{
  "best_metric": 0.4798590838909149,
  "best_model_checkpoint": "saves/CADICA_qwenvl_stenosis_detect_scale4/lora/sft/checkpoint-1100",
  "epoch": 0.28328611898017,
  "eval_steps": 50,
  "global_step": 1100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012876641771825909,
      "grad_norm": 13.245840411597928,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 2.8889,
      "num_input_tokens_seen": 52840,
      "step": 5
    },
    {
      "epoch": 0.0025753283543651817,
      "grad_norm": 12.237619501215374,
      "learning_rate": 5.882352941176471e-06,
      "loss": 2.8165,
      "num_input_tokens_seen": 105528,
      "step": 10
    },
    {
      "epoch": 0.0038629925315477724,
      "grad_norm": 16.29688816410412,
      "learning_rate": 8.823529411764707e-06,
      "loss": 2.8363,
      "num_input_tokens_seen": 158768,
      "step": 15
    },
    {
      "epoch": 0.0051506567087303634,
      "grad_norm": 11.576419511120797,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 2.6853,
      "num_input_tokens_seen": 210816,
      "step": 20
    },
    {
      "epoch": 0.006438320885912954,
      "grad_norm": 6.9672256792859,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 2.2992,
      "num_input_tokens_seen": 262936,
      "step": 25
    },
    {
      "epoch": 0.007725985063095545,
      "grad_norm": 3.1837818528204305,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 1.8923,
      "num_input_tokens_seen": 315264,
      "step": 30
    },
    {
      "epoch": 0.009013649240278136,
      "grad_norm": 2.835950303969337,
      "learning_rate": 2.058823529411765e-05,
      "loss": 1.6984,
      "num_input_tokens_seen": 367840,
      "step": 35
    },
    {
      "epoch": 0.010301313417460727,
      "grad_norm": 2.223740001042382,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 1.6434,
      "num_input_tokens_seen": 420112,
      "step": 40
    },
    {
      "epoch": 0.011588977594643318,
      "grad_norm": 1.9880935044313244,
      "learning_rate": 2.647058823529412e-05,
      "loss": 1.4659,
      "num_input_tokens_seen": 472728,
      "step": 45
    },
    {
      "epoch": 0.012876641771825908,
      "grad_norm": 1.7151131700495934,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 1.3506,
      "num_input_tokens_seen": 524648,
      "step": 50
    },
    {
      "epoch": 0.012876641771825908,
      "eval_loss": 1.1727452278137207,
      "eval_runtime": 66.3207,
      "eval_samples_per_second": 1.809,
      "eval_steps_per_second": 0.452,
      "num_input_tokens_seen": 524648,
      "step": 50
    },
    {
      "epoch": 0.014164305949008499,
      "grad_norm": 1.47475981537851,
      "learning_rate": 3.235294117647059e-05,
      "loss": 1.1455,
      "num_input_tokens_seen": 576472,
      "step": 55
    },
    {
      "epoch": 0.01545197012619109,
      "grad_norm": 1.7476693647440722,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.9971,
      "num_input_tokens_seen": 628056,
      "step": 60
    },
    {
      "epoch": 0.01673963430337368,
      "grad_norm": 1.3384365493212875,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.9073,
      "num_input_tokens_seen": 680448,
      "step": 65
    },
    {
      "epoch": 0.018027298480556272,
      "grad_norm": 0.9014358219807773,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.8386,
      "num_input_tokens_seen": 733664,
      "step": 70
    },
    {
      "epoch": 0.01931496265773886,
      "grad_norm": 0.8007820009902022,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.7827,
      "num_input_tokens_seen": 786096,
      "step": 75
    },
    {
      "epoch": 0.020602626834921454,
      "grad_norm": 0.6701003454307716,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.7814,
      "num_input_tokens_seen": 838192,
      "step": 80
    },
    {
      "epoch": 0.021890291012104043,
      "grad_norm": 0.8973165751658843,
      "learning_rate": 5e-05,
      "loss": 0.7297,
      "num_input_tokens_seen": 890112,
      "step": 85
    },
    {
      "epoch": 0.023177955189286635,
      "grad_norm": 0.9060968630490469,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.7894,
      "num_input_tokens_seen": 943472,
      "step": 90
    },
    {
      "epoch": 0.024465619366469224,
      "grad_norm": 0.9520214202472889,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.7758,
      "num_input_tokens_seen": 996872,
      "step": 95
    },
    {
      "epoch": 0.025753283543651816,
      "grad_norm": 0.8226006535044261,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.7577,
      "num_input_tokens_seen": 1049816,
      "step": 100
    },
    {
      "epoch": 0.025753283543651816,
      "eval_loss": 0.7517351508140564,
      "eval_runtime": 38.7829,
      "eval_samples_per_second": 3.094,
      "eval_steps_per_second": 0.774,
      "num_input_tokens_seen": 1049816,
      "step": 100
    },
    {
      "epoch": 0.027040947720834405,
      "grad_norm": 0.7251208491150668,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.7579,
      "num_input_tokens_seen": 1102584,
      "step": 105
    },
    {
      "epoch": 0.028328611898016998,
      "grad_norm": 0.8217419839297042,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.7659,
      "num_input_tokens_seen": 1155512,
      "step": 110
    },
    {
      "epoch": 0.029616276075199587,
      "grad_norm": 0.6768053879888967,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.7469,
      "num_input_tokens_seen": 1207976,
      "step": 115
    },
    {
      "epoch": 0.03090394025238218,
      "grad_norm": 1.9562630849642013,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.7353,
      "num_input_tokens_seen": 1259776,
      "step": 120
    },
    {
      "epoch": 0.03219160442956477,
      "grad_norm": 0.6439041597153087,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.7537,
      "num_input_tokens_seen": 1312760,
      "step": 125
    },
    {
      "epoch": 0.03347926860674736,
      "grad_norm": 0.6124318582166212,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.7669,
      "num_input_tokens_seen": 1365616,
      "step": 130
    },
    {
      "epoch": 0.03476693278392995,
      "grad_norm": 0.7593534002488418,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.722,
      "num_input_tokens_seen": 1417544,
      "step": 135
    },
    {
      "epoch": 0.036054596961112545,
      "grad_norm": 0.7827834651032061,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.7502,
      "num_input_tokens_seen": 1469856,
      "step": 140
    },
    {
      "epoch": 0.037342261138295134,
