{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02298190175237001,
      "grad_norm": 91.13835144042969,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 2.8685,
      "step": 5
    },
    {
      "epoch": 0.04596380350474002,
      "grad_norm": 46.327125549316406,
      "learning_rate": 1.0227272727272729e-06,
      "loss": 2.4334,
      "step": 10
    },
    {
      "epoch": 0.06894570525711002,
      "grad_norm": 24.945697784423828,
      "learning_rate": 1.590909090909091e-06,
      "loss": 1.5038,
      "step": 15
    },
    {
      "epoch": 0.09192760700948004,
      "grad_norm": 11.960966110229492,
      "learning_rate": 2.1590909090909092e-06,
      "loss": 0.9891,
      "step": 20
    },
    {
      "epoch": 0.11490950876185005,
      "grad_norm": 8.8960542678833,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.6429,
      "step": 25
    },
    {
      "epoch": 0.13789141051422005,
      "grad_norm": 8.532720565795898,
      "learning_rate": 3.2954545454545456e-06,
      "loss": 0.4224,
      "step": 30
    },
    {
      "epoch": 0.16087331226659007,
      "grad_norm": 5.1687092781066895,
      "learning_rate": 3.863636363636364e-06,
      "loss": 0.2885,
      "step": 35
    },
    {
      "epoch": 0.18385521401896007,
      "grad_norm": 5.163649559020996,
      "learning_rate": 4.4318181818181824e-06,
      "loss": 0.2805,
      "step": 40
    },
    {
      "epoch": 0.20683711577133007,
      "grad_norm": 8.365943908691406,
      "learning_rate": 5e-06,
      "loss": 0.2345,
      "step": 45
    },
    {
      "epoch": 0.2298190175237001,
      "grad_norm": 5.84035587310791,
      "learning_rate": 4.9979931287121e-06,
      "loss": 0.2352,
      "step": 50
    },
    {
      "epoch": 0.25280091927607007,
      "grad_norm": 4.9680633544921875,
      "learning_rate": 4.9919757368742895e-06,
      "loss": 0.1625,
      "step": 55
    },
    {
      "epoch": 0.2757828210284401,
      "grad_norm": 3.7817749977111816,
      "learning_rate": 4.981957485391297e-06,
      "loss": 0.143,
      "step": 60
    },
    {
      "epoch": 0.2987647227808101,
      "grad_norm": 3.0659518241882324,
      "learning_rate": 4.967954458536126e-06,
      "loss": 0.128,
      "step": 65
    },
    {
      "epoch": 0.32174662453318015,
      "grad_norm": 3.268207311630249,
      "learning_rate": 4.949989138126809e-06,
      "loss": 0.1505,
      "step": 70
    },
    {
      "epoch": 0.3447285262855501,
      "grad_norm": 4.184539318084717,
      "learning_rate": 4.92809036743191e-06,
      "loss": 0.1428,
      "step": 75
    },
    {
      "epoch": 0.36771042803792015,
      "grad_norm": 2.593376874923706,
      "learning_rate": 4.9022933048627496e-06,
      "loss": 0.1279,
      "step": 80
    },
    {
      "epoch": 0.39069232979029017,
      "grad_norm": 2.0689475536346436,
      "learning_rate": 4.872639367526672e-06,
      "loss": 0.0818,
      "step": 85
    },
    {
      "epoch": 0.41367423154266014,
      "grad_norm": 1.6989630460739136,
      "learning_rate": 4.839176164732009e-06,
      "loss": 0.0947,
      "step": 90
    },
    {
      "epoch": 0.43665613329503017,
      "grad_norm": 2.2530245780944824,
      "learning_rate": 4.8019574215514705e-06,
      "loss": 0.0959,
      "step": 95
    },
    {
      "epoch": 0.4596380350474002,
      "grad_norm": 2.968066930770874,
      "learning_rate": 4.761042892566707e-06,
      "loss": 0.1278,
      "step": 100
    },
    {
      "epoch": 0.48261993679977017,
      "grad_norm": 1.7297723293304443,
      "learning_rate": 4.716498265932501e-06,
      "loss": 0.0699,
      "step": 105
    },
    {
      "epoch": 0.5056018385521401,
      "grad_norm": 2.0555076599121094,
      "learning_rate": 4.668395057914627e-06,
      "loss": 0.1004,
      "step": 110
    },
    {
      "epoch": 0.5285837403045102,
      "grad_norm": 2.8183631896972656,
