{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 2.332974433898926,
      "learning_rate": 1.9804000000000002e-05,
      "loss": 9.66635009765625,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6665964126586914,
      "learning_rate": 1.9604e-05,
      "loss": 8.088862915039062,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.068279504776001,
      "learning_rate": 1.9404e-05,
      "loss": 7.215781860351562,
      "step": 150
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.76992666721344,
      "learning_rate": 1.9204e-05,
      "loss": 6.690926513671875,
      "step": 200
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.480324149131775,
      "learning_rate": 1.9004000000000003e-05,
      "loss": 6.430185546875,
      "step": 250
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1323567628860474,
      "learning_rate": 1.8804e-05,
      "loss": 6.076779174804687,
      "step": 300
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.248739242553711,
      "learning_rate": 1.8604000000000003e-05,
      "loss": 6.090475463867188,
      "step": 350
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2340136766433716,
      "learning_rate": 1.8404000000000002e-05,
      "loss": 6.04998046875,
      "step": 400
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1271188259124756,
      "learning_rate": 1.8204e-05,
      "loss": 5.8744744873046875,
      "step": 450
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7444993257522583,
      "learning_rate": 1.8004000000000002e-05,
      "loss": 5.79928466796875,
      "step": 500
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.294396162033081,
      "learning_rate": 1.7804e-05,
      "loss": 5.830994873046875,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1351591348648071,
      "learning_rate": 1.7604e-05,
      "loss": 5.764346313476563,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.6005209684371948,
      "learning_rate": 1.7404e-05,
      "loss": 5.688011474609375,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3918774127960205,
      "learning_rate": 1.7204e-05,
      "loss": 5.67168212890625,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0998594760894775,
      "learning_rate": 1.7004000000000002e-05,
      "loss": 5.563177490234375,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2702234983444214,
      "learning_rate": 1.6804e-05,
      "loss": 5.638896484375,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2025340795516968,
      "learning_rate": 1.6604000000000002e-05,
      "loss": 5.447022705078125,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0838000774383545,
      "learning_rate": 1.6404e-05,
      "loss": 5.563392944335938,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6105167865753174,
      "learning_rate": 1.6204000000000003e-05,
      "loss": 5.441434326171875,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.195156455039978,
      "learning_rate": 1.6004e-05,
      "loss": 5.535233154296875,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3313252925872803,
      "learning_rate": 1.5804000000000003e-05,
      "loss": 5.4950341796875,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2405613660812378,
      "learning_rate": 1.5604000000000002e-05,
      "loss": 5.525492553710937,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3697841167449951,
      "learning_rate": 1.5404e-05,
      "loss": 5.394120483398438,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6689515113830566,
      "learning_rate": 1.5204e-05,
      "loss": 5.42336181640625,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6169382333755493,
      "learning_rate": 1.5004e-05,
      "loss": 5.263718872070313,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3931348323822021,
      "learning_rate": 1.4804000000000001e-05,
      "loss": 5.332710571289063,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3986672163009644,
      "learning_rate": 1.4604000000000001e-05,
      "loss": 5.300538330078125,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2856848239898682,
      "learning_rate": 1.4404e-05,
      "loss": 5.265980834960938,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4729372262954712,
      "learning_rate": 1.4204000000000002e-05,
      "loss": 5.241402587890625,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5896576642990112,
      "learning_rate": 1.4004000000000002e-05,
      "loss": 5.286559448242188,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5870860815048218,
      "learning_rate": 1.3804000000000002e-05,
      "loss": 5.2646923828125,
      "step": 1550
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.9024347066879272,
      "learning_rate": 1.3604000000000002e-05,
      "loss": 5.213834838867188,
      "step": 1600
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.545648217201233,
      "learning_rate": 1.3404e-05,
      "loss": 5.284638671875,
      "step": 1650
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7494972944259644,
      "learning_rate": 1.3204000000000001e-05,
      "loss": 5.29470703125,
      "step": 1700
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.534767746925354,
      "learning_rate": 1.3004000000000001e-05,
      "loss": 5.17380126953125,
      "step": 1750
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9378814697265625,
