{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 1,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 330.0,
      "learning_rate": 2.5e-05,
      "loss": 2.5909,
      "step": 1
    },
    {
      "epoch": 0.1111111111111111,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 3.1164491176605225,
      "eval_runtime": 0.5321,
      "eval_samples_per_second": 135.313,
      "eval_steps_per_second": 9.397,
      "step": 1
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 336.0,
      "learning_rate": 5e-05,
      "loss": 3.2519,
      "step": 2
    },
    {
      "epoch": 0.2222222222222222,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 2.3881568908691406,
      "eval_runtime": 0.6264,
      "eval_samples_per_second": 114.934,
      "eval_steps_per_second": 7.982,
      "step": 2
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 316.0,
      "learning_rate": 4.943181818181818e-05,
      "loss": 2.2592,
      "step": 3
    },
    {
      "epoch": 0.3333333333333333,
      "eval_accuracy": 0.4583333333333333,
      "eval_loss": 1.807464838027954,
      "eval_runtime": 0.6245,
      "eval_samples_per_second": 115.289,
      "eval_steps_per_second": 8.006,
      "step": 3
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 201.0,
      "learning_rate": 4.886363636363637e-05,
      "loss": 1.745,
      "step": 4
    },
    {
      "epoch": 0.4444444444444444,
      "eval_accuracy": 0.4444444444444444,
      "eval_loss": 2.4170877933502197,
      "eval_runtime": 0.6253,
      "eval_samples_per_second": 115.149,
      "eval_steps_per_second": 7.996,
      "step": 4
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 237.0,
      "learning_rate": 4.829545454545455e-05,
      "loss": 2.4192,
      "step": 5
    },
    {
      "epoch": 0.5555555555555556,
      "eval_accuracy": 0.4861111111111111,
      "eval_loss": 1.8398630619049072,
      "eval_runtime": 0.6243,
      "eval_samples_per_second": 115.327,
      "eval_steps_per_second": 8.009,
      "step": 5
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 290.0,
      "learning_rate": 4.772727272727273e-05,
      "loss": 2.2182,
      "step": 6
    },
    {
      "epoch": 0.6666666666666666,
      "eval_accuracy": 0.5694444444444444,
      "eval_loss": 1.299572467803955,
      "eval_runtime": 0.5773,
      "eval_samples_per_second": 124.719,
      "eval_steps_per_second": 8.661,
      "step": 6
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 247.0,
      "learning_rate": 4.715909090909091e-05,
      "loss": 1.5298,
      "step": 7
    },
    {
      "epoch": 0.7777777777777778,
      "eval_accuracy": 0.5138888888888888,
      "eval_loss": 1.4382593631744385,
      "eval_runtime": 0.6238,
      "eval_samples_per_second": 115.42,
      "eval_steps_per_second": 8.015,
      "step": 7
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 61.75,
      "learning_rate": 4.659090909090909e-05,
      "loss": 0.8786,
      "step": 8
    },
    {
      "epoch": 0.8888888888888888,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 1.6289072036743164,
      "eval_runtime": 0.6268,
      "eval_samples_per_second": 114.863,
      "eval_steps_per_second": 7.977,
      "step": 8
    },
    {
      "epoch": 1.0,
      "grad_norm": 156.0,
      "learning_rate": 4.602272727272727e-05,
      "loss": 0.8227,
      "step": 9
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5416666666666666,
      "eval_loss": 1.5709539651870728,
      "eval_runtime": 0.6278,
      "eval_samples_per_second": 114.692,
      "eval_steps_per_second": 7.965,
      "step": 9
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 34.0,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.4666,
      "step": 10
    },
    {
      "epoch": 1.1111111111111112,
      "eval_accuracy": 0.5277777777777778,
      "eval_loss": 1.4421333074569702,
      "eval_runtime": 0.5742,
      "eval_samples_per_second": 125.392,
      "eval_steps_per_second": 8.708,
      "step": 10
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 44.5,
      "learning_rate": 4.488636363636364e-05,
      "loss": 0.5033,
      "step": 11
    },
    {
      "epoch": 1.2222222222222223,
      "eval_accuracy": 0.4861111111111111,
      "eval_loss": 1.321220874786377,
      "eval_runtime": 0.6247,
      "eval_samples_per_second": 115.26,
      "eval_steps_per_second": 8.004,
      "step": 11
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 17.75,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.3354,
      "step": 12
    },
