{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 8500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 3.437100410461426,
      "learning_rate": 4.941764705882353e-05,
      "loss": 0.6786,
      "step": 100
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 2.240630626678467,
      "learning_rate": 4.8829411764705885e-05,
      "loss": 0.6502,
      "step": 200
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 4.658847808837891,
      "learning_rate": 4.824117647058824e-05,
      "loss": 0.554,
      "step": 300
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 2.393118143081665,
      "learning_rate": 4.765294117647059e-05,
      "loss": 0.5091,
      "step": 400
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 6.558012008666992,
      "learning_rate": 4.7064705882352944e-05,
      "loss": 0.4821,
      "step": 500
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 10.292436599731445,
      "learning_rate": 4.64764705882353e-05,
      "loss": 0.4832,
      "step": 600
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 2.2930338382720947,
      "learning_rate": 4.588823529411765e-05,
      "loss": 0.4906,
      "step": 700
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 4.4116692543029785,
      "learning_rate": 4.53e-05,
      "loss": 0.4212,
      "step": 800
    },
    {
      "epoch": 1.0588235294117647,
      "grad_norm": 11.04600715637207,
      "learning_rate": 4.4711764705882356e-05,
      "loss": 0.4047,
      "step": 900
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 3.98223614692688,
      "learning_rate": 4.412352941176471e-05,
      "loss": 0.3968,
      "step": 1000
    },
    {
      "epoch": 1.2941176470588236,
      "grad_norm": 5.051950931549072,
      "learning_rate": 4.353529411764706e-05,
      "loss": 0.3654,
      "step": 1100
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 3.390256404876709,
      "learning_rate": 4.2947058823529415e-05,
      "loss": 0.4115,
      "step": 1200
    },
    {
      "epoch": 1.5294117647058822,
      "grad_norm": 6.051326751708984,
      "learning_rate": 4.235882352941177e-05,
      "loss": 0.3766,
      "step": 1300
    },
    {
      "epoch": 1.6470588235294117,
      "grad_norm": 3.2749335765838623,
      "learning_rate": 4.1770588235294115e-05,
      "loss": 0.3872,
      "step": 1400
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 3.5865375995635986,
      "learning_rate": 4.1182352941176474e-05,
      "loss": 0.3628,
      "step": 1500
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 9.934630393981934,
      "learning_rate": 4.059411764705883e-05,
      "loss": 0.3044,
      "step": 1600
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.085042953491211,
      "learning_rate": 4.0005882352941174e-05,
      "loss": 0.3483,
      "step": 1700
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 10.278831481933594,
      "learning_rate": 3.941764705882353e-05,
      "loss": 0.3355,
      "step": 1800
    },
    {
      "epoch": 2.235294117647059,
      "grad_norm": 4.414193630218506,
      "learning_rate": 3.8829411764705886e-05,
      "loss": 0.3181,
      "step": 1900
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 9.129111289978027,
      "learning_rate": 3.824117647058823e-05,
      "loss": 0.3096,
      "step": 2000
    },
    {
      "epoch": 2.4705882352941178,
      "grad_norm": 37.28619384765625,
      "learning_rate": 3.765294117647059e-05,
      "loss": 0.2821,
      "step": 2100
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 0.5300999283790588,
      "learning_rate": 3.706470588235294e-05,
      "loss": 0.3126,
      "step": 2200
    },
    {
      "epoch": 2.7058823529411766,
      "grad_norm": 1.9004733562469482,
      "learning_rate": 3.64764705882353e-05,
      "loss": 0.3397,
      "step": 2300
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.971808910369873,
      "learning_rate": 3.588823529411765e-05,
      "loss": 0.3461,
      "step": 2400
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 10.828816413879395,