      "learning_rate": 1.2804000000000001e-05,
      "loss": 5.2462841796875,
      "step": 1800
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7414913177490234,
      "learning_rate": 1.2604e-05,
      "loss": 5.184336547851562,
      "step": 1850
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.312013030052185,
      "learning_rate": 1.2404e-05,
      "loss": 5.2285107421875,
      "step": 1900
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.7188756465911865,
      "learning_rate": 1.2204e-05,
      "loss": 5.183685302734375,
      "step": 1950
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5776464939117432,
      "learning_rate": 1.2004e-05,
      "loss": 5.212086181640625,
      "step": 2000
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.8002684116363525,
      "learning_rate": 1.1803999999999999e-05,
      "loss": 5.187786865234375,
      "step": 2050
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.690958023071289,
      "learning_rate": 1.1604000000000003e-05,
      "loss": 5.070486145019531,
      "step": 2100
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.566103219985962,
      "learning_rate": 1.1404000000000001e-05,
      "loss": 5.091676330566406,
      "step": 2150
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6997359991073608,
      "learning_rate": 1.1204000000000001e-05,
      "loss": 5.151467895507812,
      "step": 2200
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4722249507904053,
      "learning_rate": 1.1004000000000002e-05,
      "loss": 5.095538024902344,
      "step": 2250
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.7524068355560303,
      "learning_rate": 1.0804000000000002e-05,
      "loss": 5.041690673828125,
      "step": 2300
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.56508207321167,
      "learning_rate": 1.0604e-05,
      "loss": 5.047481689453125,
      "step": 2350
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5206327438354492,
      "learning_rate": 1.0404e-05,
      "loss": 5.138922119140625,
      "step": 2400
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7860289812088013,
      "learning_rate": 1.0204000000000001e-05,
      "loss": 5.1789208984375,
      "step": 2450
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.762772798538208,
      "learning_rate": 1.0004000000000001e-05,
      "loss": 4.9995901489257815,
      "step": 2500
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.9011698961257935,
      "learning_rate": 9.804000000000001e-06,
      "loss": 5.030536804199219,
      "step": 2550
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7511317729949951,
      "learning_rate": 9.604000000000002e-06,
      "loss": 5.054608459472656,
      "step": 2600
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.454971194267273,
      "learning_rate": 9.404e-06,
      "loss": 5.037495422363281,
      "step": 2650
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5396337509155273,
      "learning_rate": 9.204e-06,
      "loss": 5.031622314453125,
      "step": 2700
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.467636227607727,
      "learning_rate": 9.004e-06,
      "loss": 5.0833673095703125,
      "step": 2750
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5613443851470947,
      "learning_rate": 8.804e-06,
      "loss": 4.928188171386719,
      "step": 2800
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7963453531265259,
      "learning_rate": 8.604000000000001e-06,
      "loss": 5.066728515625,
      "step": 2850
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9899804592132568,
      "learning_rate": 8.404000000000001e-06,
      "loss": 4.9547052001953125,
      "step": 2900
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.502890944480896,
      "learning_rate": 8.204000000000001e-06,
      "loss": 5.096377868652343,
      "step": 2950
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.554468035697937,
      "learning_rate": 8.004e-06,
      "loss": 5.010852966308594,
      "step": 3000
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.768509030342102,
      "learning_rate": 7.804e-06,
      "loss": 4.954717102050782,
      "step": 3050
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5894496440887451,
      "learning_rate": 7.604e-06,
      "loss": 4.946051025390625,
      "step": 3100
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7610846757888794,
      "learning_rate": 7.404e-06,
      "loss": 4.855561828613281,
      "step": 3150
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7571163177490234,
      "learning_rate": 7.204000000000001e-06,
      "loss": 4.930430297851562,
      "step": 3200
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6853914260864258,
      "learning_rate": 7.004000000000001e-06,
      "loss": 4.855824584960938,
      "step": 3250
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4607149362564087,
      "learning_rate": 6.804e-06,
      "loss": 5.078975524902344,
      "step": 3300
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.9196784496307373,
      "learning_rate": 6.604000000000001e-06,
      "loss": 4.9388720703125,
      "step": 3350
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8639936447143555,
      "learning_rate": 6.404e-06,
      "loss": 4.99400146484375,
      "step": 3400