    {
      "epoch": 1.3333333333333333,
      "eval_accuracy": 0.4305555555555556,
      "eval_loss": 1.329500675201416,
      "eval_runtime": 0.6271,
      "eval_samples_per_second": 114.814,
      "eval_steps_per_second": 7.973,
      "step": 12
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 17.5,
      "learning_rate": 4.375e-05,
      "loss": 0.2448,
      "step": 13
    },
    {
      "epoch": 1.4444444444444444,
      "eval_accuracy": 0.3888888888888889,
      "eval_loss": 1.3822489976882935,
      "eval_runtime": 0.6265,
      "eval_samples_per_second": 114.93,
      "eval_steps_per_second": 7.981,
      "step": 13
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 29.0,
      "learning_rate": 4.318181818181819e-05,
      "loss": 0.7798,
      "step": 14
    },
    {
      "epoch": 1.5555555555555556,
      "eval_accuracy": 0.4305555555555556,
      "eval_loss": 1.4287919998168945,
      "eval_runtime": 0.6272,
      "eval_samples_per_second": 114.8,
      "eval_steps_per_second": 7.972,
      "step": 14
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 66.0,
      "learning_rate": 4.261363636363637e-05,
      "loss": 0.4917,
      "step": 15
    },
    {
      "epoch": 1.6666666666666665,
      "eval_accuracy": 0.4722222222222222,
      "eval_loss": 1.5045852661132812,
      "eval_runtime": 0.6244,
      "eval_samples_per_second": 115.305,
      "eval_steps_per_second": 8.007,
      "step": 15
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 36.5,
      "learning_rate": 4.204545454545455e-05,
      "loss": 0.2832,
      "step": 16
    },
    {
      "epoch": 1.7777777777777777,
      "eval_accuracy": 0.5416666666666666,
      "eval_loss": 1.6484360694885254,
      "eval_runtime": 0.6251,
      "eval_samples_per_second": 115.18,
      "eval_steps_per_second": 7.999,
      "step": 16
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 18.5,
      "learning_rate": 4.1477272727272734e-05,
      "loss": 0.1876,
      "step": 17
    },
    {
      "epoch": 1.8888888888888888,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 1.70353102684021,
      "eval_runtime": 0.6234,
      "eval_samples_per_second": 115.489,
      "eval_steps_per_second": 8.02,
      "step": 17
    },
    {
      "epoch": 2.0,
      "grad_norm": 70.0,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 0.3145,
      "step": 18
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 1.67561936378479,
      "eval_runtime": 0.6238,
      "eval_samples_per_second": 115.418,
      "eval_steps_per_second": 8.015,
      "step": 18
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 26.625,
      "learning_rate": 4.034090909090909e-05,
      "loss": 0.1101,
      "step": 19
    },
    {
      "epoch": 2.111111111111111,
      "eval_accuracy": 0.4722222222222222,
      "eval_loss": 1.6205506324768066,
      "eval_runtime": 0.6263,
      "eval_samples_per_second": 114.968,
      "eval_steps_per_second": 7.984,
      "step": 19
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.625,
      "learning_rate": 3.9772727272727275e-05,
      "loss": 0.06,
      "step": 20
    },
    {
      "epoch": 2.2222222222222223,
      "eval_accuracy": 0.4583333333333333,
      "eval_loss": 1.606693148612976,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 115.355,
      "eval_steps_per_second": 8.011,
      "step": 20
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 13.375,
      "learning_rate": 3.9204545454545456e-05,
      "loss": 0.1102,
      "step": 21
    },
    {
      "epoch": 2.3333333333333335,
      "eval_accuracy": 0.4861111111111111,
      "eval_loss": 1.5763201713562012,
      "eval_runtime": 0.6243,
      "eval_samples_per_second": 115.323,
      "eval_steps_per_second": 8.009,
      "step": 21
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 11.875,
      "learning_rate": 3.8636363636363636e-05,
      "loss": 0.0756,
      "step": 22
    },
    {
      "epoch": 2.4444444444444446,
      "eval_accuracy": 0.5277777777777778,
      "eval_loss": 1.517237901687622,
      "eval_runtime": 0.6287,
      "eval_samples_per_second": 114.513,
      "eval_steps_per_second": 7.952,
      "step": 22
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 16.875,
      "learning_rate": 3.8068181818181816e-05,
      "loss": 0.0421,
      "step": 23
    },
    {
      "epoch": 2.5555555555555554,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 1.479959487915039,
      "eval_runtime": 0.6268,
      "eval_samples_per_second": 114.877,