      "learning_rate": 4.6168104980707105e-06,
      "loss": 0.0795,
      "step": 115
    },
    {
      "epoch": 0.5515656420568802,
      "grad_norm": 2.409313201904297,
      "learning_rate": 4.56182740525839e-06,
      "loss": 0.0964,
      "step": 120
    },
    {
      "epoch": 0.5745475438092502,
      "grad_norm": 2.7853007316589355,
      "learning_rate": 4.5035340546698915e-06,
      "loss": 0.1205,
      "step": 125
    },
    {
      "epoch": 0.5975294455616202,
      "grad_norm": 2.8365917205810547,
      "learning_rate": 4.442024036106476e-06,
      "loss": 0.1459,
      "step": 130
    },
    {
      "epoch": 0.6205113473139903,
      "grad_norm": 1.289715051651001,
      "learning_rate": 4.3773961037202784e-06,
      "loss": 0.0792,
      "step": 135
    },
    {
      "epoch": 0.6434932490663603,
      "grad_norm": 1.1410036087036133,
      "learning_rate": 4.309754017464823e-06,
      "loss": 0.1066,
      "step": 140
    },
    {
      "epoch": 0.6664751508187302,
      "grad_norm": 1.8665039539337158,
      "learning_rate": 4.239206376508716e-06,
      "loss": 0.0849,
      "step": 145
    },
    {
      "epoch": 0.6894570525711002,
      "grad_norm": 2.022641181945801,
      "learning_rate": 4.1658664448800105e-06,
      "loss": 0.0861,
      "step": 150
    },
    {
      "epoch": 0.7124389543234703,
      "grad_norm": 1.1586041450500488,
      "learning_rate": 4.089851969621138e-06,
      "loss": 0.0766,
      "step": 155
    },
    {
      "epoch": 0.7354208560758403,
      "grad_norm": 1.0550047159194946,
      "learning_rate": 4.0112849917463905e-06,
      "loss": 0.0711,
      "step": 160
    },
    {
      "epoch": 0.7584027578282103,
      "grad_norm": 1.9794331789016724,
      "learning_rate": 3.930291650305424e-06,
      "loss": 0.123,
      "step": 165
    },
    {
      "epoch": 0.7813846595805803,
      "grad_norm": 1.611553430557251,
      "learning_rate": 3.847001979867401e-06,
      "loss": 0.1239,
      "step": 170
    },
    {
      "epoch": 0.8043665613329503,
      "grad_norm": 2.660640239715576,
      "learning_rate": 3.761549701750865e-06,
      "loss": 0.0938,
      "step": 175
    },
    {
      "epoch": 0.8273484630853203,
      "grad_norm": 2.221693277359009,
      "learning_rate": 3.6740720093345646e-06,
      "loss": 0.0886,
      "step": 180
    },
    {
      "epoch": 0.8503303648376903,
      "grad_norm": 2.124843120574951,
      "learning_rate": 3.5847093477938955e-06,
      "loss": 0.0995,
      "step": 185
    },
    {
      "epoch": 0.8733122665900603,
      "grad_norm": 1.9118784666061401,
      "learning_rate": 3.4936051886165825e-06,
      "loss": 0.0884,
      "step": 190
    },
    {
      "epoch": 0.8962941683424304,
      "grad_norm": 2.094235897064209,
      "learning_rate": 3.400905799259634e-06,
      "loss": 0.0776,
      "step": 195
    },
    {
      "epoch": 0.9192760700948004,
      "grad_norm": 1.5878815650939941,
      "learning_rate": 3.3067600083173752e-06,
      "loss": 0.0829,
      "step": 200
    },
    {
      "epoch": 0.9422579718471703,
      "grad_norm": 1.0571810007095337,
      "learning_rate": 3.2113189665775812e-06,
      "loss": 0.0766,
      "step": 205
    },
    {
      "epoch": 0.9652398735995403,
      "grad_norm": 2.3485960960388184,
      "learning_rate": 3.1147359043493365e-06,
      "loss": 0.0781,
      "step": 210
    },
    {
      "epoch": 0.9882217753519104,
      "grad_norm": 1.392433762550354,
      "learning_rate": 3.0171658854522274e-06,
      "loss": 0.0705,
      "step": 215
    },
    {
      "epoch": 1.009192760700948,
      "grad_norm": 1.2191481590270996,
      "learning_rate": 2.9187655582618413e-06,
      "loss": 0.0635,
      "step": 220
    },
    {
      "epoch": 1.032174662453318,
      "grad_norm": 1.9181779623031616,
      "learning_rate": 2.8196929042112652e-06,
