
      "epoch": 1.8781725888324874,
      "grad_norm": 1.7926485538482666,
      "learning_rate": 2.866879125531437e-06,
      "loss": 0.1286,
      "step": 1110
    },
    {
      "epoch": 1.895093062605753,
      "grad_norm": 1.778939962387085,
      "learning_rate": 2.832964594526748e-06,
      "loss": 0.1617,
      "step": 1120
    },
    {
      "epoch": 1.9120135363790185,
      "grad_norm": 1.9225724935531616,
      "learning_rate": 2.798987563107488e-06,
      "loss": 0.1467,
      "step": 1130
    },
    {
      "epoch": 1.9289340101522843,
      "grad_norm": 1.0806710720062256,
      "learning_rate": 2.7649544090645226e-06,
      "loss": 0.156,
      "step": 1140
    },
    {
      "epoch": 1.94585448392555,
      "grad_norm": 1.5917881727218628,
      "learning_rate": 2.7308715207234326e-06,
      "loss": 0.1377,
      "step": 1150
    },
    {
      "epoch": 1.9627749576988156,
      "grad_norm": 1.3541003465652466,
      "learning_rate": 2.69674529574537e-06,
      "loss": 0.141,
      "step": 1160
    },
    {
      "epoch": 1.9796954314720812,
      "grad_norm": 1.592405915260315,
      "learning_rate": 2.6625821399261562e-06,
      "loss": 0.1578,
      "step": 1170
    },
    {
      "epoch": 1.996615905245347,
      "grad_norm": 1.7678178548812866,
      "learning_rate": 2.628388465993864e-06,
      "loss": 0.172,
      "step": 1180
    },
    {
      "epoch": 2.0135363790186127,
      "grad_norm": 0.7831181287765503,
      "learning_rate": 2.594170692405083e-06,
      "loss": 0.1007,
      "step": 1190
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 1.5708684921264648,
      "learning_rate": 2.559935242140125e-06,
      "loss": 0.0863,
      "step": 1200
    },
    {
      "epoch": 2.047377326565144,
      "grad_norm": 1.2178500890731812,
      "learning_rate": 2.5256885414973715e-06,
      "loss": 0.0815,
      "step": 1210
    },
    {
      "epoch": 2.0642978003384096,
      "grad_norm": 1.4404493570327759,
      "learning_rate": 2.491437018886998e-06,
      "loss": 0.0924,
      "step": 1220
    },
    {
      "epoch": 2.081218274111675,
      "grad_norm": 1.701210618019104,
      "learning_rate": 2.457187103624308e-06,
      "loss": 0.0843,
      "step": 1230
    },
    {
      "epoch": 2.0981387478849407,
      "grad_norm": 1.3683816194534302,
      "learning_rate": 2.4229452247228895e-06,
      "loss": 0.0712,
      "step": 1240
    },
    {
      "epoch": 2.1150592216582065,
      "grad_norm": 2.7988007068634033,
      "learning_rate": 2.3887178096878363e-06,
      "loss": 0.0883,
      "step": 1250
    },
    {
      "epoch": 2.1319796954314723,
      "grad_norm": 1.0649851560592651,
      "learning_rate": 2.354511283309244e-06,
      "loss": 0.0706,
      "step": 1260
    },
    {
      "epoch": 2.1489001692047376,
      "grad_norm": 2.3365893363952637,
      "learning_rate": 2.320332066456224e-06,
      "loss": 0.077,
      "step": 1270
    },
    {
      "epoch": 2.1658206429780034,
      "grad_norm": 1.3339556455612183,
      "learning_rate": 2.2861865748716448e-06,
      "loss": 0.0864,
      "step": 1280
    },
    {
      "epoch": 2.182741116751269,
      "grad_norm": 1.6249768733978271,
      "learning_rate": 2.2520812179678422e-06,
      "loss": 0.0783,
      "step": 1290
    },
    {
      "epoch": 2.199661590524535,
      "grad_norm": 1.3000891208648682,
      "learning_rate": 2.218022397623517e-06,
      "loss": 0.075,
      "step": 1300
    },
    {
      "epoch": 2.2165820642978002,
      "grad_norm": 1.5419507026672363,
      "learning_rate": 2.1840165069820434e-06,
      "loss": 0.0745,
      "step": 1310
    },
    {
      "epoch": 2.233502538071066,
      "grad_norm": 1.4701499938964844,