      "eval_steps_per_second": 7.978,
      "step": 23
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 23.125,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0681,
      "step": 24
    },
    {
      "epoch": 2.6666666666666665,
      "eval_accuracy": 0.5555555555555556,
      "eval_loss": 1.4677960872650146,
      "eval_runtime": 0.6261,
      "eval_samples_per_second": 114.994,
      "eval_steps_per_second": 7.986,
      "step": 24
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 6.21875,
      "learning_rate": 3.6931818181818184e-05,
      "loss": 0.0477,
      "step": 25
    },
    {
      "epoch": 2.7777777777777777,
      "eval_accuracy": 0.5694444444444444,
      "eval_loss": 1.4750635623931885,
      "eval_runtime": 0.6231,
      "eval_samples_per_second": 115.55,
      "eval_steps_per_second": 8.024,
      "step": 25
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 6.0,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.0144,
      "step": 26
    },
    {
      "epoch": 2.888888888888889,
      "eval_accuracy": 0.5972222222222222,
      "eval_loss": 1.4336278438568115,
      "eval_runtime": 0.6225,
      "eval_samples_per_second": 115.661,
      "eval_steps_per_second": 8.032,
      "step": 26
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.375,
      "learning_rate": 3.579545454545455e-05,
      "loss": 0.0773,
      "step": 27
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5972222222222222,
      "eval_loss": 1.3547403812408447,
      "eval_runtime": 0.6235,
      "eval_samples_per_second": 115.474,
      "eval_steps_per_second": 8.019,
      "step": 27
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 1.46875,
      "learning_rate": 3.522727272727273e-05,
      "loss": 0.004,
      "step": 28
    },
    {
      "epoch": 3.111111111111111,
      "eval_accuracy": 0.5972222222222222,
      "eval_loss": 1.2619075775146484,
      "eval_runtime": 0.6216,
      "eval_samples_per_second": 115.826,
      "eval_steps_per_second": 8.043,
      "step": 28
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 1.8671875,
      "learning_rate": 3.465909090909091e-05,
      "loss": 0.0073,
      "step": 29
    },
    {
      "epoch": 3.2222222222222223,
      "eval_accuracy": 0.5972222222222222,
      "eval_loss": 1.2116299867630005,
      "eval_runtime": 0.6311,
      "eval_samples_per_second": 114.086,
      "eval_steps_per_second": 7.923,
      "step": 29
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.40234375,
      "learning_rate": 3.409090909090909e-05,
      "loss": 0.0034,
      "step": 30
    },
    {
      "epoch": 3.3333333333333335,
      "eval_accuracy": 0.7222222222222222,
      "eval_loss": 1.2432918548583984,
      "eval_runtime": 0.6245,
      "eval_samples_per_second": 115.286,
      "eval_steps_per_second": 8.006,
      "step": 30
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 3.3125,
      "learning_rate": 3.352272727272727e-05,
      "loss": 0.0059,
      "step": 31
    },
    {
      "epoch": 3.4444444444444446,
      "eval_accuracy": 0.6944444444444444,
      "eval_loss": 1.2839678525924683,
      "eval_runtime": 0.6231,
      "eval_samples_per_second": 115.559,
      "eval_steps_per_second": 8.025,
      "step": 31
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 2.890625,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.0083,
      "step": 32
    },
    {
      "epoch": 3.5555555555555554,
      "eval_accuracy": 0.7222222222222222,
      "eval_loss": 1.2911839485168457,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 115.344,
      "eval_steps_per_second": 8.01,
      "step": 32
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.6796875,
      "learning_rate": 3.238636363636364e-05,
      "loss": 0.0012,
      "step": 33
    },
    {
      "epoch": 3.6666666666666665,
      "eval_accuracy": 0.7222222222222222,
      "eval_loss": 1.2517170906066895,
      "eval_runtime": 0.6243,
      "eval_samples_per_second": 115.335,
      "eval_steps_per_second": 8.009,
      "step": 33
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 0.86328125,
      "learning_rate": 3.181818181818182e-05,
      "loss": 0.0014,
      "step": 34
    },
    {
      "epoch": 3.7777777777777777,
      "eval_accuracy": 0.7222222222222222,
      "eval_loss": 1.1986007690429688,
      "eval_runtime": 0.5252,
      "eval_samples_per_second": 137.095,
      "eval_steps_per_second": 9.52,
      "step": 34
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.70703125,
      "learning_rate": 3.125e-05,