      "grad_norm": 0.5444126155596626,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.7174,
      "num_input_tokens_seen": 1521496,
      "step": 145
    },
    {
      "epoch": 0.03862992531547772,
      "grad_norm": 0.40878703812837747,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.7018,
      "num_input_tokens_seen": 1573376,
      "step": 150
    },
    {
      "epoch": 0.03862992531547772,
      "eval_loss": 0.7309949994087219,
      "eval_runtime": 38.2005,
      "eval_samples_per_second": 3.141,
      "eval_steps_per_second": 0.785,
      "num_input_tokens_seen": 1573376,
      "step": 150
    },
    {
      "epoch": 0.03991758949266031,
      "grad_norm": 0.5536144453733772,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.738,
      "num_input_tokens_seen": 1626136,
      "step": 155
    },
    {
      "epoch": 0.04120525366984291,
      "grad_norm": 0.5151715191704441,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.7579,
      "num_input_tokens_seen": 1678760,
      "step": 160
    },
    {
      "epoch": 0.042492917847025496,
      "grad_norm": 0.5209077394596254,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.7502,
      "num_input_tokens_seen": 1731240,
      "step": 165
    },
    {
      "epoch": 0.043780582024208085,
      "grad_norm": 0.721213601237688,
      "learning_rate": 0.0001,
      "loss": 0.7448,
      "num_input_tokens_seen": 1783816,
      "step": 170
    },
    {
      "epoch": 0.045068246201390674,
      "grad_norm": 0.48666007914879555,
      "learning_rate": 9.999940874631277e-05,
      "loss": 0.6648,
      "num_input_tokens_seen": 1834592,
      "step": 175
    },
    {
      "epoch": 0.04635591037857327,
      "grad_norm": 0.5136600613696797,
      "learning_rate": 9.999763499923432e-05,
      "loss": 0.7759,
      "num_input_tokens_seen": 1888176,
      "step": 180
    },
    {
      "epoch": 0.04764357455575586,
      "grad_norm": 0.6706281530046975,
      "learning_rate": 9.999467880071402e-05,
      "loss": 0.7167,
      "num_input_tokens_seen": 1940280,
      "step": 185
    },
    {
      "epoch": 0.04893123873293845,
      "grad_norm": 0.5159139445497618,
      "learning_rate": 9.999054022066641e-05,
      "loss": 0.7483,
      "num_input_tokens_seen": 1993096,
      "step": 190
    },
    {
      "epoch": 0.050218902910121044,
      "grad_norm": 0.40251006129746847,
      "learning_rate": 9.998521935696953e-05,
      "loss": 0.7464,
      "num_input_tokens_seen": 2045648,
      "step": 195
    },
    {
      "epoch": 0.05150656708730363,
      "grad_norm": 0.4811730853311867,
      "learning_rate": 9.997871633546257e-05,
      "loss": 0.7594,
      "num_input_tokens_seen": 2099008,
      "step": 200
    },
    {
      "epoch": 0.05150656708730363,
      "eval_loss": 0.7274295687675476,
      "eval_runtime": 38.079,
      "eval_samples_per_second": 3.151,
      "eval_steps_per_second": 0.788,
      "num_input_tokens_seen": 2099008,
      "step": 200
    },
    {
      "epoch": 0.05279423126448622,
      "grad_norm": 0.591934959695668,
      "learning_rate": 9.997103130994296e-05,
      "loss": 0.706,
      "num_input_tokens_seen": 2151680,
      "step": 205
    },
    {
      "epoch": 0.05408189544166881,
      "grad_norm": 0.48253717444489286,
      "learning_rate": 9.996216446216267e-05,
      "loss": 0.7186,
      "num_input_tokens_seen": 2203784,
      "step": 210
    },
    {
      "epoch": 0.055369559618851406,
      "grad_norm": 0.5274315079401322,
      "learning_rate": 9.995211600182397e-05,
      "loss": 0.7009,
      "num_input_tokens_seen": 2255632,
      "step": 215
    },
    {
      "epoch": 0.056657223796033995,
      "grad_norm": 0.32879215224292613,
      "learning_rate": 9.994088616657444e-05,
      "loss": 0.6801,
      "num_input_tokens_seen": 2308096,
      "step": 220
    },
    {
      "epoch": 0.057944887973216584,
      "grad_norm": 0.37171195071448215,
      "learning_rate": 9.992847522200133e-05,
      "loss": 0.7569,
      "num_input_tokens_seen": 2361168,
      "step": 225
    },
    {
      "epoch": 0.05923255215039917,
      "grad_norm": 0.4120941016934064,
      "learning_rate": 9.99148834616253e-05,
      "loss": 0.7402,
      "num_input_tokens_seen": 2413896,
      "step": 230
    },
    {
      "epoch": 0.06052021632758177,
      "grad_norm": 0.5998680948310651,
      "learning_rate": 9.990011120689351e-05,
      "loss": 0.7191,
      "num_input_tokens_seen": 2466136,
      "step": 235
    },
    {
      "epoch": 0.06180788050476436,
      "grad_norm": 0.538488141249078,
      "learning_rate": 9.988415880717194e-05,
      "loss": 0.7274,
      "num_input_tokens_seen": 2518848,
      "step": 240
    },
    {
      "epoch": 0.06309554468194695,
      "grad_norm": 0.4393093124760277,
      "learning_rate": 9.986702663973722e-05,
      "loss": 0.7704,
      "num_input_tokens_seen": 2572384,
      "step": 245
    },
    {
      "epoch": 0.06438320885912954,
      "grad_norm": 0.6116643616510118,
      "learning_rate": 9.98487151097676e-05,
      "loss": 0.7346,
      "num_input_tokens_seen": 2625352,
      "step": 250
    },
    {
      "epoch": 0.06438320885912954,
      "eval_loss": 0.7181503176689148,
      "eval_runtime": 38.0986,
      "eval_samples_per_second": 3.15,
      "eval_steps_per_second": 0.787,
      "num_input_tokens_seen": 2625352,
      "step": 250
    },
    {
      "epoch": 0.06567087303631212,
      "grad_norm": 0.41200227731339506,
      "learning_rate": 9.98292246503335e-05,
      "loss": 0.7408,
      "num_input_tokens_seen": 2678216,
      "step": 255
    },
    {
      "epoch": 0.06695853721349472,
      "grad_norm": 0.44521059732114987,
      "learning_rate": 9.980855572238714e-05,
      "loss": 0.7044,
      "num_input_tokens_seen": 2730664,
      "step": 260
    },
    {
      "epoch": 0.06824620139067732,
      "grad_norm": 0.571896859428363,
      "learning_rate": 9.978670881475172e-05,
      "loss": 0.7334,
      "num_input_tokens_seen": 2783584,
      "step": 265
    },
    {
      "epoch": 0.0695338655678599,
      "grad_norm": 0.3907697039722125,
      "learning_rate": 9.976368444410985e-05,
      "loss": 0.7075,
      "num_input_tokens_seen": 2836152,
      "step": 270
    },
    {
      "epoch": 0.0708215297450425,
      "grad_norm": 0.4507806825752261,
      "learning_rate": 9.973948315499126e-05,
      "loss": 0.7039,
      "num_input_tokens_seen": 2887808,
      "step": 275
    },
    {
      "epoch": 0.07210919392222509,
      "grad_norm": 0.41330504132984697,