      "learning_rate": 2.150069929251419e-06,
      "loss": 0.0652,
      "step": 1320
    },
    {
      "epoch": 2.250423011844332,
      "grad_norm": 1.1285984516143799,
      "learning_rate": 2.1161890365060838e-06,
      "loss": 0.0842,
      "step": 1330
    },
    {
      "epoch": 2.267343485617597,
      "grad_norm": 1.860626459121704,
      "learning_rate": 2.082380188490817e-06,
      "loss": 0.0824,
      "step": 1340
    },
    {
      "epoch": 2.284263959390863,
      "grad_norm": 1.9373877048492432,
      "learning_rate": 2.048649731426965e-06,
      "loss": 0.0861,
      "step": 1350
    },
    {
      "epoch": 2.3011844331641287,
      "grad_norm": 1.4356682300567627,
      "learning_rate": 2.0150039968211958e-06,
      "loss": 0.0793,
      "step": 1360
    },
    {
      "epoch": 2.3181049069373945,
      "grad_norm": 1.4798492193222046,
      "learning_rate": 1.9814493002770186e-06,
      "loss": 0.0761,
      "step": 1370
    },
    {
      "epoch": 2.33502538071066,
      "grad_norm": 2.065126895904541,
      "learning_rate": 1.9479919403092863e-06,
      "loss": 0.0802,
      "step": 1380
    },
    {
      "epoch": 2.3519458544839256,
      "grad_norm": 1.5476558208465576,
      "learning_rate": 1.914638197161914e-06,
      "loss": 0.0804,
      "step": 1390
    },
    {
      "epoch": 2.3688663282571913,
      "grad_norm": 2.891502618789673,
      "learning_rate": 1.8813943316290177e-06,
      "loss": 0.08,
      "step": 1400
    },
    {
      "epoch": 2.3857868020304567,
      "grad_norm": 2.407071113586426,
      "learning_rate": 1.8482665838797087e-06,
      "loss": 0.0958,
      "step": 1410
    },
    {
      "epoch": 2.4027072758037225,
      "grad_norm": 2.067172050476074,
      "learning_rate": 1.8152611722867575e-06,
      "loss": 0.0772,
      "step": 1420
    },
    {
      "epoch": 2.4196277495769882,
      "grad_norm": 1.3346515893936157,
      "learning_rate": 1.78238429225935e-06,
      "loss": 0.0794,
      "step": 1430
    },
    {
      "epoch": 2.436548223350254,
      "grad_norm": 1.4311262369155884,
      "learning_rate": 1.7496421150801547e-06,
      "loss": 0.1038,
      "step": 1440
    },
    {
      "epoch": 2.4534686971235193,
      "grad_norm": 1.4051713943481445,
      "learning_rate": 1.7170407867469185e-06,
      "loss": 0.0854,
      "step": 1450
    },
    {
      "epoch": 2.470389170896785,
      "grad_norm": 1.4341652393341064,
      "learning_rate": 1.6845864268188073e-06,
      "loss": 0.0722,
      "step": 1460
    },
    {
      "epoch": 2.487309644670051,
      "grad_norm": 2.1416544914245605,
      "learning_rate": 1.6522851272677126e-06,
      "loss": 0.0662,
      "step": 1470
    },
    {
      "epoch": 2.504230118443316,
      "grad_norm": 1.5771617889404297,
      "learning_rate": 1.6201429513347275e-06,
      "loss": 0.0949,
      "step": 1480
    },
    {
      "epoch": 2.521150592216582,
      "grad_norm": 1.555201768875122,
      "learning_rate": 1.5881659323920329e-06,
      "loss": 0.069,
      "step": 1490
    },
    {
      "epoch": 2.5380710659898478,
      "grad_norm": 2.297422170639038,
      "learning_rate": 1.556360072810371e-06,
      "loss": 0.0822,
      "step": 1500
    },
    {
      "epoch": 2.5549915397631136,
      "grad_norm": 1.4423418045043945,
      "learning_rate": 1.5247313428323521e-06,
      "loss": 0.0739,
      "step": 1510
    },
    {
      "epoch": 2.571912013536379,
      "grad_norm": 1.7135895490646362,
      "learning_rate": 1.4932856794517828e-06,
      "loss": 0.0701,
      "step": 1520
    },
    {
      "epoch": 2.5888324873096447,
      "grad_norm": 1.2873613834381104,
      "learning_rate": 1.4620289852992436e-06,
      "loss": 0.0765,
