{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0004804814424052901,
  "eval_steps": 6,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.008024040088168e-06,
      "grad_norm": 282.1022644042969,
      "learning_rate": 0.0,
      "loss": 7.5794,
      "num_input_tokens_seen": 406,
      "step": 1,
      "train_runtime": 5.5308,
      "train_tokens_per_second": 73.407
    },
    {
      "epoch": 1.6016048080176336e-05,
      "grad_norm": 367.8074645996094,
      "learning_rate": 4.000000000000001e-06,
      "loss": 7.2589,
      "num_input_tokens_seen": 798,
      "step": 2,
      "train_runtime": 6.2015,
      "train_tokens_per_second": 128.678
    },
    {
      "epoch": 2.4024072120264505e-05,
      "grad_norm": 183.4268035888672,
      "learning_rate": 8.000000000000001e-06,
      "loss": 7.2901,
      "num_input_tokens_seen": 1236,
      "step": 3,
      "train_runtime": 6.8675,
      "train_tokens_per_second": 179.978
    },
    {
      "epoch": 3.203209616035267e-05,
      "grad_norm": 157.8712158203125,
      "learning_rate": 1.2e-05,
      "loss": 4.8904,
      "num_input_tokens_seen": 1676,
      "step": 4,
      "train_runtime": 7.5348,
      "train_tokens_per_second": 222.434
    },
    {
      "epoch": 4.0040120200440844e-05,
      "grad_norm": 153.35537719726562,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 4.4882,
      "num_input_tokens_seen": 2068,
      "step": 5,
      "train_runtime": 8.1988,
      "train_tokens_per_second": 252.231
    },
    {
      "epoch": 4.804814424052901e-05,
      "grad_norm": 154.29891967773438,
      "learning_rate": 2e-05,
      "loss": 4.2164,
      "num_input_tokens_seen": 2496,
      "step": 6,
      "train_runtime": 8.8628,
      "train_tokens_per_second": 281.628
    },
    {
      "epoch": 4.804814424052901e-05,
      "eval_loss": 3.4333748817443848,
      "eval_runtime": 17.1905,
      "eval_samples_per_second": 58.172,
      "eval_steps_per_second": 29.086,
      "num_input_tokens_seen": 2496,
      "step": 6
    },
    {
      "epoch": 5.605616828061718e-05,
      "grad_norm": 43.233089447021484,
      "learning_rate": 1.963636363636364e-05,
      "loss": 2.0546,
      "num_input_tokens_seen": 3238,
      "step": 7,
      "train_runtime": 26.7603,
      "train_tokens_per_second": 121.0
    },
    {
      "epoch": 6.406419232070534e-05,
      "grad_norm": 41.88556671142578,
      "learning_rate": 1.9272727272727275e-05,
      "loss": 2.372,
      "num_input_tokens_seen": 3828,
      "step": 8,
      "train_runtime": 27.4263,
      "train_tokens_per_second": 139.574
    },
    {
      "epoch": 7.207221636079351e-05,
      "grad_norm": 48.73846435546875,
      "learning_rate": 1.8909090909090912e-05,
      "loss": 2.7182,
      "num_input_tokens_seen": 4280,
      "step": 9,
      "train_runtime": 28.0921,
      "train_tokens_per_second": 152.356
    },
    {
      "epoch": 8.008024040088169e-05,
      "grad_norm": 46.16472244262695,
      "learning_rate": 1.8545454545454545e-05,
      "loss": 2.7735,
      "num_input_tokens_seen": 4702,
      "step": 10,
      "train_runtime": 28.7563,
      "train_tokens_per_second": 163.512
    },
    {
      "epoch": 8.808826444096985e-05,
      "grad_norm": 25.96290397644043,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 1.8736,
      "num_input_tokens_seen": 5366,
      "step": 11,
      "train_runtime": 29.4216,
      "train_tokens_per_second": 182.383
    },
    {
      "epoch": 9.609628848105802e-05,
      "grad_norm": 35.26546096801758,
      "learning_rate": 1.781818181818182e-05,
      "loss": 2.4054,
      "num_input_tokens_seen": 5898,
      "step": 12,
      "train_runtime": 30.0864,
      "train_tokens_per_second": 196.036
    },
    {
      "epoch": 9.609628848105802e-05,
      "eval_loss": 2.551903486251831,
      "eval_runtime": 16.1332,
      "eval_samples_per_second": 61.984,
      "eval_steps_per_second": 30.992,
      "num_input_tokens_seen": 5898,
      "step": 12
    },
    {