      "loss": 0.0557,
      "step": 225
    },
    {
      "epoch": 1.055156564205688,
      "grad_norm": 0.8788214325904846,
      "learning_rate": 2.720106984152364e-06,
      "loss": 0.051,
      "step": 230
    },
    {
      "epoch": 1.078138465958058,
      "grad_norm": 2.0608415603637695,
      "learning_rate": 2.620167682984052e-06,
      "loss": 0.0687,
      "step": 235
    },
    {
      "epoch": 1.101120367710428,
      "grad_norm": 1.9271199703216553,
      "learning_rate": 2.5200354529575693e-06,
      "loss": 0.0565,
      "step": 240
    },
    {
      "epoch": 1.124102269462798,
      "grad_norm": 0.821776270866394,
      "learning_rate": 2.4198710560708623e-06,
      "loss": 0.0769,
      "step": 245
    },
    {
      "epoch": 1.147084171215168,
      "grad_norm": 0.9624284505844116,
      "learning_rate": 2.3198353059656763e-06,
      "loss": 0.0568,
      "step": 250
    },
    {
      "epoch": 1.170066072967538,
      "grad_norm": 0.6843684315681458,
      "learning_rate": 2.2200888097417308e-06,
      "loss": 0.0504,
      "step": 255
    },
    {
      "epoch": 1.193047974719908,
      "grad_norm": 0.6667521595954895,
      "learning_rate": 2.120791710102499e-06,
      "loss": 0.0533,
      "step": 260
    },
    {
      "epoch": 1.2160298764722781,
      "grad_norm": 1.7258788347244263,
      "learning_rate": 2.02210342824657e-06,
      "loss": 0.0565,
      "step": 265
    },
    {
      "epoch": 1.2390117782246481,
      "grad_norm": 0.9336230754852295,
      "learning_rate": 1.924182407917391e-06,
      "loss": 0.0554,
      "step": 270
    },
    {
      "epoch": 1.2619936799770182,
      "grad_norm": 0.8951128721237183,
      "learning_rate": 1.827185861022308e-06,
      "loss": 0.0612,
      "step": 275
    },
    {
      "epoch": 1.284975581729388,
      "grad_norm": 0.8194484114646912,
      "learning_rate": 1.7312695152293112e-06,
      "loss": 0.0494,
      "step": 280
    },
    {
      "epoch": 1.3079574834817582,
      "grad_norm": 1.3918014764785767,
      "learning_rate": 1.6365873639467314e-06,
      "loss": 0.0594,
      "step": 285
    },
    {
      "epoch": 1.330939385234128,
      "grad_norm": 0.7676392197608948,
      "learning_rate": 1.5432914190872757e-06,
      "loss": 0.0566,
      "step": 290
    },
    {
      "epoch": 1.353921286986498,
      "grad_norm": 1.5796483755111694,
      "learning_rate": 1.4515314670133582e-06,
      "loss": 0.0532,
      "step": 295
    },
    {
      "epoch": 1.376903188738868,
      "grad_norm": 2.4066555500030518,
      "learning_rate": 1.3614548280555351e-06,
      "loss": 0.0642,
      "step": 300
    },
    {
      "epoch": 1.399885090491238,
      "grad_norm": 1.918955683708191,
      "learning_rate": 1.2732061199901563e-06,
      "loss": 0.0596,
      "step": 305
    },
    {
      "epoch": 1.4228669922436081,
      "grad_norm": 1.1317230463027954,
      "learning_rate": 1.1869270258559477e-06,
      "loss": 0.0463,
      "step": 310
    },
    {
      "epoch": 1.4458488939959782,
      "grad_norm": 1.0399028062820435,
      "learning_rate": 1.1027560664823208e-06,
      "loss": 0.0564,
      "step": 315
    },
    {
      "epoch": 1.4688307957483482,
      "grad_norm": 1.9771984815597534,
      "learning_rate": 1.020828378094592e-06,
      "loss": 0.0589,
      "step": 320
    },
    {
      "epoch": 1.4918126975007182,
      "grad_norm": 1.3911211490631104,
      "learning_rate": 9.412754953531664e-07,
      "loss": 0.0571,
      "step": 325
    },
    {
      "epoch": 1.5147945992530882,
      "grad_norm": 0.7040560841560364,
      "learning_rate": 8.642251401750395e-07,
      "loss": 0.0597,
      "step": 330
    },
    {
      "epoch": 1.537776501005458,
      "grad_norm": 1.6459221839904785,
      "learning_rate": 7.898010166766348e-07,
      "loss": 0.0603,
      "step": 335
    },
    {