      "step": 1530
    },
    {
      "epoch": 2.6057529610829104,
      "grad_norm": 1.5120519399642944,
      "learning_rate": 1.4309671275341115e-06,
      "loss": 0.0733,
      "step": 1540
    },
    {
      "epoch": 2.6226734348561758,
      "grad_norm": 1.4765515327453613,
      "learning_rate": 1.4001059367432387e-06,
      "loss": 0.0759,
      "step": 1550
    },
    {
      "epoch": 2.6395939086294415,
      "grad_norm": 1.6146334409713745,
      "learning_rate": 1.3694512058465064e-06,
      "loss": 0.0731,
      "step": 1560
    },
    {
      "epoch": 2.6565143824027073,
      "grad_norm": 2.312690258026123,
      "learning_rate": 1.3390086890094346e-06,
      "loss": 0.0826,
      "step": 1570
    },
    {
      "epoch": 2.673434856175973,
      "grad_norm": 1.315255880355835,
      "learning_rate": 1.3087841005630774e-06,
      "loss": 0.0753,
      "step": 1580
    },
    {
      "epoch": 2.6903553299492384,
      "grad_norm": 1.6095589399337769,
      "learning_rate": 1.2787831139313957e-06,
      "loss": 0.0942,
      "step": 1590
    },
    {
      "epoch": 2.707275803722504,
      "grad_norm": 2.9680864810943604,
      "learning_rate": 1.249011360566301e-06,
      "loss": 0.083,
      "step": 1600
    },
    {
      "epoch": 2.72419627749577,
      "grad_norm": 1.3204880952835083,
      "learning_rate": 1.2194744288905807e-06,
      "loss": 0.087,
      "step": 1610
    },
    {
      "epoch": 2.7411167512690353,
      "grad_norm": 1.4213435649871826,
      "learning_rate": 1.1901778632489037e-06,
      "loss": 0.0793,
      "step": 1620
    },
    {
      "epoch": 2.758037225042301,
      "grad_norm": 2.258690595626831,
      "learning_rate": 1.1611271628671e-06,
      "loss": 0.0802,
      "step": 1630
    },
    {
      "epoch": 2.774957698815567,
      "grad_norm": 1.5296716690063477,
      "learning_rate": 1.132327780819901e-06,
      "loss": 0.0789,
      "step": 1640
    },
    {
      "epoch": 2.7918781725888326,
      "grad_norm": 1.8829957246780396,
      "learning_rate": 1.1037851230073563e-06,
      "loss": 0.0779,
      "step": 1650
    },
    {
      "epoch": 2.808798646362098,
      "grad_norm": 1.4642276763916016,
      "learning_rate": 1.0755045471400897e-06,
      "loss": 0.0813,
      "step": 1660
    },
    {
      "epoch": 2.8257191201353637,
      "grad_norm": 1.2218215465545654,
      "learning_rate": 1.0474913617336184e-06,
      "loss": 0.0763,
      "step": 1670
    },
    {
      "epoch": 2.8426395939086295,
      "grad_norm": 2.2098135948181152,
      "learning_rate": 1.0197508251118892e-06,
      "loss": 0.0731,
      "step": 1680
    },
    {
      "epoch": 2.859560067681895,
      "grad_norm": 1.3853759765625,
      "learning_rate": 9.922881444202514e-07,
      "loss": 0.0861,
      "step": 1690
    },
    {
      "epoch": 2.8764805414551606,
      "grad_norm": 1.280692458152771,
      "learning_rate": 9.651084746480229e-07,
      "loss": 0.078,
      "step": 1700
    },
    {
      "epoch": 2.8934010152284264,
      "grad_norm": 1.077873945236206,
      "learning_rate": 9.382169176608568e-07,
      "loss": 0.0807,
      "step": 1710
    },
    {
      "epoch": 2.910321489001692,
      "grad_norm": 1.4427660703659058,
      "learning_rate": 9.116185212430765e-07,
      "loss": 0.0866,
      "step": 1720
    },
    {
      "epoch": 2.927241962774958,
      "grad_norm": 1.9770188331604004,
      "learning_rate": 8.853182781501598e-07,
      "loss": 0.0728,
      "step": 1730
    },
    {
      "epoch": 2.9441624365482233,
      "grad_norm": 1.3886785507202148,
      "learning_rate": 8.593211251715533e-07,
      "loss": 0.0726,
      "step": 1740
    },
    {
      "epoch": 2.961082910321489,