      "epoch": 0.00010410431252114619,
      "grad_norm": 45.30570983886719,
      "learning_rate": 1.7454545454545456e-05,
      "loss": 2.6956,
      "num_input_tokens_seen": 6368,
      "step": 13,
      "train_runtime": 46.917,
      "train_tokens_per_second": 135.729
    },
    {
      "epoch": 0.00011211233656123435,
      "grad_norm": 25.79911231994629,
      "learning_rate": 1.7090909090909092e-05,
      "loss": 1.9536,
      "num_input_tokens_seen": 7166,
      "step": 14,
      "train_runtime": 47.6787,
      "train_tokens_per_second": 150.298
    },
    {
      "epoch": 0.00012012036060132252,
      "grad_norm": 35.28297805786133,
      "learning_rate": 1.672727272727273e-05,
      "loss": 1.8325,
      "num_input_tokens_seen": 7580,
      "step": 15,
      "train_runtime": 48.3409,
      "train_tokens_per_second": 156.803
    },
    {
      "epoch": 0.00012812838464141069,
      "grad_norm": 33.03861618041992,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 2.3691,
      "num_input_tokens_seen": 8078,
      "step": 16,
      "train_runtime": 49.006,
      "train_tokens_per_second": 164.837
    },
    {
      "epoch": 0.00013613640868149885,
      "grad_norm": 45.185157775878906,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.3737,
      "num_input_tokens_seen": 8448,
      "step": 17,
      "train_runtime": 49.6695,
      "train_tokens_per_second": 170.084
    },
    {
      "epoch": 0.00014414443272158702,
      "grad_norm": 31.473684310913086,
      "learning_rate": 1.563636363636364e-05,
      "loss": 2.427,
      "num_input_tokens_seen": 8936,
      "step": 18,
      "train_runtime": 50.3334,
      "train_tokens_per_second": 177.536
    },
    {
      "epoch": 0.00014414443272158702,
      "eval_loss": 2.4002556800842285,
      "eval_runtime": 15.9406,
      "eval_samples_per_second": 62.733,
      "eval_steps_per_second": 31.366,
      "num_input_tokens_seen": 8936,
      "step": 18
    },
    {
      "epoch": 0.0001521524567616752,
      "grad_norm": 33.58867263793945,
      "learning_rate": 1.5272727272727276e-05,
      "loss": 2.1379,
      "num_input_tokens_seen": 9360,
      "step": 19,
      "train_runtime": 66.9757,
      "train_tokens_per_second": 139.752
    },
    {
      "epoch": 0.00016016048080176338,
      "grad_norm": 33.38470458984375,
      "learning_rate": 1.4909090909090911e-05,
      "loss": 2.8262,
      "num_input_tokens_seen": 9786,
      "step": 20,
      "train_runtime": 67.641,
      "train_tokens_per_second": 144.676
    },
    {
      "epoch": 0.00016816850484185154,
      "grad_norm": 33.72075271606445,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 2.515,
      "num_input_tokens_seen": 10198,
      "step": 21,
      "train_runtime": 68.304,
      "train_tokens_per_second": 149.303
    },
    {
      "epoch": 0.0001761765288819397,
      "grad_norm": 35.71574401855469,
      "learning_rate": 1.4181818181818183e-05,
      "loss": 2.4477,
      "num_input_tokens_seen": 10644,
      "step": 22,
      "train_runtime": 68.9749,
      "train_tokens_per_second": 154.317
    },
    {
      "epoch": 0.00018418455292202788,
      "grad_norm": 33.68728256225586,
      "learning_rate": 1.381818181818182e-05,
      "loss": 2.4155,
      "num_input_tokens_seen": 11096,
      "step": 23,
      "train_runtime": 69.64,
      "train_tokens_per_second": 159.334
    },
    {
      "epoch": 0.00019219257696211604,
      "grad_norm": 35.62799072265625,
      "learning_rate": 1.3454545454545455e-05,
      "loss": 2.4388,
      "num_input_tokens_seen": 11546,
      "step": 24,
      "train_runtime": 70.3069,
      "train_tokens_per_second": 164.223
    },
    {
      "epoch": 0.00019219257696211604,
      "eval_loss": 2.3586089611053467,
      "eval_runtime": 15.9544,
      "eval_samples_per_second": 62.679,
      "eval_steps_per_second": 31.339,
      "num_input_tokens_seen": 11546,
      "step": 24
    },
    {
      "epoch": 0.0002002006010022042,
      "grad_norm": 33.08585739135742,