      "loss": 0.0013,
      "step": 35
    },
    {
      "epoch": 3.888888888888889,
      "eval_accuracy": 0.7361111111111112,
      "eval_loss": 1.1582382917404175,
      "eval_runtime": 0.6232,
      "eval_samples_per_second": 115.53,
      "eval_steps_per_second": 8.023,
      "step": 35
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.07568359375,
      "learning_rate": 3.068181818181818e-05,
      "loss": 0.0004,
      "step": 36
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7361111111111112,
      "eval_loss": 1.1583126783370972,
      "eval_runtime": 0.6233,
      "eval_samples_per_second": 115.513,
      "eval_steps_per_second": 8.022,
      "step": 36
    },
    {
      "epoch": 4.111111111111111,
      "grad_norm": 0.06591796875,
      "learning_rate": 3.0113636363636365e-05,
      "loss": 0.0002,
      "step": 37
    },
    {
      "epoch": 4.111111111111111,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.1871634721755981,
      "eval_runtime": 0.6234,
      "eval_samples_per_second": 115.494,
      "eval_steps_per_second": 8.02,
      "step": 37
    },
    {
      "epoch": 4.222222222222222,
      "grad_norm": 0.06201171875,
      "learning_rate": 2.954545454545455e-05,
      "loss": 0.0001,
      "step": 38
    },
    {
      "epoch": 4.222222222222222,
      "eval_accuracy": 0.7361111111111112,
      "eval_loss": 1.2237037420272827,
      "eval_runtime": 0.6228,
      "eval_samples_per_second": 115.606,
      "eval_steps_per_second": 8.028,
      "step": 38
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.0615234375,
      "learning_rate": 2.8977272727272732e-05,
      "loss": 0.0001,
      "step": 39
    },
    {
      "epoch": 4.333333333333333,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2508012056350708,
      "eval_runtime": 0.623,
      "eval_samples_per_second": 115.574,
      "eval_steps_per_second": 8.026,
      "step": 39
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.734375,
      "learning_rate": 2.8409090909090912e-05,
      "loss": 0.0011,
      "step": 40
    },
    {
      "epoch": 4.444444444444445,
      "eval_accuracy": 0.7361111111111112,
      "eval_loss": 1.2626367807388306,
      "eval_runtime": 0.625,
      "eval_samples_per_second": 115.193,
      "eval_steps_per_second": 8.0,
      "step": 40
    },
    {
      "epoch": 4.555555555555555,
      "grad_norm": 0.07177734375,
      "learning_rate": 2.784090909090909e-05,
      "loss": 0.0001,
      "step": 41
    },
    {
      "epoch": 4.555555555555555,
      "eval_accuracy": 0.7361111111111112,
      "eval_loss": 1.2784370183944702,
      "eval_runtime": 0.623,
      "eval_samples_per_second": 115.577,
      "eval_steps_per_second": 8.026,
      "step": 41
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.84765625,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.0006,
      "step": 42
    },
    {
      "epoch": 4.666666666666667,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.2676023244857788,
      "eval_runtime": 0.6247,
      "eval_samples_per_second": 115.264,
      "eval_steps_per_second": 8.004,
      "step": 42
    },
    {
      "epoch": 4.777777777777778,
      "grad_norm": 0.00823974609375,
      "learning_rate": 2.6704545454545453e-05,
      "loss": 0.0,
      "step": 43
    },
    {
      "epoch": 4.777777777777778,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.3084845542907715,
      "eval_runtime": 0.6226,
      "eval_samples_per_second": 115.635,
      "eval_steps_per_second": 8.03,
      "step": 43
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 0.1455078125,
      "learning_rate": 2.6136363636363637e-05,
      "loss": 0.0002,
      "step": 44
    },
    {
      "epoch": 4.888888888888889,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.3400903940200806,
      "eval_runtime": 0.6218,
      "eval_samples_per_second": 115.794,
      "eval_steps_per_second": 8.041,
      "step": 44
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.006134033203125,
      "learning_rate": 2.5568181818181817e-05,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.3613409996032715,
      "eval_runtime": 0.6243,
      "eval_samples_per_second": 115.324,
      "eval_steps_per_second": 8.009,
      "step": 45
    },
    {
      "epoch": 5.111111111111111,
      "grad_norm": 0.0120849609375,
      "learning_rate": 2.5e-05,
      "loss": 0.0,
      "step": 46
    },
    {
      "epoch": 5.111111111111111,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.3976116180419922,