      "learning_rate": 9.971410551976002e-05,
      "loss": 0.6953,
      "num_input_tokens_seen": 2939656,
      "step": 280
    },
    {
      "epoch": 0.07339685809940767,
      "grad_norm": 0.4625671909482009,
      "learning_rate": 9.968755213860094e-05,
      "loss": 0.7022,
      "num_input_tokens_seen": 2991632,
      "step": 285
    },
    {
      "epoch": 0.07468452227659027,
      "grad_norm": 0.6553627840267285,
      "learning_rate": 9.96598236395054e-05,
      "loss": 0.6796,
      "num_input_tokens_seen": 3043616,
      "step": 290
    },
    {
      "epoch": 0.07597218645377285,
      "grad_norm": 0.5157886895754477,
      "learning_rate": 9.96309206782565e-05,
      "loss": 0.7346,
      "num_input_tokens_seen": 3096920,
      "step": 295
    },
    {
      "epoch": 0.07725985063095545,
      "grad_norm": 0.5672965149433489,
      "learning_rate": 9.960084393841355e-05,
      "loss": 0.6815,
      "num_input_tokens_seen": 3149032,
      "step": 300
    },
    {
      "epoch": 0.07725985063095545,
      "eval_loss": 0.7073924541473389,
      "eval_runtime": 38.1842,
      "eval_samples_per_second": 3.143,
      "eval_steps_per_second": 0.786,
      "num_input_tokens_seen": 3149032,
      "step": 300
    },
    {
      "epoch": 0.07854751480813804,
      "grad_norm": 0.4479276285203507,
      "learning_rate": 9.956959413129585e-05,
      "loss": 0.7208,
      "num_input_tokens_seen": 3201560,
      "step": 305
    },
    {
      "epoch": 0.07983517898532062,
      "grad_norm": 0.368457437106614,
      "learning_rate": 9.953717199596598e-05,
      "loss": 0.7144,
      "num_input_tokens_seen": 3254632,
      "step": 310
    },
    {
      "epoch": 0.08112284316250322,
      "grad_norm": 0.5531413254856732,
      "learning_rate": 9.95035782992122e-05,
      "loss": 0.6861,
      "num_input_tokens_seen": 3306432,
      "step": 315
    },
    {
      "epoch": 0.08241050733968582,
      "grad_norm": 0.41513991799613037,
      "learning_rate": 9.94688138355304e-05,
      "loss": 0.6836,
      "num_input_tokens_seen": 3358392,
      "step": 320
    },
    {
      "epoch": 0.0836981715168684,
      "grad_norm": 0.47052274706452957,
      "learning_rate": 9.943287942710527e-05,
      "loss": 0.7353,
      "num_input_tokens_seen": 3411424,
      "step": 325
    },
    {
      "epoch": 0.08498583569405099,
      "grad_norm": 0.6322586593511644,
      "learning_rate": 9.939577592379088e-05,
      "loss": 0.6774,
      "num_input_tokens_seen": 3462992,
      "step": 330
    },
    {
      "epoch": 0.08627349987123359,
      "grad_norm": 0.4129597798905344,
      "learning_rate": 9.935750420309055e-05,
      "loss": 0.7331,
      "num_input_tokens_seen": 3516136,
      "step": 335
    },
    {
      "epoch": 0.08756116404841617,
      "grad_norm": 0.4031509882699161,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.6939,
      "num_input_tokens_seen": 3568360,
      "step": 340
    },
    {
      "epoch": 0.08884882822559877,
      "grad_norm": 0.4444358747076587,
      "learning_rate": 9.927745975766654e-05,
      "loss": 0.7158,
      "num_input_tokens_seen": 3620696,
      "step": 345
    },
    {
      "epoch": 0.09013649240278135,
      "grad_norm": 0.5290547365449167,
      "learning_rate": 9.923568892600578e-05,
      "loss": 0.6932,
      "num_input_tokens_seen": 3673152,
      "step": 350
    },
    {
      "epoch": 0.09013649240278135,
      "eval_loss": 0.7044599056243896,
      "eval_runtime": 38.2709,
      "eval_samples_per_second": 3.136,
      "eval_steps_per_second": 0.784,
      "num_input_tokens_seen": 3673152,
      "step": 350
    },
    {
      "epoch": 0.09142415657996394,
      "grad_norm": 0.47530311368359207,
      "learning_rate": 9.91927536630402e-05,
      "loss": 0.6778,
      "num_input_tokens_seen": 3725296,
      "step": 355
    },
    {
      "epoch": 0.09271182075714654,
      "grad_norm": 0.38913022785688944,
      "learning_rate": 9.91486549841951e-05,
      "loss": 0.6857,
      "num_input_tokens_seen": 3777552,
      "step": 360
    },
    {
      "epoch": 0.09399948493432912,
      "grad_norm": 0.4834773141333328,
      "learning_rate": 9.91033939324107e-05,
      "loss": 0.7184,
      "num_input_tokens_seen": 3830200,
      "step": 365
    },
    {
      "epoch": 0.09528714911151172,
      "grad_norm": 0.5862045807150876,
      "learning_rate": 9.905697157811761e-05,
      "loss": 0.7196,
      "num_input_tokens_seen": 3883200,
      "step": 370
    },
    {
      "epoch": 0.09657481328869431,
      "grad_norm": 0.4576971522205563,
      "learning_rate": 9.900938901921131e-05,
      "loss": 0.6914,
      "num_input_tokens_seen": 3935576,
      "step": 375
    },
    {
      "epoch": 0.0978624774658769,
      "grad_norm": 0.49551517524520683,
      "learning_rate": 9.896064738102635e-05,
      "loss": 0.6681,
      "num_input_tokens_seen": 3987624,
      "step": 380
    },
    {
      "epoch": 0.09915014164305949,
      "grad_norm": 0.8198390819787913,
      "learning_rate": 9.891074781630966e-05,
      "loss": 0.6723,
      "num_input_tokens_seen": 4039680,
      "step": 385
    },
    {
      "epoch": 0.10043780582024209,
      "grad_norm": 0.7034626469978683,
      "learning_rate": 9.885969150519331e-05,
      "loss": 0.6498,
      "num_input_tokens_seen": 4091216,
      "step": 390
    },
    {
      "epoch": 0.10172546999742467,
      "grad_norm": 0.8838075623197742,
      "learning_rate": 9.88074796551666e-05,
      "loss": 0.7311,
      "num_input_tokens_seen": 4144264,
      "step": 395
    },
    {
      "epoch": 0.10301313417460727,
      "grad_norm": 0.7342758386202114,
      "learning_rate": 9.875411350104744e-05,
      "loss": 0.7089,
      "num_input_tokens_seen": 4197072,
      "step": 400
    },
    {
      "epoch": 0.10301313417460727,
      "eval_loss": 0.6847750544548035,
      "eval_runtime": 37.9238,
      "eval_samples_per_second": 3.164,
      "eval_steps_per_second": 0.791,
      "num_input_tokens_seen": 4197072,
      "step": 400
    },
    {
      "epoch": 0.10430079835178985,
      "grad_norm": 0.8113533605928532,
      "learning_rate": 9.86995943049533e-05,
      "loss": 0.7021,
      "num_input_tokens_seen": 4249656,
      "step": 405
    },
    {
      "epoch": 0.10558846252897244,
      "grad_norm": 1.1772677082041305,
      "learning_rate": 9.864392335627117e-05,
      "loss": 0.6943,
      "num_input_tokens_seen": 4302944,
      "step": 410
    },
    {
      "epoch": 0.10687612670615504,
      "grad_norm": 1.6493280510697776,