      "learning_rate": 1.3090909090909092e-05,
      "loss": 2.0016,
      "num_input_tokens_seen": 12020,
      "step": 25,
      "train_runtime": 86.9671,
      "train_tokens_per_second": 138.213
    },
    {
      "epoch": 0.00020820862504229237,
      "grad_norm": 37.98166275024414,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 2.5803,
      "num_input_tokens_seen": 12430,
      "step": 26,
      "train_runtime": 87.6331,
      "train_tokens_per_second": 141.841
    },
    {
      "epoch": 0.00021621664908238054,
      "grad_norm": 36.4505729675293,
      "learning_rate": 1.2363636363636364e-05,
      "loss": 3.0606,
      "num_input_tokens_seen": 12860,
      "step": 27,
      "train_runtime": 88.295,
      "train_tokens_per_second": 145.648
    },
    {
      "epoch": 0.0002242246731224687,
      "grad_norm": 34.81946563720703,
      "learning_rate": 1.2e-05,
      "loss": 1.7622,
      "num_input_tokens_seen": 13302,
      "step": 28,
      "train_runtime": 88.9585,
      "train_tokens_per_second": 149.53
    },
    {
      "epoch": 0.00023223269716255687,
      "grad_norm": 31.250852584838867,
      "learning_rate": 1.1636363636363637e-05,
      "loss": 2.3633,
      "num_input_tokens_seen": 13796,
      "step": 29,
      "train_runtime": 89.6239,
      "train_tokens_per_second": 153.932
    },
    {
      "epoch": 0.00024024072120264504,
      "grad_norm": 32.98413848876953,
      "learning_rate": 1.1272727272727272e-05,
      "loss": 2.5788,
      "num_input_tokens_seen": 14286,
      "step": 30,
      "train_runtime": 90.2907,
      "train_tokens_per_second": 158.222
    },
    {
      "epoch": 0.00024024072120264504,
      "eval_loss": 2.321453332901001,
      "eval_runtime": 15.9524,
      "eval_samples_per_second": 62.686,
      "eval_steps_per_second": 31.343,
      "num_input_tokens_seen": 14286,
      "step": 30
    },
    {
      "epoch": 0.0002482487452427332,
      "grad_norm": 36.93943405151367,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 2.5933,
      "num_input_tokens_seen": 14690,
      "step": 31,
      "train_runtime": 125.5173,
      "train_tokens_per_second": 117.036
    },
    {
      "epoch": 0.00025625676928282137,
      "grad_norm": 29.403127670288086,
      "learning_rate": 1.0545454545454546e-05,
      "loss": 1.8715,
      "num_input_tokens_seen": 15182,
      "step": 32,
      "train_runtime": 126.1813,
      "train_tokens_per_second": 120.319
    },
    {
      "epoch": 0.00026426479332290954,
      "grad_norm": 26.887046813964844,
      "learning_rate": 1.0181818181818182e-05,
      "loss": 2.3258,
      "num_input_tokens_seen": 15742,
      "step": 33,
      "train_runtime": 126.8482,
      "train_tokens_per_second": 124.101
    },
    {
      "epoch": 0.0002722728173629977,
      "grad_norm": 35.79985046386719,
      "learning_rate": 9.81818181818182e-06,
      "loss": 2.6422,
      "num_input_tokens_seen": 16170,
      "step": 34,
      "train_runtime": 127.5112,
      "train_tokens_per_second": 126.812
    },
    {
      "epoch": 0.00028028084140308587,
      "grad_norm": 29.967466354370117,
      "learning_rate": 9.454545454545456e-06,
      "loss": 2.1683,
      "num_input_tokens_seen": 16602,
      "step": 35,
      "train_runtime": 128.1746,
      "train_tokens_per_second": 129.526
    },
    {
      "epoch": 0.00028828886544317404,
      "grad_norm": 23.023691177368164,
      "learning_rate": 9.090909090909091e-06,
      "loss": 2.2433,
      "num_input_tokens_seen": 17328,
      "step": 36,
      "train_runtime": 128.8377,
      "train_tokens_per_second": 134.495
    },
    {
      "epoch": 0.00028828886544317404,
      "eval_loss": 2.3060390949249268,
      "eval_runtime": 15.9837,
      "eval_samples_per_second": 62.564,
      "eval_steps_per_second": 31.282,
      "num_input_tokens_seen": 17328,
      "step": 36
    },
    {
      "epoch": 0.0002962968894832622,
      "grad_norm": 36.571800231933594,
      "learning_rate": 8.727272727272728e-06,
