{
  "best_metric": 0.8163847923278809,
  "best_model_checkpoint": "runs/deepseek_lora_20240423-223943/checkpoint-2500",
  "epoch": 0.0625,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 3.086414337158203,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.7892,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.478134155273438,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.7746,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.574502468109131,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.8222,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.6497371196746826,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.7423,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.116753339767456,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.7622,
      "step": 50
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.179832696914673,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.8183,
      "step": 60
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.9869463443756104,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.822,
      "step": 70
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.093494415283203,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.7966,
      "step": 80
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.230633735656738,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.8113,
      "step": 90
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.374403953552246,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.7582,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.465492248535156,
      "learning_rate": 4.4e-06,
      "loss": 0.7662,
      "step": 110
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.279934883117676,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.8376,
      "step": 120
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.799221992492676,
      "learning_rate": 5.2e-06,
      "loss": 0.7965,
      "step": 130
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.222240686416626,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.8855,
      "step": 140
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.009174346923828,
      "learning_rate": 6e-06,
      "loss": 0.8394,
      "step": 150
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.040350914001465,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.8426,
      "step": 160
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.131030559539795,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.7747,
      "step": 170
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.31986927986145,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.7125,
      "step": 180
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.7623395919799805,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.7854,
      "step": 190
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.848206520080566,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.7756,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.455166816711426,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.7894,
      "step": 210
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.759767532348633,
      "learning_rate": 8.8e-06,
      "loss": 0.7454,
      "step": 220
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.262899875640869,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.8555,
      "step": 230
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.28985071182251,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.6845,
      "step": 240
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.174241542816162,
      "learning_rate": 1e-05,
      "loss": 0.7983,
      "step": 250
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.931599617004395,
      "learning_rate": 1.04e-05,
      "loss": 0.9041,
      "step": 260
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.004627227783203,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.817,
      "step": 270
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.6102757453918457,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.7292,
      "step": 280
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.764902353286743,
      "learning_rate": 1.16e-05,
      "loss": 0.9042,
      "step": 290
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.958317995071411,
      "learning_rate": 1.2e-05,
      "loss": 0.7539,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.098923683166504,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.7955,
      "step": 310
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.129098892211914,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.849,
      "step": 320
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.054119825363159,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.8645,
      "step": 330
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.205028057098389,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.8175,
      "step": 340
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.614790439605713,
      "learning_rate": 1.4e-05,
      "loss": 0.8998,
      "step": 350
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.9891204833984375,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.8108,
      "step": 360
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.152099609375,
      "learning_rate": 1.48e-05,
      "loss": 0.7855,
      "step": 370
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.833850860595703,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.7736,
      "step": 380
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.849621295928955,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.7668,