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7838945388793945,
      "learning_rate": 6.204e-06,
      "loss": 4.887441101074219,
      "step": 3450
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.659825086593628,
      "learning_rate": 6.004000000000001e-06,
      "loss": 4.677823181152344,
      "step": 3500
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4775367975234985,
      "learning_rate": 5.804000000000001e-06,
      "loss": 4.8725653076171875,
      "step": 3550
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.4383823871612549,
      "learning_rate": 5.604000000000001e-06,
      "loss": 4.973861389160156,
      "step": 3600
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7640321254730225,
      "learning_rate": 5.404e-06,
      "loss": 4.926856994628906,
      "step": 3650
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3515348434448242,
      "learning_rate": 5.2040000000000005e-06,
      "loss": 4.904393005371094,
      "step": 3700
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.3852506875991821,
      "learning_rate": 5.004e-06,
      "loss": 4.93823486328125,
      "step": 3750
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5573221445083618,
      "learning_rate": 4.804e-06,
      "loss": 4.901606140136718,
      "step": 3800
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7575520277023315,
      "learning_rate": 4.604e-06,
      "loss": 4.8115087890625,
      "step": 3850
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5037921667099,
      "learning_rate": 4.4040000000000005e-06,
      "loss": 4.910712890625,
      "step": 3900
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.2725980281829834,
      "learning_rate": 4.204e-06,
      "loss": 4.771448364257813,
      "step": 3950
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4463030099868774,
      "learning_rate": 4.004e-06,
      "loss": 4.809368286132813,
      "step": 4000
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.726515293121338,
      "learning_rate": 3.8040000000000003e-06,
      "loss": 4.944158935546875,
      "step": 4050
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6566334962844849,
      "learning_rate": 3.604e-06,
      "loss": 4.9215576171875,
      "step": 4100
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5525240898132324,
      "learning_rate": 3.404e-06,
      "loss": 4.8733984375,
      "step": 4150
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.4844027757644653,
      "learning_rate": 3.2040000000000006e-06,
      "loss": 4.825590515136719,
      "step": 4200
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5394399166107178,
      "learning_rate": 3.0040000000000004e-06,
      "loss": 4.887457580566406,
      "step": 4250
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.463205099105835,
      "learning_rate": 2.804e-06,
      "loss": 4.960470275878906,
      "step": 4300
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.674668550491333,
      "learning_rate": 2.6040000000000004e-06,
      "loss": 4.964275207519531,
      "step": 4350
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.597741723060608,
      "learning_rate": 2.404e-06,
      "loss": 4.844057312011719,
      "step": 4400
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.4939367771148682,
      "learning_rate": 2.2040000000000004e-06,
      "loss": 4.818087463378906,
      "step": 4450
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5718450546264648,
      "learning_rate": 2.004e-06,
      "loss": 4.905419006347656,
      "step": 4500
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.5969394445419312,
      "learning_rate": 1.8040000000000002e-06,
      "loss": 4.818899230957031,
      "step": 4550
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.6721904277801514,
      "learning_rate": 1.604e-06,
      "loss": 4.901571044921875,
      "step": 4600
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.3588262796401978,
      "learning_rate": 1.404e-06,
      "loss": 4.869562377929688,
      "step": 4650
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.6911239624023438,
      "learning_rate": 1.204e-06,
      "loss": 4.782047729492188,
      "step": 4700
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.359279990196228,
      "learning_rate": 1.004e-06,
      "loss": 4.815296936035156,
      "step": 4750
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.4526113271713257,
      "learning_rate": 8.04e-07,
      "loss": 4.827503051757812,
      "step": 4800
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.6256376504898071,
      "learning_rate": 6.040000000000001e-07,
      "loss": 4.810068359375,
      "step": 4850
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.2820252180099487,
      "learning_rate": 4.04e-07,
      "loss": 4.8670956420898435,
      "step": 4900
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3670392036437988,
      "learning_rate": 2.0400000000000003e-07,
      "loss": 4.911158447265625,
      "step": 4950
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4537466764450073,
      "learning_rate": 4e-09,
      "loss": 4.827361145019531,
      "step": 5000
    }
  ],
  "logging_steps": 50,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2.090336256e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}