      "learning_rate": 9.858710197162721e-05,
      "loss": 0.7146,
      "num_input_tokens_seen": 4355480,
      "step": 415
    },
    {
      "epoch": 0.10816379088333762,
      "grad_norm": 3.0159798803441715,
      "learning_rate": 9.852913149485556e-05,
      "loss": 0.6312,
      "num_input_tokens_seen": 4407688,
      "step": 420
    },
    {
      "epoch": 0.10945145506052022,
      "grad_norm": 1.7981196843056153,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.6877,
      "num_input_tokens_seen": 4459736,
      "step": 425
    },
    {
      "epoch": 0.11073911923770281,
      "grad_norm": 1.5783278376799834,
      "learning_rate": 9.840974877611422e-05,
      "loss": 0.6975,
      "num_input_tokens_seen": 4512928,
      "step": 430
    },
    {
      "epoch": 0.1120267834148854,
      "grad_norm": 3.306646516615779,
      "learning_rate": 9.834833935756344e-05,
      "loss": 0.651,
      "num_input_tokens_seen": 4565840,
      "step": 435
    },
    {
      "epoch": 0.11331444759206799,
      "grad_norm": 2.3184973874904005,
      "learning_rate": 9.828578649365601e-05,
      "loss": 0.685,
      "num_input_tokens_seen": 4618168,
      "step": 440
    },
    {
      "epoch": 0.11460211176925057,
      "grad_norm": 1.602690016495642,
      "learning_rate": 9.822209166377635e-05,
      "loss": 0.6258,
      "num_input_tokens_seen": 4669784,
      "step": 445
    },
    {
      "epoch": 0.11588977594643317,
      "grad_norm": 2.6770797227308196,
      "learning_rate": 9.815725637431662e-05,
      "loss": 0.6732,
      "num_input_tokens_seen": 4722528,
      "step": 450
    },
    {
      "epoch": 0.11588977594643317,
      "eval_loss": 0.6526497006416321,
      "eval_runtime": 39.085,
      "eval_samples_per_second": 3.07,
      "eval_steps_per_second": 0.768,
      "num_input_tokens_seen": 4722528,
      "step": 450
    },
    {
      "epoch": 0.11717744012361576,
      "grad_norm": 2.1823349329218074,
      "learning_rate": 9.809128215864097e-05,
      "loss": 0.6544,
      "num_input_tokens_seen": 4774400,
      "step": 455
    },
    {
      "epoch": 0.11846510430079835,
      "grad_norm": 1.434521593914191,
      "learning_rate": 9.802417057704931e-05,
      "loss": 0.652,
      "num_input_tokens_seen": 4826704,
      "step": 460
    },
    {
      "epoch": 0.11975276847798094,
      "grad_norm": 2.399754385687283,
      "learning_rate": 9.795592321674045e-05,
      "loss": 0.6582,
      "num_input_tokens_seen": 4880072,
      "step": 465
    },
    {
      "epoch": 0.12104043265516354,
      "grad_norm": 3.9235176077985536,
      "learning_rate": 9.788654169177453e-05,
      "loss": 0.6506,
      "num_input_tokens_seen": 4931968,
      "step": 470
    },
    {
      "epoch": 0.12232809683234612,
      "grad_norm": 3.659330745777227,
      "learning_rate": 9.781602764303487e-05,
      "loss": 0.6551,
      "num_input_tokens_seen": 4983656,
      "step": 475
    },
    {
      "epoch": 0.12361576100952872,
      "grad_norm": 1.9670601503398757,
      "learning_rate": 9.774438273818911e-05,
      "loss": 0.6978,
      "num_input_tokens_seen": 5036528,
      "step": 480
    },
    {
      "epoch": 0.12490342518671131,
      "grad_norm": 1.308580869419328,
      "learning_rate": 9.767160867164979e-05,
      "loss": 0.6407,
      "num_input_tokens_seen": 5088768,
      "step": 485
    },
    {
      "epoch": 0.1261910893638939,
      "grad_norm": 1.7349486072682865,
      "learning_rate": 9.759770716453436e-05,
      "loss": 0.6641,
      "num_input_tokens_seen": 5142080,
      "step": 490
    },
    {
      "epoch": 0.1274787535410765,
      "grad_norm": 2.993327939872198,
      "learning_rate": 9.752267996462434e-05,
      "loss": 0.6588,
      "num_input_tokens_seen": 5194432,
      "step": 495
    },
    {
      "epoch": 0.12876641771825909,
      "grad_norm": 2.6430988002320976,
      "learning_rate": 9.744652884632406e-05,
      "loss": 0.6304,
      "num_input_tokens_seen": 5246640,
      "step": 500
    },
    {
      "epoch": 0.12876641771825909,
      "eval_loss": 0.6272165775299072,
      "eval_runtime": 39.4177,
      "eval_samples_per_second": 3.044,
      "eval_steps_per_second": 0.761,
      "num_input_tokens_seen": 5246640,
      "step": 500
    },
    {
      "epoch": 0.13005408189544168,
      "grad_norm": 2.6047672112920286,
      "learning_rate": 9.736925561061871e-05,
      "loss": 0.5741,
      "num_input_tokens_seen": 5299024,
      "step": 505
    },
    {
      "epoch": 0.13134174607262425,
      "grad_norm": 2.4706517190834063,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.6535,
      "num_input_tokens_seen": 5352664,
      "step": 510
    },
    {
      "epoch": 0.13262941024980685,
      "grad_norm": 2.031672226684599,
      "learning_rate": 9.721135012358156e-05,
      "loss": 0.6081,
      "num_input_tokens_seen": 5406008,
      "step": 515
    },
    {
      "epoch": 0.13391707442698944,
      "grad_norm": 2.773997809426142,
      "learning_rate": 9.713072160673777e-05,
      "loss": 0.6792,
      "num_input_tokens_seen": 5459368,
      "step": 520
    },
    {
      "epoch": 0.13520473860417204,
      "grad_norm": 5.083057729524855,
      "learning_rate": 9.704897844137673e-05,
      "loss": 0.6821,
      "num_input_tokens_seen": 5512960,
      "step": 525
    },
    {
      "epoch": 0.13649240278135463,
      "grad_norm": 3.0440654843385584,
      "learning_rate": 9.696612256073633e-05,
      "loss": 0.5835,
      "num_input_tokens_seen": 5565368,
      "step": 530
    },
    {
      "epoch": 0.1377800669585372,
      "grad_norm": 3.7400231170971323,
      "learning_rate": 9.688215592437039e-05,
      "loss": 0.6129,
      "num_input_tokens_seen": 5618008,
      "step": 535
    },
    {
      "epoch": 0.1390677311357198,
      "grad_norm": 6.340287952379529,
      "learning_rate": 9.679708051810221e-05,
      "loss": 0.5765,
      "num_input_tokens_seen": 5670072,
      "step": 540
    },
    {
      "epoch": 0.1403553953129024,
      "grad_norm": 3.6351560550229207,
      "learning_rate": 9.67108983539777e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 5722936,
      "step": 545
    },
    {
      "epoch": 0.141643059490085,
      "grad_norm": 3.8363425916745117,
      "learning_rate": 9.662361147021779e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 5774880,
      "step": 550
    },
    {
      "epoch": 0.141643059490085,
      "eval_loss": 0.5832681059837341,
      "eval_runtime": 38.2495,
      "eval_samples_per_second": 3.137,
      "eval_steps_per_second": 0.784,