      "step": 390
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.4542975425720215,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.7781,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.197661876678467,
      "learning_rate": 1.64e-05,
      "loss": 0.8654,
      "step": 410
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.2606770992279053,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.7565,
      "step": 420
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.9680209159851074,
      "learning_rate": 1.72e-05,
      "loss": 0.7886,
      "step": 430
    },
    {
      "epoch": 0.01,
      "grad_norm": 18.749984741210938,
      "learning_rate": 1.76e-05,
      "loss": 0.7305,
      "step": 440
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.822000503540039,
      "learning_rate": 1.8e-05,
      "loss": 0.7833,
      "step": 450
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.999715805053711,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.8483,
      "step": 460
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.193736553192139,
      "learning_rate": 1.88e-05,
      "loss": 0.84,
      "step": 470
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.573124885559082,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.8437,
      "step": 480
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.4221601486206055,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.6836,
      "step": 490
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.0399410724639893,
      "learning_rate": 2e-05,
      "loss": 0.8264,
      "step": 500
    },
    {
      "epoch": 0.01,
      "eval_loss": 0.8175864219665527,
      "eval_runtime": 67.7802,
      "eval_samples_per_second": 14.754,
      "eval_steps_per_second": 14.754,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.971303701400757,
      "learning_rate": 1.9978947368421054e-05,
      "loss": 0.7385,
      "step": 510
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.8043839931488037,
      "learning_rate": 1.9957894736842107e-05,
      "loss": 0.7826,
      "step": 520
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.702253341674805,
      "learning_rate": 1.993684210526316e-05,
      "loss": 0.7971,
      "step": 530
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.176826000213623,
      "learning_rate": 1.9915789473684212e-05,
      "loss": 0.748,
      "step": 540
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.120133876800537,
      "learning_rate": 1.9894736842105265e-05,
      "loss": 0.8461,
      "step": 550
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.286151885986328,
      "learning_rate": 1.9873684210526318e-05,
      "loss": 0.8335,
      "step": 560
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.857172966003418,
      "learning_rate": 1.985263157894737e-05,
      "loss": 0.7231,
      "step": 570
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.327859401702881,
      "learning_rate": 1.9831578947368423e-05,
      "loss": 0.877,
      "step": 580
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.9340362548828125,
      "learning_rate": 1.9810526315789476e-05,
      "loss": 0.8984,
      "step": 590
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.1034326553344727,
      "learning_rate": 1.9789473684210528e-05,
      "loss": 0.7045,
      "step": 600
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.853721857070923,
      "learning_rate": 1.976842105263158e-05,
      "loss": 0.761,
      "step": 610
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.6926398277282715,
      "learning_rate": 1.9747368421052633e-05,
      "loss": 0.9493,
      "step": 620
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.261799335479736,
      "learning_rate": 1.9726315789473686e-05,
      "loss": 0.7719,
      "step": 630
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.864114284515381,
      "learning_rate": 1.970526315789474e-05,
      "loss": 0.9406,
      "step": 640
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.093533515930176,
      "learning_rate": 1.968421052631579e-05,
      "loss": 0.7951,
      "step": 650
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.3724496364593506,
      "learning_rate": 1.9663157894736844e-05,
      "loss": 0.8648,
      "step": 660
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.12341022491455,
      "learning_rate": 1.9642105263157897e-05,
      "loss": 0.7823,
      "step": 670
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.80940842628479,
      "learning_rate": 1.962105263157895e-05,
      "loss": 0.706,
      "step": 680
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.243487358093262,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.8244,
      "step": 690
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.420123100280762,
      "learning_rate": 1.9578947368421055e-05,
      "loss": 0.6753,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 63.8618278503418,
      "learning_rate": 1.9557894736842107e-05,
      "loss": 0.8309,
      "step": 710
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.521258354187012,
      "learning_rate": 1.953684210526316e-05,
      "loss": 0.8101,
      "step": 720
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.9532318115234375,
      "learning_rate": 1.9515789473684213e-05,
      "loss": 0.8533,
      "step": 730
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.792180061340332,
      "learning_rate": 1.9494736842105265e-05,
      "loss": 0.7573,
      "step": 740
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.155513286590576,
      "learning_rate": 1.9473684210526318e-05,
      "loss": 0.8961,