      "loss": 2.2402,
      "num_input_tokens_seen": 17764,
      "step": 37,
      "train_runtime": 145.5214,
      "train_tokens_per_second": 122.071
    },
    {
      "epoch": 0.0003043049135233504,
      "grad_norm": 33.204383850097656,
      "learning_rate": 8.363636363636365e-06,
      "loss": 2.469,
      "num_input_tokens_seen": 18190,
      "step": 38,
      "train_runtime": 146.1828,
      "train_tokens_per_second": 124.433
    },
    {
      "epoch": 0.0003123129375634386,
      "grad_norm": 42.04680252075195,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.7519,
      "num_input_tokens_seen": 18618,
      "step": 39,
      "train_runtime": 146.8516,
      "train_tokens_per_second": 126.781
    },
    {
      "epoch": 0.00032032096160352676,
      "grad_norm": 30.66261100769043,
      "learning_rate": 7.636363636363638e-06,
      "loss": 2.5034,
      "num_input_tokens_seen": 19024,
      "step": 40,
      "train_runtime": 147.5219,
      "train_tokens_per_second": 128.957
    },
    {
      "epoch": 0.0003283289856436149,
      "grad_norm": 30.992774963378906,
      "learning_rate": 7.272727272727273e-06,
      "loss": 1.8883,
      "num_input_tokens_seen": 19492,
      "step": 41,
      "train_runtime": 148.1868,
      "train_tokens_per_second": 131.537
    },
    {
      "epoch": 0.0003363370096837031,
      "grad_norm": 33.176795959472656,
      "learning_rate": 6.90909090909091e-06,
      "loss": 2.329,
      "num_input_tokens_seen": 19906,
      "step": 42,
      "train_runtime": 148.8521,
      "train_tokens_per_second": 133.73
    },
    {
      "epoch": 0.0003363370096837031,
      "eval_loss": 2.2987871170043945,
      "eval_runtime": 15.9526,
      "eval_samples_per_second": 62.686,
      "eval_steps_per_second": 31.343,
      "num_input_tokens_seen": 19906,
      "step": 42
    },
    {
      "epoch": 0.00034434503372379125,
      "grad_norm": 36.44293975830078,
      "learning_rate": 6.545454545454546e-06,
      "loss": 2.4833,
      "num_input_tokens_seen": 20328,
      "step": 43,
      "train_runtime": 165.5052,
      "train_tokens_per_second": 122.824
    },
    {
      "epoch": 0.0003523530577638794,
      "grad_norm": 32.9749870300293,
      "learning_rate": 6.181818181818182e-06,
      "loss": 2.1394,
      "num_input_tokens_seen": 20756,
      "step": 44,
      "train_runtime": 166.1716,
      "train_tokens_per_second": 124.907
    },
    {
      "epoch": 0.0003603610818039676,
      "grad_norm": 28.772348403930664,
      "learning_rate": 5.8181818181818185e-06,
      "loss": 2.0539,
      "num_input_tokens_seen": 21202,
      "step": 45,
      "train_runtime": 166.8343,
      "train_tokens_per_second": 127.084
    },
    {
      "epoch": 0.00036836910584405575,
      "grad_norm": 28.623464584350586,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 2.2397,
      "num_input_tokens_seen": 21836,
      "step": 46,
      "train_runtime": 167.5037,
      "train_tokens_per_second": 130.361
    },
    {
      "epoch": 0.0003763771298841439,
      "grad_norm": 24.052433013916016,
      "learning_rate": 5.090909090909091e-06,
      "loss": 2.3161,
      "num_input_tokens_seen": 22416,
      "step": 47,
      "train_runtime": 168.1683,
      "train_tokens_per_second": 133.295
    },
    {
      "epoch": 0.0003843851539242321,
      "grad_norm": 27.369709014892578,
      "learning_rate": 4.727272727272728e-06,
      "loss": 1.9488,
      "num_input_tokens_seen": 23096,
      "step": 48,
      "train_runtime": 168.8317,
      "train_tokens_per_second": 136.799
    },
    {
      "epoch": 0.0003843851539242321,
      "eval_loss": 2.295372486114502,
      "eval_runtime": 15.9296,
      "eval_samples_per_second": 62.776,
      "eval_steps_per_second": 31.388,
      "num_input_tokens_seen": 23096,
      "step": 48
    },
    {
      "epoch": 0.00039239317796432025,
      "grad_norm": 21.76813507080078,
      "learning_rate": 4.363636363636364e-06,
      "loss": 2.3741,
      "num_input_tokens_seen": 23970,