      "num_input_tokens_seen": 5774880,
      "step": 550
    },
    {
      "epoch": 0.14293072366726758,
      "grad_norm": 3.911447203674744,
      "learning_rate": 9.653522193117013e-05,
      "loss": 0.5073,
      "num_input_tokens_seen": 5826608,
      "step": 555
    },
    {
      "epoch": 0.14421838784445018,
      "grad_norm": 3.3501835856945763,
      "learning_rate": 9.644573182726035e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 5879776,
      "step": 560
    },
    {
      "epoch": 0.14550605202163275,
      "grad_norm": 8.75758822201328,
      "learning_rate": 9.63551432749426e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 5932888,
      "step": 565
    },
    {
      "epoch": 0.14679371619881534,
      "grad_norm": 4.351029258458384,
      "learning_rate": 9.626345841664953e-05,
      "loss": 0.6251,
      "num_input_tokens_seen": 5984648,
      "step": 570
    },
    {
      "epoch": 0.14808138037599794,
      "grad_norm": 7.617020699535255,
      "learning_rate": 9.617067942074153e-05,
      "loss": 0.6508,
      "num_input_tokens_seen": 6037000,
      "step": 575
    },
    {
      "epoch": 0.14936904455318054,
      "grad_norm": 7.293430172750479,
      "learning_rate": 9.607680848145558e-05,
      "loss": 0.6686,
      "num_input_tokens_seen": 6090512,
      "step": 580
    },
    {
      "epoch": 0.15065670873036313,
      "grad_norm": 3.3635276124166653,
      "learning_rate": 9.598184781885318e-05,
      "loss": 0.5793,
      "num_input_tokens_seen": 6143320,
      "step": 585
    },
    {
      "epoch": 0.1519443729075457,
      "grad_norm": 2.7589160396339407,
      "learning_rate": 9.588579967876806e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 6195720,
      "step": 590
    },
    {
      "epoch": 0.1532320370847283,
      "grad_norm": 1.582169884399532,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.5644,
      "num_input_tokens_seen": 6247592,
      "step": 595
    },
    {
      "epoch": 0.1545197012619109,
      "grad_norm": 3.891844940061855,
      "learning_rate": 9.569045007802559e-05,
      "loss": 0.5794,
      "num_input_tokens_seen": 6299656,
      "step": 600
    },
    {
      "epoch": 0.1545197012619109,
      "eval_loss": 0.6039358973503113,
      "eval_runtime": 38.3138,
      "eval_samples_per_second": 3.132,
      "eval_steps_per_second": 0.783,
      "num_input_tokens_seen": 6299656,
      "step": 600
    },
    {
      "epoch": 0.1558073654390935,
      "grad_norm": 5.90634634073773,
      "learning_rate": 9.55911532374151e-05,
      "loss": 0.6106,
      "num_input_tokens_seen": 6351680,
      "step": 605
    },
    {
      "epoch": 0.15709502961627608,
      "grad_norm": 3.5429043559071034,
      "learning_rate": 9.549077815930636e-05,
      "loss": 0.5812,
      "num_input_tokens_seen": 6403648,
      "step": 610
    },
    {
      "epoch": 0.15838269379345868,
      "grad_norm": 2.8753548663225144,
      "learning_rate": 9.538932721758474e-05,
      "loss": 0.5992,
      "num_input_tokens_seen": 6456328,
      "step": 615
    },
    {
      "epoch": 0.15967035797064125,
      "grad_norm": 2.4013005755622467,
      "learning_rate": 9.528680281157999e-05,
      "loss": 0.587,
      "num_input_tokens_seen": 6509024,
      "step": 620
    },
    {
      "epoch": 0.16095802214782384,
      "grad_norm": 3.860358696946306,
      "learning_rate": 9.518320736600943e-05,
      "loss": 0.5836,
      "num_input_tokens_seen": 6561336,
      "step": 625
    },
    {
      "epoch": 0.16224568632500644,
      "grad_norm": 3.187917212328382,
      "learning_rate": 9.507854333092063e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 6614024,
      "step": 630
    },
    {
      "epoch": 0.16353335050218903,
      "grad_norm": 3.5342177024321586,
      "learning_rate": 9.497281318163346e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 6666416,
      "step": 635
    },
    {
      "epoch": 0.16482101467937163,
      "grad_norm": 3.90374612709263,
      "learning_rate": 9.486601941868154e-05,
      "loss": 0.572,
      "num_input_tokens_seen": 6718200,
      "step": 640
    },
    {
      "epoch": 0.1661086788565542,
      "grad_norm": 4.4270591027201665,
      "learning_rate": 9.475816456775313e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 6771256,
      "step": 645
    },
    {
      "epoch": 0.1673963430337368,
      "grad_norm": 5.04761388655614,
      "learning_rate": 9.464925117963133e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 6824008,
      "step": 650
    },
    {
      "epoch": 0.1673963430337368,
      "eval_loss": 0.5542036890983582,
      "eval_runtime": 68.9048,
      "eval_samples_per_second": 1.742,
      "eval_steps_per_second": 0.435,
      "num_input_tokens_seen": 6824008,
      "step": 650
    },
    {
      "epoch": 0.1686840072109194,
      "grad_norm": 3.428410481447858,
      "learning_rate": 9.453928183013385e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 6875432,
      "step": 655
    },
    {
      "epoch": 0.16997167138810199,
      "grad_norm": 2.9137495299009846,
      "learning_rate": 9.442825912005202e-05,
      "loss": 0.56,
      "num_input_tokens_seen": 6927768,
      "step": 660
    },
    {
      "epoch": 0.17125933556528458,
      "grad_norm": 4.2956604210715925,
      "learning_rate": 9.431618567508933e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 6980544,
      "step": 665
    },
    {
      "epoch": 0.17254699974246718,
      "grad_norm": 4.3977584083656405,
      "learning_rate": 9.420306414579925e-05,
      "loss": 0.5604,
      "num_input_tokens_seen": 7032584,
      "step": 670
    },
    {
      "epoch": 0.17383466391964975,
      "grad_norm": 4.48381006313936,
      "learning_rate": 9.408889720752266e-05,
      "loss": 0.5763,
      "num_input_tokens_seen": 7085048,
      "step": 675
    },
    {
      "epoch": 0.17512232809683234,
      "grad_norm": 2.189534287393346,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 7137952,
      "step": 680
    },
    {
      "epoch": 0.17640999227401494,
      "grad_norm": 3.34591241093722,
      "learning_rate": 9.385743792892982e-05,
      "loss": 0.5935,
      "num_input_tokens_seen": 7190584,
      "step": 685
    },
    {
      "epoch": 0.17769765645119753,
      "grad_norm": 2.7509902524242507,
      "learning_rate": 9.374015106265968e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 7243440,
      "step": 690
    },
    {
      "epoch": 0.17898532062838013,
      "grad_norm": 2.322454948468365,