      "step": 750
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.195950508117676,
      "learning_rate": 1.945263157894737e-05,
      "loss": 0.8398,
      "step": 760
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.699478626251221,
      "learning_rate": 1.9431578947368423e-05,
      "loss": 0.8018,
      "step": 770
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.254507541656494,
      "learning_rate": 1.9410526315789476e-05,
      "loss": 0.8408,
      "step": 780
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.351966857910156,
      "learning_rate": 1.9389473684210525e-05,
      "loss": 0.7323,
      "step": 790
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.361276626586914,
      "learning_rate": 1.936842105263158e-05,
      "loss": 0.8401,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.449990272521973,
      "learning_rate": 1.9347368421052634e-05,
      "loss": 0.726,
      "step": 810
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.375738143920898,
      "learning_rate": 1.9326315789473687e-05,
      "loss": 0.8305,
      "step": 820
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.601025342941284,
      "learning_rate": 1.930526315789474e-05,
      "loss": 0.9152,
      "step": 830
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.153268814086914,
      "learning_rate": 1.9284210526315792e-05,
      "loss": 0.8423,
      "step": 840
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.785663604736328,
      "learning_rate": 1.9263157894736845e-05,
      "loss": 0.7733,
      "step": 850
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.162787437438965,
      "learning_rate": 1.9242105263157894e-05,
      "loss": 0.893,
      "step": 860
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.871621608734131,
      "learning_rate": 1.922105263157895e-05,
      "loss": 0.798,
      "step": 870
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.9919800758361816,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.8484,
      "step": 880
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.40109920501709,
      "learning_rate": 1.9178947368421055e-05,
      "loss": 0.9129,
      "step": 890
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.794926643371582,
      "learning_rate": 1.9157894736842108e-05,
      "loss": 0.8687,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.942440986633301,
      "learning_rate": 1.913684210526316e-05,
      "loss": 0.8564,
      "step": 910
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.968307018280029,
      "learning_rate": 1.9115789473684213e-05,
      "loss": 0.8495,
      "step": 920
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.425616264343262,
      "learning_rate": 1.9094736842105262e-05,
      "loss": 0.7242,
      "step": 930
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.819301128387451,
      "learning_rate": 1.907368421052632e-05,
      "loss": 0.8381,
      "step": 940
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.81688117980957,
      "learning_rate": 1.9052631578947368e-05,
      "loss": 0.8817,
      "step": 950
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.102423191070557,
      "learning_rate": 1.9031578947368424e-05,
      "loss": 0.8274,
      "step": 960
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.12994909286499,
      "learning_rate": 1.9010526315789476e-05,
      "loss": 0.7052,
      "step": 970
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.15468692779541,
      "learning_rate": 1.898947368421053e-05,
      "loss": 0.772,
      "step": 980
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.62323796749115,
      "learning_rate": 1.8968421052631582e-05,
      "loss": 0.7764,
      "step": 990
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.546677589416504,
      "learning_rate": 1.894736842105263e-05,
      "loss": 0.8365,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "eval_loss": 0.7952949404716492,
      "eval_runtime": 67.7544,
      "eval_samples_per_second": 14.759,
      "eval_steps_per_second": 14.759,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.28386402130127,
      "learning_rate": 1.8926315789473687e-05,
      "loss": 0.8765,
      "step": 1010
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.3430304527282715,
      "learning_rate": 1.8905263157894736e-05,
      "loss": 0.8763,
      "step": 1020
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.0531206130981445,
      "learning_rate": 1.8884210526315792e-05,
      "loss": 0.7943,
      "step": 1030
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.028320074081421,
      "learning_rate": 1.886315789473684e-05,
      "loss": 0.836,
      "step": 1040
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.3861188888549805,
      "learning_rate": 1.8842105263157898e-05,
      "loss": 0.7336,
      "step": 1050
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.7832908630371094,
      "learning_rate": 1.882105263157895e-05,
      "loss": 0.9283,
      "step": 1060
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.8170342445373535,
      "learning_rate": 1.88e-05,
      "loss": 0.7655,
      "step": 1070
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.15322732925415,
      "learning_rate": 1.8778947368421056e-05,
      "loss": 0.9341,
      "step": 1080
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.066686153411865,
      "learning_rate": 1.8757894736842105e-05,
      "loss": 0.85,
      "step": 1090
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.986961603164673,
      "learning_rate": 1.873684210526316e-05,
      "loss": 0.8943,
      "step": 1100
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.8456902503967285,
