{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1060723371308483,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011060723371308484,
      "grad_norm": 2677.314208984375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 227.5515,
      "step": 100
    },
    {
      "epoch": 0.022121446742616967,
      "grad_norm": 635.984375,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 94.427,
      "step": 200
    },
    {
      "epoch": 0.03318217011392545,
      "grad_norm": 661.0513305664062,
      "learning_rate": 3e-06,
      "loss": 73.1687,
      "step": 300
    },
    {
      "epoch": 0.044242893485233935,
      "grad_norm": 3017.78662109375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 173.9834,
      "step": 400
    },
    {
      "epoch": 0.05530361685654242,
      "grad_norm": 519.4037475585938,
      "learning_rate": 5e-06,
      "loss": 137.7788,
      "step": 500
    },
    {
      "epoch": 0.0663643402278509,
      "grad_norm": 8911.849609375,
      "learning_rate": 6e-06,
      "loss": 187.6141,
      "step": 600
    },
    {
      "epoch": 0.07742506359915939,
      "grad_norm": 471.71282958984375,
      "learning_rate": 7e-06,
      "loss": 58.5892,
      "step": 700
    },
    {
      "epoch": 0.08848578697046787,
      "grad_norm": 3600.653564453125,
      "learning_rate": 8.000000000000001e-06,
      "loss": 58.4563,
      "step": 800
    },
    {
      "epoch": 0.09954651034177635,
      "grad_norm": 593.4425048828125,
      "learning_rate": 9e-06,
      "loss": 90.9575,
      "step": 900
    },
    {
      "epoch": 0.11060723371308484,
      "grad_norm": 520.9338989257812,
      "learning_rate": 1e-05,
      "loss": 86.6733,
      "step": 1000
    },
    {
      "epoch": 0.12166795708439332,
      "grad_norm": 2047.7371826171875,
      "learning_rate": 9.88888888888889e-06,
      "loss": 173.2546,
      "step": 1100
    },
    {
      "epoch": 0.1327286804557018,
      "grad_norm": 814.9510498046875,
      "learning_rate": 9.777777777777779e-06,
      "loss": 101.3527,
      "step": 1200
    },
    {
      "epoch": 0.14378940382701028,
      "grad_norm": 334.23162841796875,
      "learning_rate": 9.666666666666667e-06,
      "loss": 106.3003,
      "step": 1300
    },
    {
      "epoch": 0.15485012719831878,
      "grad_norm": 714.3182983398438,
      "learning_rate": 9.555555555555556e-06,
      "loss": 47.5703,
      "step": 1400
    },
    {
      "epoch": 0.16591085056962726,
      "grad_norm": 180.67575073242188,
      "learning_rate": 9.444444444444445e-06,
      "loss": 51.4053,
      "step": 1500
    },
    {
      "epoch": 0.17697157394093574,
      "grad_norm": 489.76275634765625,
      "learning_rate": 9.333333333333334e-06,
      "loss": 101.9069,
      "step": 1600
    },
    {
      "epoch": 0.18803229731224422,
      "grad_norm": 1117.69970703125,
      "learning_rate": 9.222222222222224e-06,
      "loss": 101.1306,
      "step": 1700
    },
    {
      "epoch": 0.1990930206835527,
      "grad_norm": 482.6600646972656,
      "learning_rate": 9.111111111111112e-06,
      "loss": 105.3138,
      "step": 1800
    },
    {
      "epoch": 0.21015374405486117,
      "grad_norm": 9838.111328125,
      "learning_rate": 9e-06,
      "loss": 100.1129,
      "step": 1900
    },
    {
      "epoch": 0.22121446742616968,
      "grad_norm": 1945.24951171875,
      "learning_rate": 8.888888888888888e-06,
      "loss": 127.0558,
      "step": 2000
    },
    {
      "epoch": 0.23227519079747816,
      "grad_norm": 474.45806884765625,
      "learning_rate": 8.777777777777778e-06,
      "loss": 76.4998,
      "step": 2100
    },
    {
      "epoch": 0.24333591416878664,
      "grad_norm": 866.0442504882812,
      "learning_rate": 8.666666666666668e-06,
      "loss": 84.0351,
      "step": 2200
    },
    {
      "epoch": 0.2543966375400951,
      "grad_norm": 983.19384765625,
      "learning_rate": 8.555555555555556e-06,
      "loss": 124.6492,
      "step": 2300
    },
    {
      "epoch": 0.2654573609114036,
      "grad_norm": 539.8901977539062,
      "learning_rate": 8.444444444444446e-06,
      "loss": 67.798,
      "step": 2400
    },
    {
      "epoch": 0.2765180842827121,
      "grad_norm": 143.80145263671875,
      "learning_rate": 8.333333333333334e-06,
      "loss": 134.1144,
      "step": 2500
    },
    {