      "eval_runtime": 0.6228,
      "eval_samples_per_second": 115.607,
      "eval_steps_per_second": 8.028,
      "step": 46
    },
    {
      "epoch": 5.222222222222222,
      "grad_norm": 0.0164794921875,
      "learning_rate": 2.4431818181818185e-05,
      "loss": 0.0,
      "step": 47
    },
    {
      "epoch": 5.222222222222222,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.418382167816162,
      "eval_runtime": 0.6217,
      "eval_samples_per_second": 115.803,
      "eval_steps_per_second": 8.042,
      "step": 47
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.0301513671875,
      "learning_rate": 2.3863636363636365e-05,
      "loss": 0.0001,
      "step": 48
    },
    {
      "epoch": 5.333333333333333,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.4322800636291504,
      "eval_runtime": 0.6241,
      "eval_samples_per_second": 115.372,
      "eval_steps_per_second": 8.012,
      "step": 48
    },
    {
      "epoch": 5.444444444444445,
      "grad_norm": 0.00092315673828125,
      "learning_rate": 2.3295454545454546e-05,
      "loss": 0.0,
      "step": 49
    },
    {
      "epoch": 5.444444444444445,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.4491381645202637,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 115.342,
      "eval_steps_per_second": 8.01,
      "step": 49
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.000576019287109375,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 5.555555555555555,
      "eval_accuracy": 0.7638888888888888,
      "eval_loss": 1.4777884483337402,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 115.346,
      "eval_steps_per_second": 8.01,
      "step": 50
    },
    {
      "epoch": 5.666666666666667,
      "grad_norm": 0.0260009765625,
      "learning_rate": 2.215909090909091e-05,
      "loss": 0.0001,
      "step": 51
    },
    {
      "epoch": 5.666666666666667,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.4906997680664062,
      "eval_runtime": 0.6249,
      "eval_samples_per_second": 115.212,
      "eval_steps_per_second": 8.001,
      "step": 51
    },
    {
      "epoch": 5.777777777777778,
      "grad_norm": 0.00079345703125,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.0,
      "step": 52
    },
    {
      "epoch": 5.777777777777778,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.4938912391662598,
      "eval_runtime": 0.6254,
      "eval_samples_per_second": 115.131,
      "eval_steps_per_second": 7.995,
      "step": 52
    },
    {
      "epoch": 5.888888888888889,
      "grad_norm": 0.0172119140625,
      "learning_rate": 2.1022727272727274e-05,
      "loss": 0.0,
      "step": 53
    },
    {
      "epoch": 5.888888888888889,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.5104340314865112,
      "eval_runtime": 0.625,
      "eval_samples_per_second": 115.196,
      "eval_steps_per_second": 8.0,
      "step": 53
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.057861328125,
      "learning_rate": 2.0454545454545457e-05,
      "loss": 0.0001,
      "step": 54
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.5254335403442383,
      "eval_runtime": 0.625,
      "eval_samples_per_second": 115.196,
      "eval_steps_per_second": 8.0,
      "step": 54
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.0032806396484375,
      "learning_rate": 1.9886363636363638e-05,
      "loss": 0.0,
      "step": 55
    },
    {
      "epoch": 6.111111111111111,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.542032241821289,
      "eval_runtime": 0.6326,
      "eval_samples_per_second": 113.813,
      "eval_steps_per_second": 7.904,
      "step": 55
    },
    {
      "epoch": 6.222222222222222,
      "grad_norm": 0.021728515625,
      "learning_rate": 1.9318181818181818e-05,
      "loss": 0.0,
      "step": 56
    },
    {
      "epoch": 6.222222222222222,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.5513436794281006,
      "eval_runtime": 0.6244,
      "eval_samples_per_second": 115.315,
      "eval_steps_per_second": 8.008,
      "step": 56
    },
    {
      "epoch": 6.333333333333333,
      "grad_norm": 0.00115203857421875,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0,
      "step": 57
    },
    {
      "epoch": 6.333333333333333,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.5597673654556274,
      "eval_runtime": 0.6252,
      "eval_samples_per_second": 115.161,
      "eval_steps_per_second": 7.997,
      "step": 57
    },
    {
      "epoch": 6.444444444444445,