      "grad_norm": 1.6050931215286255,
      "learning_rate": 8.336319422039945e-07,
      "loss": 0.0806,
      "step": 1750
    },
    {
      "epoch": 2.9780033840947544,
      "grad_norm": 2.0447897911071777,
      "learning_rate": 8.082555513355148e-07,
      "loss": 0.0809,
      "step": 1760
    },
    {
      "epoch": 2.99492385786802,
      "grad_norm": 1.4558676481246948,
      "learning_rate": 7.831967159402837e-07,
      "loss": 0.0871,
      "step": 1770
    },
    {
      "epoch": 3.011844331641286,
      "grad_norm": 1.2891719341278076,
      "learning_rate": 7.584601397844882e-07,
      "loss": 0.0494,
      "step": 1780
    },
    {
      "epoch": 3.0287648054145517,
      "grad_norm": 1.3743082284927368,
      "learning_rate": 7.340504661433918e-07,
      "loss": 0.0507,
      "step": 1790
    },
    {
      "epoch": 3.045685279187817,
      "grad_norm": 1.0931307077407837,
      "learning_rate": 7.099722769297445e-07,
      "loss": 0.0465,
      "step": 1800
    },
    {
      "epoch": 3.062605752961083,
      "grad_norm": 1.310681939125061,
      "learning_rate": 6.862300918337275e-07,
      "loss": 0.0527,
      "step": 1810
    },
    {
      "epoch": 3.0795262267343486,
      "grad_norm": 1.3822815418243408,
      "learning_rate": 6.628283674745608e-07,
      "loss": 0.0412,
      "step": 1820
    },
    {
      "epoch": 3.0964467005076144,
      "grad_norm": 1.4022243022918701,
      "learning_rate": 6.397714965639556e-07,
      "loss": 0.0557,
      "step": 1830
    },
    {
      "epoch": 3.1133671742808797,
      "grad_norm": 0.8449741005897522,
      "learning_rate": 6.170638070815668e-07,
      "loss": 0.0465,
      "step": 1840
    },
    {
      "epoch": 3.1302876480541455,
      "grad_norm": 1.6387944221496582,
      "learning_rate": 5.947095614625894e-07,
      "loss": 0.0408,
      "step": 1850
    },
    {
      "epoch": 3.1472081218274113,
      "grad_norm": 1.1685737371444702,
      "learning_rate": 5.727129557976632e-07,
      "loss": 0.0369,
      "step": 1860
    },
    {
      "epoch": 3.164128595600677,
      "grad_norm": 1.1949836015701294,
      "learning_rate": 5.510781190452299e-07,
      "loss": 0.0413,
      "step": 1870
    },
    {
      "epoch": 3.1810490693739424,
      "grad_norm": 1.5283178091049194,
      "learning_rate": 5.298091122564911e-07,
      "loss": 0.0394,
      "step": 1880
    },
    {
      "epoch": 3.197969543147208,
      "grad_norm": 1.7606452703475952,
      "learning_rate": 5.089099278131079e-07,
      "loss": 0.0539,
      "step": 1890
    },
    {
      "epoch": 3.214890016920474,
      "grad_norm": 1.3933011293411255,
      "learning_rate": 4.883844886778028e-07,
      "loss": 0.0437,
      "step": 1900
    },
    {
      "epoch": 3.2318104906937393,
      "grad_norm": 1.65366530418396,
      "learning_rate": 4.6823664765798136e-07,
      "loss": 0.0472,
      "step": 1910
    },
    {
      "epoch": 3.248730964467005,
      "grad_norm": 1.2451823949813843,
      "learning_rate": 4.484701866825231e-07,
      "loss": 0.0408,
      "step": 1920
    },
    {
      "epoch": 3.265651438240271,
      "grad_norm": 0.8365059494972229,
      "learning_rate": 4.290888160918855e-07,
      "loss": 0.0417,
      "step": 1930
    },
    {
      "epoch": 3.2825719120135366,
      "grad_norm": 1.6369590759277344,
      "learning_rate": 4.1009617394163566e-07,
      "loss": 0.0483,
      "step": 1940
    },
    {
      "epoch": 3.299492385786802,
      "grad_norm": 1.3412925004959106,
      "learning_rate": 3.914958253195561e-07,
      "loss": 0.0422,
      "step": 1950
    },
    {
      "epoch": 3.3164128595600677,
      "grad_norm": 1.2426025867462158,