      "learning_rate": 9.362182973536569e-05,
      "loss": 0.5351,
      "num_input_tokens_seen": 7295568,
      "step": 695
    },
    {
      "epoch": 0.1802729848055627,
      "grad_norm": 3.4615171229405046,
      "learning_rate": 9.35024767453647e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 7347040,
      "step": 700
    },
    {
      "epoch": 0.1802729848055627,
      "eval_loss": 0.5440100431442261,
      "eval_runtime": 39.1181,
      "eval_samples_per_second": 3.068,
      "eval_steps_per_second": 0.767,
      "num_input_tokens_seen": 7347040,
      "step": 700
    },
    {
      "epoch": 0.1815606489827453,
      "grad_norm": 4.815426816055898,
      "learning_rate": 9.338209491537257e-05,
      "loss": 0.543,
      "num_input_tokens_seen": 7399584,
      "step": 705
    },
    {
      "epoch": 0.1828483131599279,
      "grad_norm": 7.294932559918336,
      "learning_rate": 9.326068709243727e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 7452928,
      "step": 710
    },
    {
      "epoch": 0.18413597733711048,
      "grad_norm": 3.6946433405013495,
      "learning_rate": 9.313825614787177e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 7505112,
      "step": 715
    },
    {
      "epoch": 0.18542364151429308,
      "grad_norm": 4.339671310261357,
      "learning_rate": 9.301480497718593e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 7557608,
      "step": 720
    },
    {
      "epoch": 0.18671130569147568,
      "grad_norm": 11.604530853746237,
      "learning_rate": 9.289033650001817e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 7610048,
      "step": 725
    },
    {
      "epoch": 0.18799896986865824,
      "grad_norm": 5.990020165378009,
      "learning_rate": 9.276485366006634e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 7662056,
      "step": 730
    },
    {
      "epoch": 0.18928663404584084,
      "grad_norm": 4.709895983169237,
      "learning_rate": 9.263835942501807e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 7713656,
      "step": 735
    },
    {
      "epoch": 0.19057429822302344,
      "grad_norm": 4.873824727341975,
      "learning_rate": 9.251085678648072e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 7765992,
      "step": 740
    },
    {
      "epoch": 0.19186196240020603,
      "grad_norm": 3.288968567031419,
      "learning_rate": 9.238234875991046e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 7818448,
      "step": 745
    },
    {
      "epoch": 0.19314962657738863,
      "grad_norm": 4.778741391076671,
      "learning_rate": 9.225283838454111e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 7870520,
      "step": 750
    },
    {
      "epoch": 0.19314962657738863,
      "eval_loss": 0.5273815989494324,
      "eval_runtime": 39.1812,
      "eval_samples_per_second": 3.063,
      "eval_steps_per_second": 0.766,
      "num_input_tokens_seen": 7870520,
      "step": 750
    },
    {
      "epoch": 0.1944372907545712,
      "grad_norm": 4.544356566141105,
      "learning_rate": 9.21223287233121e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 7922736,
      "step": 755
    },
    {
      "epoch": 0.1957249549317538,
      "grad_norm": 7.025876813077666,
      "learning_rate": 9.199082286279622e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 7975304,
      "step": 760
    },
    {
      "epoch": 0.1970126191089364,
      "grad_norm": 4.9360968239249985,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 8027448,
      "step": 765
    },
    {
      "epoch": 0.19830028328611898,
      "grad_norm": 10.528361984915874,
      "learning_rate": 9.172483500792244e-05,
      "loss": 0.5214,
      "num_input_tokens_seen": 8080944,
      "step": 770
    },
    {
      "epoch": 0.19958794746330158,
      "grad_norm": 9.264531258094065,
      "learning_rate": 9.159035930421658e-05,
      "loss": 0.6098,
      "num_input_tokens_seen": 8133392,
      "step": 775
    },
    {
      "epoch": 0.20087561164048418,
      "grad_norm": 1.9709167614209242,
      "learning_rate": 9.145489998237902e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 8185360,
      "step": 780
    },
    {
      "epoch": 0.20216327581766674,
      "grad_norm": 7.5915211434567595,
      "learning_rate": 9.131846024604274e-05,
      "loss": 0.5803,
      "num_input_tokens_seen": 8237672,
      "step": 785
    },
    {
      "epoch": 0.20345093999484934,
      "grad_norm": 3.251682970663388,
      "learning_rate": 9.11810433220276e-05,
      "loss": 0.5365,
      "num_input_tokens_seen": 8289688,
      "step": 790
    },
    {
      "epoch": 0.20473860417203193,
      "grad_norm": 4.341533737034294,
      "learning_rate": 9.104265246026415e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 8341624,
      "step": 795
    },
    {
      "epoch": 0.20602626834921453,
      "grad_norm": 5.463180544339495,
      "learning_rate": 9.090329093371666e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 8393696,
      "step": 800
    },
    {
      "epoch": 0.20602626834921453,
      "eval_loss": 0.5219093561172485,
      "eval_runtime": 39.7455,
      "eval_samples_per_second": 3.019,
      "eval_steps_per_second": 0.755,
      "num_input_tokens_seen": 8393696,
      "step": 800
    },
    {
      "epoch": 0.20731393252639713,
      "grad_norm": 4.254130676908817,
      "learning_rate": 9.076296203830579e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 8446496,
      "step": 805
    },
    {
      "epoch": 0.2086015967035797,
      "grad_norm": 5.6525741285524145,
      "learning_rate": 9.062166909283062e-05,
      "loss": 0.5625,
      "num_input_tokens_seen": 8499544,
      "step": 810
    },
    {
      "epoch": 0.2098892608807623,
      "grad_norm": 3.8041246225911345,
      "learning_rate": 9.047941543889014e-05,
      "loss": 0.5564,
      "num_input_tokens_seen": 8552568,
      "step": 815
    },
    {
      "epoch": 0.2111769250579449,
      "grad_norm": 3.803732280546421,
      "learning_rate": 9.033620444080428e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 8605560,
      "step": 820
    },
    {
      "epoch": 0.21246458923512748,
      "grad_norm": 2.8518948364927925,
      "learning_rate": 9.019203948553422e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 8657704,
      "step": 825
    },
    {
      "epoch": 0.21375225341231008,
      "grad_norm": 3.939376115862177,
      "learning_rate": 9.004692398260244e-05,
      "loss": 0.5235,
      "num_input_tokens_seen": 8711088,
      "step": 830
    },
    {
      "epoch": 0.21503991758949267,