      "grad_norm": 0.029052734375,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.0,
      "step": 58
    },
    {
      "epoch": 6.444444444444445,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.575224757194519,
      "eval_runtime": 0.6248,
      "eval_samples_per_second": 115.228,
      "eval_steps_per_second": 8.002,
      "step": 58
    },
    {
      "epoch": 6.555555555555555,
      "grad_norm": 0.004058837890625,
      "learning_rate": 1.7613636363636366e-05,
      "loss": 0.0,
      "step": 59
    },
    {
      "epoch": 6.555555555555555,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.5835950374603271,
      "eval_runtime": 0.6244,
      "eval_samples_per_second": 115.306,
      "eval_steps_per_second": 8.007,
      "step": 59
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.0010528564453125,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.0,
      "step": 60
    },
    {
      "epoch": 6.666666666666667,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.587098479270935,
      "eval_runtime": 0.645,
      "eval_samples_per_second": 111.621,
      "eval_steps_per_second": 7.751,
      "step": 60
    },
    {
      "epoch": 6.777777777777778,
      "grad_norm": 0.004058837890625,
      "learning_rate": 1.6477272727272726e-05,
      "loss": 0.0,
      "step": 61
    },
    {
      "epoch": 6.777777777777778,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.5953326225280762,
      "eval_runtime": 0.6226,
      "eval_samples_per_second": 115.648,
      "eval_steps_per_second": 8.031,
      "step": 61
    },
    {
      "epoch": 6.888888888888889,
      "grad_norm": 0.0030670166015625,
      "learning_rate": 1.590909090909091e-05,
      "loss": 0.0,
      "step": 62
    },
    {
      "epoch": 6.888888888888889,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.5967087745666504,
      "eval_runtime": 0.6253,
      "eval_samples_per_second": 115.146,
      "eval_steps_per_second": 7.996,
      "step": 62
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.002593994140625,
      "learning_rate": 1.534090909090909e-05,
      "loss": 0.0,
      "step": 63
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6085822582244873,
      "eval_runtime": 0.6253,
      "eval_samples_per_second": 115.153,
      "eval_steps_per_second": 7.997,
      "step": 63
    },
    {
      "epoch": 7.111111111111111,
      "grad_norm": 0.00738525390625,
      "learning_rate": 1.4772727272727274e-05,
      "loss": 0.0,
      "step": 64
    },
    {
      "epoch": 7.111111111111111,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.597582459449768,
      "eval_runtime": 0.5737,
      "eval_samples_per_second": 125.507,
      "eval_steps_per_second": 8.716,
      "step": 64
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.0018463134765625,
      "learning_rate": 1.4204545454545456e-05,
      "loss": 0.0,
      "step": 65
    },
    {
      "epoch": 7.222222222222222,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.600056529045105,
      "eval_runtime": 0.6217,
      "eval_samples_per_second": 115.809,
      "eval_steps_per_second": 8.042,
      "step": 65
    },
    {
      "epoch": 7.333333333333333,
      "grad_norm": 0.000949859619140625,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.0,
      "step": 66
    },
    {
      "epoch": 7.333333333333333,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6038472652435303,
      "eval_runtime": 0.6271,
      "eval_samples_per_second": 114.813,
      "eval_steps_per_second": 7.973,
      "step": 66
    },
    {
      "epoch": 7.444444444444445,
      "grad_norm": 0.0047607421875,
      "learning_rate": 1.3068181818181819e-05,
      "loss": 0.0,
      "step": 67
    },
    {
      "epoch": 7.444444444444445,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6238569021224976,
      "eval_runtime": 0.6258,
      "eval_samples_per_second": 115.045,
      "eval_steps_per_second": 7.989,
      "step": 67
    },
    {
      "epoch": 7.555555555555555,
      "grad_norm": 0.0016021728515625,
      "learning_rate": 1.25e-05,
      "loss": 0.0,
      "step": 68
    },
    {
      "epoch": 7.555555555555555,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.6014671325683594,
      "eval_runtime": 0.6227,
      "eval_samples_per_second": 115.634,
      "eval_steps_per_second": 8.03,
      "step": 68
    },
    {
      "epoch": 7.666666666666667,
      "grad_norm": 0.0003643035888671875,
      "learning_rate": 1.1931818181818183e-05,
      "loss": 0.0,