      "learning_rate": 1.871578947368421e-05,
      "loss": 0.8279,
      "step": 1110
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.6177377700805664,
      "learning_rate": 1.8694736842105266e-05,
      "loss": 0.8192,
      "step": 1120
    },
    {
      "epoch": 0.03,
      "grad_norm": 14.768010139465332,
      "learning_rate": 1.8673684210526316e-05,
      "loss": 0.8005,
      "step": 1130
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.347342491149902,
      "learning_rate": 1.8652631578947368e-05,
      "loss": 0.8081,
      "step": 1140
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.0560150146484375,
      "learning_rate": 1.8631578947368424e-05,
      "loss": 0.9389,
      "step": 1150
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.3164710998535156,
      "learning_rate": 1.8610526315789473e-05,
      "loss": 0.8501,
      "step": 1160
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.112225532531738,
      "learning_rate": 1.858947368421053e-05,
      "loss": 0.7162,
      "step": 1170
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.200588703155518,
      "learning_rate": 1.856842105263158e-05,
      "loss": 0.7448,
      "step": 1180
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.573482513427734,
      "learning_rate": 1.8547368421052635e-05,
      "loss": 0.8071,
      "step": 1190
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.153548717498779,
      "learning_rate": 1.8526315789473684e-05,
      "loss": 0.7957,
      "step": 1200
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.3308305740356445,
      "learning_rate": 1.8505263157894737e-05,
      "loss": 0.7301,
      "step": 1210
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.269808769226074,
      "learning_rate": 1.8484210526315793e-05,
      "loss": 0.8072,
      "step": 1220
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.588324546813965,
      "learning_rate": 1.8463157894736842e-05,
      "loss": 0.8587,
      "step": 1230
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.593557357788086,
      "learning_rate": 1.8442105263157898e-05,
      "loss": 0.856,
      "step": 1240
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.2591094970703125,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 0.7717,
      "step": 1250
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.052567958831787,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.7823,
      "step": 1260
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.447838306427002,
      "learning_rate": 1.8378947368421053e-05,
      "loss": 0.83,
      "step": 1270
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.029257774353027,
      "learning_rate": 1.8357894736842105e-05,
      "loss": 0.7504,
      "step": 1280
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.053960800170898,
      "learning_rate": 1.8336842105263158e-05,
      "loss": 0.9074,
      "step": 1290
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.2877705097198486,
      "learning_rate": 1.831578947368421e-05,
      "loss": 0.772,
      "step": 1300
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.4482290744781494,
      "learning_rate": 1.8294736842105267e-05,
      "loss": 0.8658,
      "step": 1310
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.684794902801514,
      "learning_rate": 1.8273684210526316e-05,
      "loss": 0.7848,
      "step": 1320
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.553828716278076,
      "learning_rate": 1.8252631578947372e-05,
      "loss": 0.8219,
      "step": 1330
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.5203397274017334,
      "learning_rate": 1.823157894736842e-05,
      "loss": 0.9071,
      "step": 1340
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.961795806884766,
      "learning_rate": 1.8210526315789477e-05,
      "loss": 0.6542,
      "step": 1350
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.663081645965576,
      "learning_rate": 1.8189473684210527e-05,
      "loss": 0.7402,
      "step": 1360
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.785040855407715,
      "learning_rate": 1.816842105263158e-05,
      "loss": 0.7462,
      "step": 1370
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.659074783325195,
      "learning_rate": 1.8147368421052632e-05,
      "loss": 0.6951,
      "step": 1380
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.5885703563690186,
      "learning_rate": 1.8126315789473685e-05,
      "loss": 0.7008,
      "step": 1390
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.1295347213745117,
      "learning_rate": 1.810526315789474e-05,
      "loss": 0.9103,
      "step": 1400
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.4699888229370117,
      "learning_rate": 1.808421052631579e-05,
      "loss": 0.841,
      "step": 1410
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.3273444175720215,
      "learning_rate": 1.8063157894736846e-05,
      "loss": 0.9041,
      "step": 1420
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.149638652801514,
      "learning_rate": 1.8042105263157895e-05,
      "loss": 0.7784,
      "step": 1430
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.4124910831451416,
      "learning_rate": 1.8021052631578948e-05,
      "loss": 0.8208,
      "step": 1440
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.9231085777282715,
      "learning_rate": 1.8e-05,
      "loss": 0.7173,
      "step": 1450
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.008113384246826,
      "learning_rate": 1.7978947368421053e-05,
      "loss": 0.7383,
      "step": 1460
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.1748046875,