      "epoch": 0.28757880765402055,
      "grad_norm": 609.1976318359375,
      "learning_rate": 8.222222222222222e-06,
      "loss": 89.9258,
      "step": 2600
    },
    {
      "epoch": 0.29863953102532903,
      "grad_norm": 1026.5499267578125,
      "learning_rate": 8.111111111111112e-06,
      "loss": 60.0396,
      "step": 2700
    },
    {
      "epoch": 0.30970025439663756,
      "grad_norm": 760.7676391601562,
      "learning_rate": 8.000000000000001e-06,
      "loss": 130.439,
      "step": 2800
    },
    {
      "epoch": 0.32076097776794604,
      "grad_norm": 1219.1871337890625,
      "learning_rate": 7.88888888888889e-06,
      "loss": 55.5903,
      "step": 2900
    },
    {
      "epoch": 0.3318217011392545,
      "grad_norm": 171.1070556640625,
      "learning_rate": 7.77777777777778e-06,
      "loss": 88.8811,
      "step": 3000
    },
    {
      "epoch": 0.342882424510563,
      "grad_norm": 344.57537841796875,
      "learning_rate": 7.666666666666667e-06,
      "loss": 96.9235,
      "step": 3100
    },
    {
      "epoch": 0.3539431478818715,
      "grad_norm": 908.0119018554688,
      "learning_rate": 7.555555555555556e-06,
      "loss": 100.6853,
      "step": 3200
    },
    {
      "epoch": 0.36500387125317996,
      "grad_norm": 108.43099975585938,
      "learning_rate": 7.444444444444445e-06,
      "loss": 100.0964,
      "step": 3300
    },
    {
      "epoch": 0.37606459462448844,
      "grad_norm": 417.69342041015625,
      "learning_rate": 7.333333333333333e-06,
      "loss": 94.0516,
      "step": 3400
    },
    {
      "epoch": 0.3871253179957969,
      "grad_norm": 438.87359619140625,
      "learning_rate": 7.222222222222223e-06,
      "loss": 91.7939,
      "step": 3500
    },
    {
      "epoch": 0.3981860413671054,
      "grad_norm": 539.2892456054688,
      "learning_rate": 7.111111111111112e-06,
      "loss": 59.7248,
      "step": 3600
    },
    {
      "epoch": 0.40924676473841387,
      "grad_norm": 1536.0924072265625,
      "learning_rate": 7e-06,
      "loss": 166.3196,
      "step": 3700
    },
    {
      "epoch": 0.42030748810972235,
      "grad_norm": 147.51727294921875,
      "learning_rate": 6.88888888888889e-06,
      "loss": 93.8634,
      "step": 3800
    },
    {
      "epoch": 0.4313682114810309,
      "grad_norm": 327.2782287597656,
      "learning_rate": 6.777777777777779e-06,
      "loss": 70.0039,
      "step": 3900
    },
    {
      "epoch": 0.44242893485233936,
      "grad_norm": 454.30511474609375,
      "learning_rate": 6.666666666666667e-06,
      "loss": 109.2739,
      "step": 4000
    },
    {
      "epoch": 0.45348965822364784,
      "grad_norm": 188.76947021484375,
      "learning_rate": 6.555555555555556e-06,
      "loss": 86.5162,
      "step": 4100
    },
    {
      "epoch": 0.4645503815949563,
      "grad_norm": 270.0735778808594,
      "learning_rate": 6.444444444444445e-06,
      "loss": 148.0442,
      "step": 4200
    },
    {
      "epoch": 0.4756111049662648,
      "grad_norm": 866.9409790039062,
      "learning_rate": 6.333333333333333e-06,
      "loss": 112.571,
      "step": 4300
    },
    {
      "epoch": 0.4866718283375733,
      "grad_norm": 304.2856750488281,
      "learning_rate": 6.222222222222223e-06,
      "loss": 71.1989,
      "step": 4400
    },
    {
      "epoch": 0.49773255170888175,
      "grad_norm": 115.608154296875,
      "learning_rate": 6.111111111111112e-06,
      "loss": 61.8786,
      "step": 4500
    },
    {
      "epoch": 0.5087932750801902,
      "grad_norm": 590.6244506835938,
      "learning_rate": 6e-06,
      "loss": 84.0039,
      "step": 4600
    },
    {
      "epoch": 0.5198539984514987,
      "grad_norm": 764.8768920898438,
      "learning_rate": 5.88888888888889e-06,
      "loss": 110.3587,
      "step": 4700
    },
    {
      "epoch": 0.5309147218228072,
      "grad_norm": 970.3759765625,
      "learning_rate": 5.777777777777778e-06,
      "loss": 69.0591,
      "step": 4800
    },
    {
      "epoch": 0.5419754451941157,
      "grad_norm": 835.3163452148438,
      "learning_rate": 5.666666666666667e-06,
      "loss": 61.6823,
      "step": 4900
    },
    {
      "epoch": 0.5530361685654241,
      "grad_norm": 329.2018737792969,
      "learning_rate": 5.555555555555557e-06,
      "loss": 81.483,
      "step": 5000
    },
    {
      "epoch": 0.5640968919367326,