      "step": 69
    },
    {
      "epoch": 7.666666666666667,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6172735691070557,
      "eval_runtime": 0.6399,
      "eval_samples_per_second": 112.522,
      "eval_steps_per_second": 7.814,
      "step": 69
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.0137939453125,
      "learning_rate": 1.1363636363636365e-05,
      "loss": 0.0,
      "step": 70
    },
    {
      "epoch": 7.777777777777778,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6042041778564453,
      "eval_runtime": 0.6241,
      "eval_samples_per_second": 115.365,
      "eval_steps_per_second": 8.011,
      "step": 70
    },
    {
      "epoch": 7.888888888888889,
      "grad_norm": 0.01336669921875,
      "learning_rate": 1.0795454545454547e-05,
      "loss": 0.0,
      "step": 71
    },
    {
      "epoch": 7.888888888888889,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6198363304138184,
      "eval_runtime": 0.5514,
      "eval_samples_per_second": 130.587,
      "eval_steps_per_second": 9.069,
      "step": 71
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.00157928466796875,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.0,
      "step": 72
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6046463251113892,
      "eval_runtime": 0.6274,
      "eval_samples_per_second": 114.752,
      "eval_steps_per_second": 7.969,
      "step": 72
    },
    {
      "epoch": 8.11111111111111,
      "grad_norm": 0.006805419921875,
      "learning_rate": 9.659090909090909e-06,
      "loss": 0.0,
      "step": 73
    },
    {
      "epoch": 8.11111111111111,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6086045503616333,
      "eval_runtime": 0.5492,
      "eval_samples_per_second": 131.099,
      "eval_steps_per_second": 9.104,
      "step": 73
    },
    {
      "epoch": 8.222222222222221,
      "grad_norm": 0.00159454345703125,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.0,
      "step": 74
    },
    {
      "epoch": 8.222222222222221,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6164934635162354,
      "eval_runtime": 0.6241,
      "eval_samples_per_second": 115.374,
      "eval_steps_per_second": 8.012,
      "step": 74
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.007659912109375,
      "learning_rate": 8.522727272727273e-06,
      "loss": 0.0,
      "step": 75
    },
    {
      "epoch": 8.333333333333334,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6239551305770874,
      "eval_runtime": 0.6237,
      "eval_samples_per_second": 115.434,
      "eval_steps_per_second": 8.016,
      "step": 75
    },
    {
      "epoch": 8.444444444444445,
      "grad_norm": 0.01190185546875,
      "learning_rate": 7.954545454545455e-06,
      "loss": 0.0,
      "step": 76
    },
    {
      "epoch": 8.444444444444445,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6252025365829468,
      "eval_runtime": 0.6251,
      "eval_samples_per_second": 115.189,
      "eval_steps_per_second": 7.999,
      "step": 76
    },
    {
      "epoch": 8.555555555555555,
      "grad_norm": 0.005950927734375,
      "learning_rate": 7.386363636363637e-06,
      "loss": 0.0,
      "step": 77
    },
    {
      "epoch": 8.555555555555555,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6214745044708252,
      "eval_runtime": 0.6243,
      "eval_samples_per_second": 115.333,
      "eval_steps_per_second": 8.009,
      "step": 77
    },
    {
      "epoch": 8.666666666666666,
      "grad_norm": 0.0004711151123046875,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.0,
      "step": 78
    },
    {
      "epoch": 8.666666666666666,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.621131420135498,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 115.354,
      "eval_steps_per_second": 8.011,
      "step": 78
    },
    {
      "epoch": 8.777777777777779,
      "grad_norm": 0.006195068359375,
      "learning_rate": 6.25e-06,
      "loss": 0.0,
      "step": 79
    },
    {
      "epoch": 8.777777777777779,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6142359972000122,
      "eval_runtime": 0.6237,
      "eval_samples_per_second": 115.441,
      "eval_steps_per_second": 8.017,
      "step": 79
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.00060272216796875,
      "learning_rate": 5.681818181818182e-06,
      "loss": 0.0,
      "step": 80
    },
    {
      "epoch": 8.88888888888889,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6200090646743774,