      "step": 49,
      "train_runtime": 185.4643,
      "train_tokens_per_second": 129.243
    },
    {
      "epoch": 0.0004004012020044084,
      "grad_norm": 28.296794891357422,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.7647,
      "num_input_tokens_seen": 24458,
      "step": 50,
      "train_runtime": 186.1284,
      "train_tokens_per_second": 131.404
    },
    {
      "epoch": 0.0004084092260444966,
      "grad_norm": 34.54985427856445,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 2.2798,
      "num_input_tokens_seen": 24884,
      "step": 51,
      "train_runtime": 186.7918,
      "train_tokens_per_second": 133.218
    },
    {
      "epoch": 0.00041641725008458475,
      "grad_norm": 25.983028411865234,
      "learning_rate": 3.272727272727273e-06,
      "loss": 1.8212,
      "num_input_tokens_seen": 25394,
      "step": 52,
      "train_runtime": 187.454,
      "train_tokens_per_second": 135.468
    },
    {
      "epoch": 0.0004244252741246729,
      "grad_norm": 33.13299560546875,
      "learning_rate": 2.9090909090909093e-06,
      "loss": 2.4243,
      "num_input_tokens_seen": 25814,
      "step": 53,
      "train_runtime": 188.1191,
      "train_tokens_per_second": 137.222
    },
    {
      "epoch": 0.0004324332981647611,
      "grad_norm": 32.00737380981445,
      "learning_rate": 2.5454545454545456e-06,
      "loss": 2.0507,
      "num_input_tokens_seen": 26232,
      "step": 54,
      "train_runtime": 188.7874,
      "train_tokens_per_second": 138.95
    },
    {
      "epoch": 0.0004324332981647611,
      "eval_loss": 2.2933812141418457,
      "eval_runtime": 15.8975,
      "eval_samples_per_second": 62.903,
      "eval_steps_per_second": 31.451,
      "num_input_tokens_seen": 26232,
      "step": 54
    },
    {
      "epoch": 0.00044044132220484925,
      "grad_norm": 29.870973587036133,
      "learning_rate": 2.181818181818182e-06,
      "loss": 2.7572,
      "num_input_tokens_seen": 26724,
      "step": 55,
      "train_runtime": 205.3777,
      "train_tokens_per_second": 130.121
    },
    {
      "epoch": 0.0004484493462449374,
      "grad_norm": 32.467864990234375,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 2.5342,
      "num_input_tokens_seen": 27188,
      "step": 56,
      "train_runtime": 206.0422,
      "train_tokens_per_second": 131.954
    },
    {
      "epoch": 0.0004564573702850256,
      "grad_norm": 23.056095123291016,
      "learning_rate": 1.4545454545454546e-06,
      "loss": 1.8859,
      "num_input_tokens_seen": 27760,
      "step": 57,
      "train_runtime": 206.7074,
      "train_tokens_per_second": 134.296
    },
    {
      "epoch": 0.00046446539432511375,
      "grad_norm": 28.30994415283203,
      "learning_rate": 1.090909090909091e-06,
      "loss": 2.0119,
      "num_input_tokens_seen": 28288,
      "step": 58,
      "train_runtime": 207.3729,
      "train_tokens_per_second": 136.411
    },
    {
      "epoch": 0.0004724734183652019,
      "grad_norm": 32.38734817504883,
      "learning_rate": 7.272727272727273e-07,
      "loss": 2.3612,
      "num_input_tokens_seen": 28672,
      "step": 59,
      "train_runtime": 208.0357,
      "train_tokens_per_second": 137.822
    },
    {
      "epoch": 0.0004804814424052901,
      "grad_norm": 25.652999877929688,
      "learning_rate": 3.6363636363636366e-07,
      "loss": 1.8457,
      "num_input_tokens_seen": 29110,
      "step": 60,
      "train_runtime": 208.6978,
      "train_tokens_per_second": 139.484
    },
    {
      "epoch": 0.0004804814424052901,
      "eval_loss": 2.2931265830993652,
      "eval_runtime": 15.8795,
      "eval_samples_per_second": 62.974,
      "eval_steps_per_second": 31.487,
      "num_input_tokens_seen": 29110,
      "step": 60
    }
  ],
  "logging_steps": 1,
  "max_steps": 60,
  "num_input_tokens_seen": 29110,
  "num_train_epochs": 1,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 484641984921600.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}