      "grad_norm": 714.6965942382812,
      "learning_rate": 5.444444444444445e-06,
      "loss": 53.5503,
      "step": 5100
    },
    {
      "epoch": 0.5751576153080411,
      "grad_norm": 634.34033203125,
      "learning_rate": 5.333333333333334e-06,
      "loss": 85.4892,
      "step": 5200
    },
    {
      "epoch": 0.5862183386793496,
      "grad_norm": 751.374755859375,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 58.7359,
      "step": 5300
    },
    {
      "epoch": 0.5972790620506581,
      "grad_norm": 320.8452453613281,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 72.5238,
      "step": 5400
    },
    {
      "epoch": 0.6083397854219667,
      "grad_norm": 1012.67333984375,
      "learning_rate": 5e-06,
      "loss": 81.7175,
      "step": 5500
    },
    {
      "epoch": 0.6194005087932751,
      "grad_norm": 545.6947631835938,
      "learning_rate": 4.888888888888889e-06,
      "loss": 73.193,
      "step": 5600
    },
    {
      "epoch": 0.6304612321645836,
      "grad_norm": 441.7222595214844,
      "learning_rate": 4.777777777777778e-06,
      "loss": 73.501,
      "step": 5700
    },
    {
      "epoch": 0.6415219555358921,
      "grad_norm": 1395.952880859375,
      "learning_rate": 4.666666666666667e-06,
      "loss": 152.7282,
      "step": 5800
    },
    {
      "epoch": 0.6525826789072006,
      "grad_norm": 1098.189697265625,
      "learning_rate": 4.555555555555556e-06,
      "loss": 68.6455,
      "step": 5900
    },
    {
      "epoch": 0.663643402278509,
      "grad_norm": 255.7035369873047,
      "learning_rate": 4.444444444444444e-06,
      "loss": 99.0397,
      "step": 6000
    },
    {
      "epoch": 0.6747041256498175,
      "grad_norm": 2269.526123046875,
      "learning_rate": 4.333333333333334e-06,
      "loss": 104.7769,
      "step": 6100
    },
    {
      "epoch": 0.685764849021126,
      "grad_norm": 753.7120361328125,
      "learning_rate": 4.222222222222223e-06,
      "loss": 40.9474,
      "step": 6200
    },
    {
      "epoch": 0.6968255723924345,
      "grad_norm": 809.7864990234375,
      "learning_rate": 4.111111111111111e-06,
      "loss": 69.8394,
      "step": 6300
    },
    {
      "epoch": 0.707886295763743,
      "grad_norm": 427.16754150390625,
      "learning_rate": 4.000000000000001e-06,
      "loss": 125.2336,
      "step": 6400
    },
    {
      "epoch": 0.7189470191350514,
      "grad_norm": 805.5377807617188,
      "learning_rate": 3.88888888888889e-06,
      "loss": 91.8197,
      "step": 6500
    },
    {
      "epoch": 0.7300077425063599,
      "grad_norm": 160.71058654785156,
      "learning_rate": 3.777777777777778e-06,
      "loss": 77.6574,
      "step": 6600
    },
    {
      "epoch": 0.7410684658776684,
      "grad_norm": 965.3944702148438,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 77.2402,
      "step": 6700
    },
    {
      "epoch": 0.7521291892489769,
      "grad_norm": 903.8233642578125,
      "learning_rate": 3.555555555555556e-06,
      "loss": 46.0224,
      "step": 6800
    },
    {
      "epoch": 0.7631899126202853,
      "grad_norm": 667.3284301757812,
      "learning_rate": 3.444444444444445e-06,
      "loss": 77.027,
      "step": 6900
    },
    {
      "epoch": 0.7742506359915938,
      "grad_norm": 1243.2249755859375,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 67.4235,
      "step": 7000
    },
    {
      "epoch": 0.7853113593629023,
      "grad_norm": 366.35675048828125,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 72.4485,
      "step": 7100
    },
    {
      "epoch": 0.7963720827342108,
      "grad_norm": 708.7745361328125,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 97.0563,
      "step": 7200
    },
    {
      "epoch": 0.8074328061055193,
      "grad_norm": 218.93495178222656,
      "learning_rate": 3e-06,
      "loss": 50.6086,
      "step": 7300
    },
    {
      "epoch": 0.8184935294768277,
      "grad_norm": 478.40936279296875,
      "learning_rate": 2.888888888888889e-06,
      "loss": 70.7387,
      "step": 7400
    },
    {
      "epoch": 0.8295542528481362,
      "grad_norm": 2558.1591796875,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 67.7935,
      "step": 7500
    },
    {
      "epoch": 0.8406149762194447,
      "grad_norm": 1078.4237060546875,
      "learning_rate": 2.666666666666667e-06,