      "grad_norm": 6.635912128499916,
      "learning_rate": 8.9900861364012e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 8763712,
      "step": 835
    },
    {
      "epoch": 0.21632758176667524,
      "grad_norm": 3.7547407090496687,
      "learning_rate": 8.975385508416532e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 8815760,
      "step": 840
    },
    {
      "epoch": 0.21761524594385784,
      "grad_norm": 4.093006904445721,
      "learning_rate": 8.960590861978265e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 8867720,
      "step": 845
    },
    {
      "epoch": 0.21890291012104043,
      "grad_norm": 11.397392997722068,
      "learning_rate": 8.945702546981969e-05,
      "loss": 0.5063,
      "num_input_tokens_seen": 8919608,
      "step": 850
    },
    {
      "epoch": 0.21890291012104043,
      "eval_loss": 0.5525640249252319,
      "eval_runtime": 39.0469,
      "eval_samples_per_second": 3.073,
      "eval_steps_per_second": 0.768,
      "num_input_tokens_seen": 8919608,
      "step": 850
    },
    {
      "epoch": 0.22019057429822303,
      "grad_norm": 4.339535962830116,
      "learning_rate": 8.930720915538487e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 8971048,
      "step": 855
    },
    {
      "epoch": 0.22147823847540563,
      "grad_norm": 6.118436891847819,
      "learning_rate": 8.915646321965614e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 9022936,
      "step": 860
    },
    {
      "epoch": 0.2227659026525882,
      "grad_norm": 3.3997835203618667,
      "learning_rate": 8.900479122779712e-05,
      "loss": 0.5623,
      "num_input_tokens_seen": 9075336,
      "step": 865
    },
    {
      "epoch": 0.2240535668297708,
      "grad_norm": 4.188326935911128,
      "learning_rate": 8.885219676687277e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 9127688,
      "step": 870
    },
    {
      "epoch": 0.22534123100695339,
      "grad_norm": 5.220175192497493,
      "learning_rate": 8.869868344576459e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 9180624,
      "step": 875
    },
    {
      "epoch": 0.22662889518413598,
      "grad_norm": 2.2022914161050577,
      "learning_rate": 8.854425489508532e-05,
      "loss": 0.5062,
      "num_input_tokens_seen": 9233176,
      "step": 880
    },
    {
      "epoch": 0.22791655936131858,
      "grad_norm": 4.62379059067999,
      "learning_rate": 8.838891476709288e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 9286688,
      "step": 885
    },
    {
      "epoch": 0.22920422353850115,
      "grad_norm": 3.639684630492015,
      "learning_rate": 8.823266673560426e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 9339600,
      "step": 890
    },
    {
      "epoch": 0.23049188771568374,
      "grad_norm": 4.131757647310936,
      "learning_rate": 8.807551449590846e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 9391536,
      "step": 895
    },
    {
      "epoch": 0.23177955189286634,
      "grad_norm": 4.771128685196347,
      "learning_rate": 8.791746176467907e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 9443616,
      "step": 900
    },
    {
      "epoch": 0.23177955189286634,
      "eval_loss": 0.49604204297065735,
      "eval_runtime": 39.5289,
      "eval_samples_per_second": 3.036,
      "eval_steps_per_second": 0.759,
      "num_input_tokens_seen": 9443616,
      "step": 900
    },
    {
      "epoch": 0.23306721607004893,
      "grad_norm": 6.849781513397169,
      "learning_rate": 8.775851227988656e-05,
      "loss": 0.5774,
      "num_input_tokens_seen": 9497304,
      "step": 905
    },
    {
      "epoch": 0.23435488024723153,
      "grad_norm": 2.526801567699946,
      "learning_rate": 8.759866980070963e-05,
      "loss": 0.5441,
      "num_input_tokens_seen": 9549416,
      "step": 910
    },
    {
      "epoch": 0.23564254442441412,
      "grad_norm": 3.1008408808291503,
      "learning_rate": 8.743793810744654e-05,
      "loss": 0.4898,
      "num_input_tokens_seen": 9601800,
      "step": 915
    },
    {
      "epoch": 0.2369302086015967,
      "grad_norm": 4.120824184689494,
      "learning_rate": 8.727632100142551e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 9653600,
      "step": 920
    },
    {
      "epoch": 0.2382178727787793,
      "grad_norm": 5.251488809494114,
      "learning_rate": 8.711382230491493e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 9707224,
      "step": 925
    },
    {
      "epoch": 0.23950553695596188,
      "grad_norm": 6.885034741125289,
      "learning_rate": 8.695044586103296e-05,
      "loss": 0.5517,
      "num_input_tokens_seen": 9760096,
      "step": 930
    },
    {
      "epoch": 0.24079320113314448,
      "grad_norm": 4.6246077239626855,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.6064,
      "num_input_tokens_seen": 9812672,
      "step": 935
    },
    {
      "epoch": 0.24208086531032708,
      "grad_norm": 5.621020693846077,
      "learning_rate": 8.662107520733027e-05,
      "loss": 0.5398,
      "num_input_tokens_seen": 9866200,
      "step": 940
    },
    {
      "epoch": 0.24336852948750964,
      "grad_norm": 3.1921985322817092,
      "learning_rate": 8.64550887871741e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 9918160,
      "step": 945
    },
    {
      "epoch": 0.24465619366469224,
      "grad_norm": 2.3689648161336465,
      "learning_rate": 8.628824019879137e-05,
      "loss": 0.5862,
      "num_input_tokens_seen": 9970600,
      "step": 950
    },
    {
      "epoch": 0.24465619366469224,
      "eval_loss": 0.5085262656211853,
      "eval_runtime": 39.0437,
      "eval_samples_per_second": 3.073,
      "eval_steps_per_second": 0.768,
      "num_input_tokens_seen": 9970600,
      "step": 950
    },
    {
      "epoch": 0.24594385784187484,
      "grad_norm": 2.8827978223065363,
      "learning_rate": 8.612053338817581e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 10022248,
      "step": 955
    },
    {
      "epoch": 0.24723152201905743,
      "grad_norm": 6.662877258417003,
      "learning_rate": 8.595197232161824e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 10075280,
      "step": 960
    },
    {
      "epoch": 0.24851918619624003,
      "grad_norm": 8.140970355143077,
      "learning_rate": 8.578256098561275e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 10128392,
      "step": 965
    },
    {
      "epoch": 0.24980685037342262,
      "grad_norm": 3.243184767888501,