      "epoch": 1.5607584027578283,
      "grad_norm": 1.1625308990478516,
      "learning_rate": 7.181226125672142e-07,
      "loss": 0.0517,
      "step": 340
    },
    {
      "epoch": 1.583740304510198,
      "grad_norm": 0.6541905999183655,
      "learning_rate": 6.493050073117115e-07,
      "loss": 0.0518,
      "step": 345
    },
    {
      "epoch": 1.6067222062625683,
      "grad_norm": 2.277409791946411,
      "learning_rate": 5.83458687371e-07,
      "loss": 0.0594,
      "step": 350
    },
    {
      "epoch": 1.6297041080149381,
      "grad_norm": 1.7359994649887085,
      "learning_rate": 5.20689368816201e-07,
      "loss": 0.0531,
      "step": 355
    },
    {
      "epoch": 1.6526860097673084,
      "grad_norm": 0.7097958326339722,
      "learning_rate": 4.610978276018496e-07,
      "loss": 0.0516,
      "step": 360
    },
    {
      "epoch": 1.6756679115196782,
      "grad_norm": 1.2193461656570435,
      "learning_rate": 4.0477973777039854e-07,
      "loss": 0.0491,
      "step": 365
    },
    {
      "epoch": 1.6986498132720482,
      "grad_norm": 0.9415901899337769,
      "learning_rate": 3.518255178478247e-07,
      "loss": 0.0447,
      "step": 370
    },
    {
      "epoch": 1.7216317150244183,
      "grad_norm": 0.737471342086792,
      "learning_rate": 3.02320185676957e-07,
      "loss": 0.0525,
      "step": 375
    },
    {
      "epoch": 1.7446136167767883,
      "grad_norm": 0.9534263610839844,
      "learning_rate": 2.563432219215814e-07,
      "loss": 0.0455,
      "step": 380
    },
    {
      "epoch": 1.7675955185291583,
      "grad_norm": 1.4818460941314697,
      "learning_rate": 2.1396844246046904e-07,
      "loss": 0.0497,
      "step": 385
    },
    {
      "epoch": 1.7905774202815283,
      "grad_norm": 1.0967113971710205,
      "learning_rate": 1.7526387987620602e-07,
      "loss": 0.0453,
      "step": 390
    },
    {
      "epoch": 1.8135593220338984,
      "grad_norm": 1.8939931392669678,
      "learning_rate": 1.4029167422908107e-07,
      "loss": 0.0539,
      "step": 395
    },
    {
      "epoch": 1.8365412237862682,
      "grad_norm": 1.2355403900146484,
      "learning_rate": 1.0910797329140466e-07,
      "loss": 0.0541,
      "step": 400
    },
    {
      "epoch": 1.8595231255386384,
      "grad_norm": 1.0511115789413452,
      "learning_rate": 8.176284240242638e-08,
      "loss": 0.0535,
      "step": 405
    },
    {
      "epoch": 1.8825050272910082,
      "grad_norm": 1.2864972352981567,
      "learning_rate": 5.830018408858246e-08,
      "loss": 0.0454,
      "step": 410
    },
    {
      "epoch": 1.9054869290433785,
      "grad_norm": 1.7390204668045044,
      "learning_rate": 3.8757667578119e-08,
      "loss": 0.0531,
      "step": 415
    },
    {
      "epoch": 1.9284688307957483,
      "grad_norm": 0.6060563921928406,
      "learning_rate": 2.31666683232587e-08,
      "loss": 0.0534,
      "step": 420
    },
    {
      "epoch": 1.9514507325481185,
      "grad_norm": 0.7836726903915405,
      "learning_rate": 1.1552217627004426e-08,
      "loss": 0.046,
      "step": 425
    },
    {
      "epoch": 1.9744326343004883,
      "grad_norm": 0.7183709144592285,
      "learning_rate": 3.932962455458489e-09,
      "loss": 0.0488,
      "step": 430
    },
    {
      "epoch": 1.9974145360528583,
      "grad_norm": 0.9619048833847046,
      "learning_rate": 3.2113550017198734e-10,
      "loss": 0.0515,
      "step": 435
    },
    {
      "epoch": 2.0,
      "step": 436,
      "total_flos": 1.0467157525694669e+18,
      "train_loss": 0.17987915195114568,
      "train_runtime": 9099.3309,
      "train_samples_per_second": 3.06,
      "train_steps_per_second": 0.048
    }
  ],
  "logging_steps": 5,
  "max_steps": 436,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0467157525694669e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}