      "loss": 50.7353,
      "step": 7600
    },
    {
      "epoch": 0.8516756995907533,
      "grad_norm": 1874.4305419921875,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 70.5969,
      "step": 7700
    },
    {
      "epoch": 0.8627364229620618,
      "grad_norm": 21617.517578125,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 128.802,
      "step": 7800
    },
    {
      "epoch": 0.8737971463333702,
      "grad_norm": 363.6821594238281,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 75.1114,
      "step": 7900
    },
    {
      "epoch": 0.8848578697046787,
      "grad_norm": 366.4112548828125,
      "learning_rate": 2.222222222222222e-06,
      "loss": 97.5475,
      "step": 8000
    },
    {
      "epoch": 0.8959185930759872,
      "grad_norm": 803.3827514648438,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 46.939,
      "step": 8100
    },
    {
      "epoch": 0.9069793164472957,
      "grad_norm": 818.63330078125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 64.0652,
      "step": 8200
    },
    {
      "epoch": 0.9180400398186042,
      "grad_norm": 441.73773193359375,
      "learning_rate": 1.888888888888889e-06,
      "loss": 79.8003,
      "step": 8300
    },
    {
      "epoch": 0.9291007631899126,
      "grad_norm": 981.2015991210938,
      "learning_rate": 1.777777777777778e-06,
      "loss": 134.9718,
      "step": 8400
    },
    {
      "epoch": 0.9401614865612211,
      "grad_norm": 688.312744140625,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 61.6925,
      "step": 8500
    },
    {
      "epoch": 0.9512222099325296,
      "grad_norm": 924.4016723632812,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 62.4553,
      "step": 8600
    },
    {
      "epoch": 0.9622829333038381,
      "grad_norm": 121.21617126464844,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 156.6417,
      "step": 8700
    },
    {
      "epoch": 0.9733436566751466,
      "grad_norm": 238.2010498046875,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 58.9229,
      "step": 8800
    },
    {
      "epoch": 0.984404380046455,
      "grad_norm": 137.2640380859375,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 43.9339,
      "step": 8900
    },
    {
      "epoch": 0.9954651034177635,
      "grad_norm": 2417.302001953125,
      "learning_rate": 1.111111111111111e-06,
      "loss": 47.7252,
      "step": 9000
    },
    {
      "epoch": 1.0,
      "eval_loss": 68.28670501708984,
      "eval_runtime": 62.8173,
      "eval_samples_per_second": 31.998,
      "eval_steps_per_second": 15.999,
      "step": 9041
    },
    {
      "epoch": 1.006525826789072,
      "grad_norm": 470.1504211425781,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 106.8215,
      "step": 9100
    },
    {
      "epoch": 1.0175865501603805,
      "grad_norm": 309.5218811035156,
      "learning_rate": 8.88888888888889e-07,
      "loss": 56.8555,
      "step": 9200
    },
    {
      "epoch": 1.028647273531689,
      "grad_norm": 785.29443359375,
      "learning_rate": 7.777777777777779e-07,
      "loss": 56.0117,
      "step": 9300
    },
    {
      "epoch": 1.0397079969029974,
      "grad_norm": 1061.9769287109375,
      "learning_rate": 6.666666666666667e-07,
      "loss": 95.929,
      "step": 9400
    },
    {
      "epoch": 1.050768720274306,
      "grad_norm": 526.0447387695312,
      "learning_rate": 5.555555555555555e-07,
      "loss": 46.0885,
      "step": 9500
    },
    {
      "epoch": 1.0618294436456144,
      "grad_norm": 1534.6865234375,
      "learning_rate": 4.444444444444445e-07,
      "loss": 45.3248,
      "step": 9600
    },
    {
      "epoch": 1.0728901670169229,
      "grad_norm": 539.015380859375,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 66.5751,
      "step": 9700
    },
    {
      "epoch": 1.0839508903882313,
      "grad_norm": 2555.524658203125,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 81.8048,
      "step": 9800
    },
    {
      "epoch": 1.0950116137595398,
      "grad_norm": 816.99072265625,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 58.4326,
      "step": 9900
    },
    {
      "epoch": 1.1060723371308483,
      "grad_norm": 1056.107666015625,
      "learning_rate": 0.0,
      "loss": 110.9712,
      "step": 10000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}