      "learning_rate": 1.795789473684211e-05,
      "loss": 0.8399,
      "step": 1470
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.4990293979644775,
      "learning_rate": 1.793684210526316e-05,
      "loss": 0.6721,
      "step": 1480
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.1186299324035645,
      "learning_rate": 1.7915789473684214e-05,
      "loss": 0.782,
      "step": 1490
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.12732458114624,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 0.7211,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "eval_loss": 0.811568021774292,
      "eval_runtime": 67.7961,
      "eval_samples_per_second": 14.75,
      "eval_steps_per_second": 14.75,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.631096124649048,
      "learning_rate": 1.7873684210526316e-05,
      "loss": 0.7557,
      "step": 1510
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.850045204162598,
      "learning_rate": 1.785263157894737e-05,
      "loss": 0.8757,
      "step": 1520
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.1114978790283203,
      "learning_rate": 1.7831578947368422e-05,
      "loss": 0.7613,
      "step": 1530
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.5038743019104,
      "learning_rate": 1.7810526315789474e-05,
      "loss": 0.8049,
      "step": 1540
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.2331156730651855,
      "learning_rate": 1.7789473684210527e-05,
      "loss": 0.8277,
      "step": 1550
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.05696964263916,
      "learning_rate": 1.7768421052631583e-05,
      "loss": 0.7973,
      "step": 1560
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.1331920623779297,
      "learning_rate": 1.7747368421052632e-05,
      "loss": 0.7688,
      "step": 1570
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.984541416168213,
      "learning_rate": 1.7726315789473685e-05,
      "loss": 0.7865,
      "step": 1580
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.149406433105469,
      "learning_rate": 1.7705263157894738e-05,
      "loss": 0.7728,
      "step": 1590
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.092243194580078,
      "learning_rate": 1.768421052631579e-05,
      "loss": 0.935,
      "step": 1600
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.16551399230957,
      "learning_rate": 1.7663157894736843e-05,
      "loss": 0.8286,
      "step": 1610
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.131350517272949,
      "learning_rate": 1.7642105263157896e-05,
      "loss": 0.7864,
      "step": 1620
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.870023727416992,
      "learning_rate": 1.7621052631578948e-05,
      "loss": 0.8645,
      "step": 1630
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.631692886352539,
      "learning_rate": 1.76e-05,
      "loss": 0.8473,
      "step": 1640
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.421032905578613,
      "learning_rate": 1.7578947368421054e-05,
      "loss": 0.7868,
      "step": 1650
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.57529878616333,
      "learning_rate": 1.7557894736842106e-05,
      "loss": 0.7882,
      "step": 1660
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.8785624504089355,
      "learning_rate": 1.753684210526316e-05,
      "loss": 0.7543,
      "step": 1670
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.722006320953369,
      "learning_rate": 1.751578947368421e-05,
      "loss": 0.9626,
      "step": 1680
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.466771364212036,
      "learning_rate": 1.7494736842105264e-05,
      "loss": 0.783,
      "step": 1690
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.072049856185913,
      "learning_rate": 1.7473684210526317e-05,
      "loss": 0.7503,
      "step": 1700
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.768575668334961,
      "learning_rate": 1.745263157894737e-05,
      "loss": 0.8193,
      "step": 1710
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.585022211074829,
      "learning_rate": 1.7431578947368422e-05,
      "loss": 0.8808,
      "step": 1720
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0711567401885986,
      "learning_rate": 1.7410526315789475e-05,
      "loss": 0.8098,
      "step": 1730
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.3020272254943848,
      "learning_rate": 1.7389473684210527e-05,
      "loss": 0.7196,
      "step": 1740
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.645238161087036,
      "learning_rate": 1.736842105263158e-05,
      "loss": 0.8904,
      "step": 1750
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.018638610839844,
      "learning_rate": 1.7347368421052633e-05,
      "loss": 0.7937,
      "step": 1760
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.629096746444702,
      "learning_rate": 1.7326315789473685e-05,
      "loss": 0.9171,
      "step": 1770
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5619189739227295,
      "learning_rate": 1.7305263157894738e-05,
      "loss": 0.9488,
      "step": 1780
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.464752197265625,
      "learning_rate": 1.728421052631579e-05,
      "loss": 0.8459,
      "step": 1790
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.9856364727020264,
      "learning_rate": 1.7263157894736843e-05,
      "loss": 0.8378,
      "step": 1800
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.753553867340088,
      "learning_rate": 1.7242105263157896e-05,
      "loss": 0.8093,
      "step": 1810