      "learning_rate": 8.561230338676239e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 10180720,
      "step": 970
    },
    {
      "epoch": 0.2510945145506052,
      "grad_norm": 6.588760068173114,
      "learning_rate": 8.544120355168451e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 10233256,
      "step": 975
    },
    {
      "epoch": 0.2523821787277878,
      "grad_norm": 2.6240987196110837,
      "learning_rate": 8.526926552691544e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 10284928,
      "step": 980
    },
    {
      "epoch": 0.2536698429049704,
      "grad_norm": 8.242761558538728,
      "learning_rate": 8.509649337881483e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 10338208,
      "step": 985
    },
    {
      "epoch": 0.254957507082153,
      "grad_norm": 8.922137566500533,
      "learning_rate": 8.492289119346943e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 10390224,
      "step": 990
    },
    {
      "epoch": 0.25624517125933555,
      "grad_norm": 4.922275874717211,
      "learning_rate": 8.474846307659658e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 10443080,
      "step": 995
    },
    {
      "epoch": 0.25753283543651817,
      "grad_norm": 6.866585614783304,
      "learning_rate": 8.457321315344694e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 10495592,
      "step": 1000
    },
    {
      "epoch": 0.25753283543651817,
      "eval_loss": 0.5305114388465881,
      "eval_runtime": 38.9297,
      "eval_samples_per_second": 3.082,
      "eval_steps_per_second": 0.771,
      "num_input_tokens_seen": 10495592,
      "step": 1000
    },
    {
      "epoch": 0.25882049961370074,
      "grad_norm": 8.233033578002926,
      "learning_rate": 8.439714556870704e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 10548136,
      "step": 1005
    },
    {
      "epoch": 0.26010816379088336,
      "grad_norm": 5.3701298824478485,
      "learning_rate": 8.422026448640124e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 10600048,
      "step": 1010
    },
    {
      "epoch": 0.26139582796806593,
      "grad_norm": 5.491882026124958,
      "learning_rate": 8.40425740897932e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 10652160,
      "step": 1015
    },
    {
      "epoch": 0.2626834921452485,
      "grad_norm": 5.479941792055548,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.5171,
      "num_input_tokens_seen": 10705208,
      "step": 1020
    },
    {
      "epoch": 0.2639711563224311,
      "grad_norm": 3.489116106033337,
      "learning_rate": 8.368478218232787e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 10758688,
      "step": 1025
    },
    {
      "epoch": 0.2652588204996137,
      "grad_norm": 5.923123692460237,
      "learning_rate": 8.350468913330192e-05,
      "loss": 0.5521,
      "num_input_tokens_seen": 10811408,
      "step": 1030
    },
    {
      "epoch": 0.2665464846767963,
      "grad_norm": 2.7605406738569824,
      "learning_rate": 8.33238036934364e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 10864144,
      "step": 1035
    },
    {
      "epoch": 0.2678341488539789,
      "grad_norm": 5.500647711838314,
      "learning_rate": 8.31421301406986e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 10916952,
      "step": 1040
    },
    {
      "epoch": 0.26912181303116145,
      "grad_norm": 6.823855575342733,
      "learning_rate": 8.29596727716949e-05,
      "loss": 0.5491,
      "num_input_tokens_seen": 10968824,
      "step": 1045
    },
    {
      "epoch": 0.2704094772083441,
      "grad_norm": 5.409054743152559,
      "learning_rate": 8.277643590156894e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 11021656,
      "step": 1050
    },
    {
      "epoch": 0.2704094772083441,
      "eval_loss": 0.5039986371994019,
      "eval_runtime": 40.3009,
      "eval_samples_per_second": 2.978,
      "eval_steps_per_second": 0.744,
      "num_input_tokens_seen": 11021656,
      "step": 1050
    },
    {
      "epoch": 0.27169714138552664,
      "grad_norm": 3.2588151986321994,
      "learning_rate": 8.259242386389973e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 11074336,
      "step": 1055
    },
    {
      "epoch": 0.27298480556270927,
      "grad_norm": 12.995641199019554,
      "learning_rate": 8.240764101059912e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 11126776,
      "step": 1060
    },
    {
      "epoch": 0.27427246973989183,
      "grad_norm": 8.713479932798109,
      "learning_rate": 8.222209171180883e-05,
      "loss": 0.4978,
      "num_input_tokens_seen": 11179680,
      "step": 1065
    },
    {
      "epoch": 0.2755601339170744,
      "grad_norm": 3.6728132957332016,
      "learning_rate": 8.203578035579715e-05,
      "loss": 0.5695,
      "num_input_tokens_seen": 11231616,
      "step": 1070
    },
    {
      "epoch": 0.276847798094257,
      "grad_norm": 9.661110166832387,
      "learning_rate": 8.184871134885513e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 11283720,
      "step": 1075
    },
    {
      "epoch": 0.2781354622714396,
      "grad_norm": 5.4096015474623576,
      "learning_rate": 8.166088911519235e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 11336144,
      "step": 1080
    },
    {
      "epoch": 0.2794231264486222,
      "grad_norm": 5.353663008589148,
      "learning_rate": 8.147231809683236e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 11389128,
      "step": 1085
    },
    {
      "epoch": 0.2807107906258048,
      "grad_norm": 3.863008112890598,
      "learning_rate": 8.128300275350756e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 11441864,
      "step": 1090
    },
    {
      "epoch": 0.2819984548029874,
      "grad_norm": 5.545035623030093,
      "learning_rate": 8.109294756255375e-05,
      "loss": 0.4895,
      "num_input_tokens_seen": 11494880,
      "step": 1095
    },
    {
      "epoch": 0.28328611898017,
      "grad_norm": 5.124762488175073,
      "learning_rate": 8.090215701880419e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 11547008,
      "step": 1100
    },
    {
      "epoch": 0.28328611898017,
      "eval_loss": 0.4798590838909149,
      "eval_runtime": 40.6942,
      "eval_samples_per_second": 2.949,
      "eval_steps_per_second": 0.737,
      "num_input_tokens_seen": 11547008,
      "step": 1100
    }
  ],
  "logging_steps": 5,
  "max_steps": 3400,
  "num_input_tokens_seen": 11547008,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 761815920803840.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}