      "eval_runtime": 0.624,
      "eval_samples_per_second": 115.375,
      "eval_steps_per_second": 8.012,
      "step": 80
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.0025634765625,
      "learning_rate": 5.113636363636364e-06,
      "loss": 0.0,
      "step": 81
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6171174049377441,
      "eval_runtime": 0.6229,
      "eval_samples_per_second": 115.582,
      "eval_steps_per_second": 8.027,
      "step": 81
    },
    {
      "epoch": 9.11111111111111,
      "grad_norm": 0.01220703125,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.0,
      "step": 82
    },
    {
      "epoch": 9.11111111111111,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6105068922042847,
      "eval_runtime": 0.6248,
      "eval_samples_per_second": 115.24,
      "eval_steps_per_second": 8.003,
      "step": 82
    },
    {
      "epoch": 9.222222222222221,
      "grad_norm": 0.00167083740234375,
      "learning_rate": 3.9772727272727275e-06,
      "loss": 0.0,
      "step": 83
    },
    {
      "epoch": 9.222222222222221,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6220049858093262,
      "eval_runtime": 0.6253,
      "eval_samples_per_second": 115.141,
      "eval_steps_per_second": 7.996,
      "step": 83
    },
    {
      "epoch": 9.333333333333334,
      "grad_norm": 0.0031280517578125,
      "learning_rate": 3.409090909090909e-06,
      "loss": 0.0,
      "step": 84
    },
    {
      "epoch": 9.333333333333334,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.621678113937378,
      "eval_runtime": 0.6233,
      "eval_samples_per_second": 115.516,
      "eval_steps_per_second": 8.022,
      "step": 84
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.01220703125,
      "learning_rate": 2.840909090909091e-06,
      "loss": 0.0,
      "step": 85
    },
    {
      "epoch": 9.444444444444445,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6148762702941895,
      "eval_runtime": 0.6224,
      "eval_samples_per_second": 115.681,
      "eval_steps_per_second": 8.033,
      "step": 85
    },
    {
      "epoch": 9.555555555555555,
      "grad_norm": 0.00014972686767578125,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.0,
      "step": 86
    },
    {
      "epoch": 9.555555555555555,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6093779802322388,
      "eval_runtime": 0.6232,
      "eval_samples_per_second": 115.534,
      "eval_steps_per_second": 8.023,
      "step": 86
    },
    {
      "epoch": 9.666666666666666,
      "grad_norm": 0.0012054443359375,
      "learning_rate": 1.7045454545454546e-06,
      "loss": 0.0,
      "step": 87
    },
    {
      "epoch": 9.666666666666666,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.612194538116455,
      "eval_runtime": 0.6254,
      "eval_samples_per_second": 115.134,
      "eval_steps_per_second": 7.995,
      "step": 87
    },
    {
      "epoch": 9.777777777777779,
      "grad_norm": 0.0096435546875,
      "learning_rate": 1.1363636363636364e-06,
      "loss": 0.0,
      "step": 88
    },
    {
      "epoch": 9.777777777777779,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.6344364881515503,
      "eval_runtime": 0.6244,
      "eval_samples_per_second": 115.306,
      "eval_steps_per_second": 8.007,
      "step": 88
    },
    {
      "epoch": 9.88888888888889,
      "grad_norm": 0.00136566162109375,
      "learning_rate": 5.681818181818182e-07,
      "loss": 0.0,
      "step": 89
    },
    {
      "epoch": 9.88888888888889,
      "eval_accuracy": 0.7916666666666666,
      "eval_loss": 1.6183820962905884,
      "eval_runtime": 0.6241,
      "eval_samples_per_second": 115.365,
      "eval_steps_per_second": 8.011,
      "step": 89
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.000919342041015625,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 90
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7777777777777778,
      "eval_loss": 1.621146321296692,
      "eval_runtime": 0.6236,
      "eval_samples_per_second": 115.455,
      "eval_steps_per_second": 8.018,
      "step": 90
    },
    {
      "epoch": 10.0,
      "step": 90,
      "total_flos": 4834098484871168.0,
      "train_loss": 0.24404720813035966,
      "train_runtime": 123.0578,
      "train_samples_per_second": 23.079,
      "train_steps_per_second": 0.731
    }
  ],
  "logging_steps": 1,
  "max_steps": 90,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4834098484871168.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}