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.4593358039855957,
      "learning_rate": 1.722105263157895e-05,
      "loss": 0.7896,
      "step": 1820
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.7163546085357666,
      "learning_rate": 1.72e-05,
      "loss": 0.7188,
      "step": 1830
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.105628728866577,
      "learning_rate": 1.7178947368421054e-05,
      "loss": 0.7643,
      "step": 1840
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.387368679046631,
      "learning_rate": 1.7157894736842107e-05,
      "loss": 0.8465,
      "step": 1850
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.020385265350342,
      "learning_rate": 1.713684210526316e-05,
      "loss": 0.7798,
      "step": 1860
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.560520172119141,
      "learning_rate": 1.7115789473684212e-05,
      "loss": 0.7704,
      "step": 1870
    },
    {
      "epoch": 0.05,
      "grad_norm": 15.739727973937988,
      "learning_rate": 1.7094736842105265e-05,
      "loss": 0.7148,
      "step": 1880
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.79690408706665,
      "learning_rate": 1.7073684210526317e-05,
      "loss": 0.798,
      "step": 1890
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6939146518707275,
      "learning_rate": 1.705263157894737e-05,
      "loss": 0.7641,
      "step": 1900
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.193384170532227,
      "learning_rate": 1.7031578947368423e-05,
      "loss": 0.7866,
      "step": 1910
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.940731525421143,
      "learning_rate": 1.7010526315789475e-05,
      "loss": 0.8261,
      "step": 1920
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.1812446117401123,
      "learning_rate": 1.6989473684210528e-05,
      "loss": 0.7973,
      "step": 1930
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.7413289546966553,
      "learning_rate": 1.696842105263158e-05,
      "loss": 0.7818,
      "step": 1940
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.024014472961426,
      "learning_rate": 1.6947368421052633e-05,
      "loss": 0.7237,
      "step": 1950
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.0871291160583496,
      "learning_rate": 1.6926315789473686e-05,
      "loss": 0.772,
      "step": 1960
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.28814435005188,
      "learning_rate": 1.690526315789474e-05,
      "loss": 0.7067,
      "step": 1970
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8241286277770996,
      "learning_rate": 1.688421052631579e-05,
      "loss": 0.8175,
      "step": 1980
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.5942068099975586,
      "learning_rate": 1.6863157894736844e-05,
      "loss": 0.9265,
      "step": 1990
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.6822662353515625,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 0.8593,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "eval_loss": 0.8064771890640259,
      "eval_runtime": 67.7887,
      "eval_samples_per_second": 14.752,
      "eval_steps_per_second": 14.752,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.032164573669434,
      "learning_rate": 1.682105263157895e-05,
      "loss": 0.8819,
      "step": 2010
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.874982833862305,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.8021,
      "step": 2020
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6172547340393066,
      "learning_rate": 1.6778947368421054e-05,
      "loss": 0.8017,
      "step": 2030
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.659741401672363,
      "learning_rate": 1.6757894736842107e-05,
      "loss": 0.8896,
      "step": 2040
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.189141750335693,
      "learning_rate": 1.673684210526316e-05,
      "loss": 0.7997,
      "step": 2050
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.523468971252441,
      "learning_rate": 1.6715789473684212e-05,
      "loss": 0.8498,
      "step": 2060
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.533658981323242,
      "learning_rate": 1.6694736842105265e-05,
      "loss": 0.8857,
      "step": 2070
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.0041606426239014,
      "learning_rate": 1.6673684210526318e-05,
      "loss": 0.8112,
      "step": 2080
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.055651664733887,
      "learning_rate": 1.665263157894737e-05,
      "loss": 0.7872,
      "step": 2090
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.761922836303711,
      "learning_rate": 1.6631578947368423e-05,
      "loss": 0.7727,
      "step": 2100
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.518223524093628,
      "learning_rate": 1.6610526315789476e-05,
      "loss": 0.7997,
      "step": 2110
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.975761890411377,
      "learning_rate": 1.658947368421053e-05,
      "loss": 0.7457,
      "step": 2120
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.2227561473846436,
      "learning_rate": 1.656842105263158e-05,
      "loss": 0.816,
      "step": 2130
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.705923080444336,
      "learning_rate": 1.6547368421052634e-05,
      "loss": 0.8113,
      "step": 2140
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.655057430267334,
      "learning_rate": 1.6526315789473686e-05,
      "loss": 0.7912,
      "step": 2150
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.0186755657196045,
      "learning_rate": 1.650526315789474e-05,
      "loss": 0.8608,
      "step": 2160