      "learning_rate": 3.732912616764478e-07,
      "loss": 0.0374,
      "step": 1960
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 1.1421136856079102,
      "learning_rate": 3.554859001707522e-07,
      "loss": 0.0427,
      "step": 1970
    },
    {
      "epoch": 3.350253807106599,
      "grad_norm": 1.1082905530929565,
      "learning_rate": 3.3808308302712007e-07,
      "loss": 0.0485,
      "step": 1980
    },
    {
      "epoch": 3.3671742808798646,
      "grad_norm": 1.8885458707809448,
      "learning_rate": 3.2108607690904595e-07,
      "loss": 0.0465,
      "step": 1990
    },
    {
      "epoch": 3.3840947546531304,
      "grad_norm": 1.5490193367004395,
      "learning_rate": 3.044980723056884e-07,
      "loss": 0.0424,
      "step": 2000
    },
    {
      "epoch": 3.401015228426396,
      "grad_norm": 1.3184294700622559,
      "learning_rate": 2.883221829329857e-07,
      "loss": 0.0469,
      "step": 2010
    },
    {
      "epoch": 3.4179357021996615,
      "grad_norm": 2.039376974105835,
      "learning_rate": 2.7256144514918204e-07,
      "loss": 0.0564,
      "step": 2020
    },
    {
      "epoch": 3.4348561759729273,
      "grad_norm": 1.3883781433105469,
      "learning_rate": 2.5721881738487776e-07,
      "loss": 0.0479,
      "step": 2030
    },
    {
      "epoch": 3.451776649746193,
      "grad_norm": 1.4026315212249756,
      "learning_rate": 2.4229717958770685e-07,
      "loss": 0.0469,
      "step": 2040
    },
    {
      "epoch": 3.4686971235194584,
      "grad_norm": 1.3182621002197266,
      "learning_rate": 2.277993326817421e-07,
      "loss": 0.0733,
      "step": 2050
    },
    {
      "epoch": 3.485617597292724,
      "grad_norm": 1.252341866493225,
      "learning_rate": 2.1372799804173894e-07,
      "loss": 0.0401,
      "step": 2060
    },
    {
      "epoch": 3.50253807106599,
      "grad_norm": 1.346508502960205,
      "learning_rate": 2.0008581698231e-07,
      "loss": 0.0486,
      "step": 2070
    },
    {
      "epoch": 3.5194585448392557,
      "grad_norm": 0.9947476983070374,
      "learning_rate": 1.8687535026212206e-07,
      "loss": 0.0448,
      "step": 2080
    },
    {
      "epoch": 3.536379018612521,
      "grad_norm": 1.5827783346176147,
      "learning_rate": 1.740990776032256e-07,
      "loss": 0.0462,
      "step": 2090
    },
    {
      "epoch": 3.553299492385787,
      "grad_norm": 1.36147940158844,
      "learning_rate": 1.617593972255846e-07,
      "loss": 0.0474,
      "step": 2100
    },
    {
      "epoch": 3.5702199661590526,
      "grad_norm": 1.3989553451538086,
      "learning_rate": 1.4985862539691033e-07,
      "loss": 0.0506,
      "step": 2110
    },
    {
      "epoch": 3.587140439932318,
      "grad_norm": 2.2407867908477783,
      "learning_rate": 1.3839899599787655e-07,
      "loss": 0.0563,
      "step": 2120
    },
    {
      "epoch": 3.6040609137055837,
      "grad_norm": 1.7094990015029907,
      "learning_rate": 1.2738266010280275e-07,
      "loss": 0.0455,
      "step": 2130
    },
    {
      "epoch": 3.6209813874788495,
      "grad_norm": 2.9613707065582275,
      "learning_rate": 1.168116855758747e-07,
      "loss": 0.0462,
      "step": 2140
    },
    {
      "epoch": 3.6379018612521152,
      "grad_norm": 1.730281949043274,
      "learning_rate": 1.066880566829917e-07,
      "loss": 0.0478,
      "step": 2150
    },
    {
      "epoch": 3.6548223350253806,
      "grad_norm": 1.7777799367904663,
      "learning_rate": 9.701367371930059e-08,
      "loss": 0.0514,
      "step": 2160
    },
    {
      "epoch": 3.6717428087986463,
      "grad_norm": 1.5566476583480835,
      "learning_rate": 8.779035265249064e-08,
      "loss": 0.0509,
      "step": 2170
    },
    {