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.232386827468872,
      "learning_rate": 1.648421052631579e-05,
      "loss": 0.8549,
      "step": 2170
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.968620300292969,
      "learning_rate": 1.6463157894736844e-05,
      "loss": 0.868,
      "step": 2180
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.5853216648101807,
      "learning_rate": 1.6442105263157897e-05,
      "loss": 0.8388,
      "step": 2190
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.375610589981079,
      "learning_rate": 1.642105263157895e-05,
      "loss": 0.9111,
      "step": 2200
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9734487533569336,
      "learning_rate": 1.64e-05,
      "loss": 0.7288,
      "step": 2210
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.517192840576172,
      "learning_rate": 1.6378947368421055e-05,
      "loss": 0.698,
      "step": 2220
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.183718204498291,
      "learning_rate": 1.6357894736842108e-05,
      "loss": 0.7759,
      "step": 2230
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.9075675010681152,
      "learning_rate": 1.633684210526316e-05,
      "loss": 0.7829,
      "step": 2240
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.287744998931885,
      "learning_rate": 1.6315789473684213e-05,
      "loss": 0.7057,
      "step": 2250
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.977657318115234,
      "learning_rate": 1.6294736842105265e-05,
      "loss": 0.8346,
      "step": 2260
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.196689128875732,
      "learning_rate": 1.6273684210526318e-05,
      "loss": 0.8508,
      "step": 2270
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.467477798461914,
      "learning_rate": 1.6252631578947367e-05,
      "loss": 0.7179,
      "step": 2280
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.059762954711914,
      "learning_rate": 1.6231578947368423e-05,
      "loss": 0.7549,
      "step": 2290
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.980865955352783,
      "learning_rate": 1.6210526315789473e-05,
      "loss": 0.814,
      "step": 2300
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.675939559936523,
      "learning_rate": 1.618947368421053e-05,
      "loss": 0.8227,
      "step": 2310
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.530073642730713,
      "learning_rate": 1.616842105263158e-05,
      "loss": 0.8517,
      "step": 2320
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.6851344108581543,
      "learning_rate": 1.6147368421052634e-05,
      "loss": 0.7684,
      "step": 2330
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.206923961639404,
      "learning_rate": 1.6126315789473687e-05,
      "loss": 0.8199,
      "step": 2340
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.220828056335449,
      "learning_rate": 1.6105263157894736e-05,
      "loss": 0.8871,
      "step": 2350
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.5062482357025146,
      "learning_rate": 1.6084210526315792e-05,
      "loss": 0.8281,
      "step": 2360
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9830796718597412,
      "learning_rate": 1.606315789473684e-05,
      "loss": 0.8678,
      "step": 2370
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.3255491256713867,
      "learning_rate": 1.6042105263157897e-05,
      "loss": 0.8337,
      "step": 2380
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.259572505950928,
      "learning_rate": 1.6021052631578947e-05,
      "loss": 0.7954,
      "step": 2390
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.6201376914978027,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.818,
      "step": 2400
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.3598544597625732,
      "learning_rate": 1.5978947368421055e-05,
      "loss": 0.7697,
      "step": 2410
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.34808349609375,
      "learning_rate": 1.5957894736842105e-05,
      "loss": 0.6347,
      "step": 2420
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.967682361602783,
      "learning_rate": 1.593684210526316e-05,
      "loss": 0.7178,
      "step": 2430
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.222978591918945,
      "learning_rate": 1.591578947368421e-05,
      "loss": 0.7642,
      "step": 2440
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.9339826107025146,
      "learning_rate": 1.5894736842105266e-05,
      "loss": 0.8197,
      "step": 2450
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.3337771892547607,
      "learning_rate": 1.5873684210526315e-05,
      "loss": 0.9375,
      "step": 2460
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.8479838371276855,
      "learning_rate": 1.585263157894737e-05,
      "loss": 0.9196,
      "step": 2470
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.294541358947754,
      "learning_rate": 1.5831578947368424e-05,
      "loss": 0.7144,
      "step": 2480
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.325323104858398,
      "learning_rate": 1.5810526315789473e-05,
      "loss": 0.7897,
      "step": 2490
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.377369403839111,
      "learning_rate": 1.578947368421053e-05,
      "loss": 0.9008,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "eval_loss": 0.8163847923278809,
      "eval_runtime": 67.7994,
      "eval_samples_per_second": 14.749,
      "eval_steps_per_second": 14.749,
      "step": 2500
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2500,
  "total_flos": 4.025531498496e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}