      "epoch": 3.688663282571912,
      "grad_norm": 1.7991079092025757,
      "learning_rate": 7.901982478192572e-08,
      "loss": 0.0432,
      "step": 2180
    },
    {
      "epoch": 3.7055837563451774,
      "grad_norm": 1.1896286010742188,
      "learning_rate": 7.07037364136609e-08,
      "loss": 0.0468,
      "step": 2190
    },
    {
      "epoch": 3.7225042301184432,
      "grad_norm": 1.1462163925170898,
      "learning_rate": 6.2843648551415e-08,
      "loss": 0.0451,
      "step": 2200
    },
    {
      "epoch": 3.739424703891709,
      "grad_norm": 1.4667885303497314,
      "learning_rate": 5.544103660355987e-08,
      "loss": 0.0416,
      "step": 2210
    },
    {
      "epoch": 3.7563451776649748,
      "grad_norm": 1.2911655902862549,
      "learning_rate": 4.849729010616949e-08,
      "loss": 0.0494,
      "step": 2220
    },
    {
      "epoch": 3.77326565143824,
      "grad_norm": 1.1728025674819946,
      "learning_rate": 4.201371246219388e-08,
      "loss": 0.0411,
      "step": 2230
    },
    {
      "epoch": 3.790186125211506,
      "grad_norm": 1.3054083585739136,
      "learning_rate": 3.599152069679812e-08,
      "loss": 0.04,
      "step": 2240
    },
    {
      "epoch": 3.8071065989847717,
      "grad_norm": 1.867836594581604,
      "learning_rate": 3.043184522891679e-08,
      "loss": 0.0567,
      "step": 2250
    },
    {
      "epoch": 3.824027072758037,
      "grad_norm": 1.3285189867019653,
      "learning_rate": 2.5335729659062002e-08,
      "loss": 0.0401,
      "step": 2260
    },
    {
      "epoch": 3.8409475465313028,
      "grad_norm": 1.121762752532959,
      "learning_rate": 2.07041305734329e-08,
      "loss": 0.0408,
      "step": 2270
    },
    {
      "epoch": 3.8578680203045685,
      "grad_norm": 0.9373975992202759,
      "learning_rate": 1.6537917364354838e-08,
      "loss": 0.0379,
      "step": 2280
    },
    {
      "epoch": 3.8747884940778343,
      "grad_norm": 1.4871065616607666,
      "learning_rate": 1.2837872067085477e-08,
      "loss": 0.0454,
      "step": 2290
    },
    {
      "epoch": 3.8917089678511,
      "grad_norm": 1.1789824962615967,
      "learning_rate": 9.604689213022767e-09,
      "loss": 0.0558,
      "step": 2300
    },
    {
      "epoch": 3.9086294416243654,
      "grad_norm": 1.1937594413757324,
      "learning_rate": 6.838975699332007e-09,
      "loss": 0.0447,
      "step": 2310
    },
    {
      "epoch": 3.925549915397631,
      "grad_norm": 1.5232540369033813,
      "learning_rate": 4.541250675028363e-09,
      "loss": 0.0535,
      "step": 2320
    },
    {
      "epoch": 3.9424703891708965,
      "grad_norm": 1.3633273839950562,
      "learning_rate": 2.711945443526209e-09,
      "loss": 0.047,
      "step": 2330
    },
    {
      "epoch": 3.9593908629441623,
      "grad_norm": 1.0834922790527344,
      "learning_rate": 1.351403381680827e-09,
      "loss": 0.0456,
      "step": 2340
    },
    {
      "epoch": 3.976311336717428,
      "grad_norm": 1.0907598733901978,
      "learning_rate": 4.598798753330247e-10,
      "loss": 0.0413,
      "step": 2350
    },
    {
      "epoch": 3.993231810490694,
      "grad_norm": 1.6505200862884521,
      "learning_rate": 3.754227137081623e-11,
      "loss": 0.046,
      "step": 2360
    },
    {
      "epoch": 4.0,
      "step": 2364,
      "total_flos": 1.5782755510340813e+17,
      "train_loss": 0.13063472979615384,
      "train_runtime": 2788.0875,
      "train_samples_per_second": 13.558,
      "train_steps_per_second": 0.848
    }
  ],
  "logging_steps": 10,
  "max_steps": 2364,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 1.5782755510340813e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}