      "learning_rate": 3.53e-05,
      "loss": 0.2998,
      "step": 2500
    },
    {
      "epoch": 3.0588235294117645,
      "grad_norm": 5.21161413192749,
      "learning_rate": 3.471176470588236e-05,
      "loss": 0.287,
      "step": 2600
    },
    {
      "epoch": 3.176470588235294,
      "grad_norm": 17.639314651489258,
      "learning_rate": 3.412352941176471e-05,
      "loss": 0.2408,
      "step": 2700
    },
    {
      "epoch": 3.2941176470588234,
      "grad_norm": 5.808989524841309,
      "learning_rate": 3.3535294117647056e-05,
      "loss": 0.2514,
      "step": 2800
    },
    {
      "epoch": 3.411764705882353,
      "grad_norm": 5.669056415557861,
      "learning_rate": 3.2947058823529416e-05,
      "loss": 0.2401,
      "step": 2900
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 17.544368743896484,
      "learning_rate": 3.235882352941176e-05,
      "loss": 0.304,
      "step": 3000
    },
    {
      "epoch": 3.6470588235294117,
      "grad_norm": 15.652566909790039,
      "learning_rate": 3.1770588235294115e-05,
      "loss": 0.2169,
      "step": 3100
    },
    {
      "epoch": 3.764705882352941,
      "grad_norm": 38.110477447509766,
      "learning_rate": 3.1182352941176475e-05,
      "loss": 0.3245,
      "step": 3200
    },
    {
      "epoch": 3.8823529411764706,
      "grad_norm": 15.126811027526855,
      "learning_rate": 3.059411764705882e-05,
      "loss": 0.2366,
      "step": 3300
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.916968822479248,
      "learning_rate": 3.0005882352941178e-05,
      "loss": 0.2394,
      "step": 3400
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 6.685044765472412,
      "learning_rate": 2.9417647058823534e-05,
      "loss": 0.239,
      "step": 3500
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 0.2686949074268341,
      "learning_rate": 2.8829411764705884e-05,
      "loss": 0.2042,
      "step": 3600
    },
    {
      "epoch": 4.352941176470588,
      "grad_norm": 4.416964530944824,
      "learning_rate": 2.8241176470588237e-05,
      "loss": 0.235,
      "step": 3700
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 14.782841682434082,
      "learning_rate": 2.7652941176470586e-05,
      "loss": 0.2145,
      "step": 3800
    },
    {
      "epoch": 4.588235294117647,
      "grad_norm": 3.3013288974761963,
      "learning_rate": 2.7064705882352943e-05,
      "loss": 0.2301,
      "step": 3900
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 12.687139511108398,
      "learning_rate": 2.6476470588235296e-05,
      "loss": 0.2459,
      "step": 4000
    },
    {
      "epoch": 4.823529411764706,
      "grad_norm": 12.49465560913086,
      "learning_rate": 2.5888235294117645e-05,
      "loss": 0.2523,
      "step": 4100
    },
    {
      "epoch": 4.9411764705882355,
      "grad_norm": 7.881172180175781,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.2301,
      "step": 4200
    },
    {
      "epoch": 5.0588235294117645,
      "grad_norm": 7.88573694229126,
      "learning_rate": 2.4711764705882355e-05,
      "loss": 0.1962,
      "step": 4300
    },
    {
      "epoch": 5.176470588235294,
      "grad_norm": 8.924189567565918,
      "learning_rate": 2.4123529411764704e-05,
      "loss": 0.1782,
      "step": 4400
    },
    {
      "epoch": 5.294117647058823,
      "grad_norm": 0.7592102289199829,
      "learning_rate": 2.353529411764706e-05,
      "loss": 0.2081,
      "step": 4500
    },
    {
      "epoch": 5.411764705882353,
      "grad_norm": 4.540928363800049,
      "learning_rate": 2.2947058823529414e-05,
      "loss": 0.1991,
      "step": 4600
    },
    {
      "epoch": 5.529411764705882,
      "grad_norm": 2.462392807006836,
      "learning_rate": 2.2358823529411767e-05,
      "loss": 0.226,
      "step": 4700
    },
    {
      "epoch": 5.647058823529412,
      "grad_norm": 0.7198309898376465,
      "learning_rate": 2.1770588235294116e-05,
      "loss": 0.2093,
      "step": 4800
    },
    {
      "epoch": 5.764705882352941,
      "grad_norm": 0.04426976293325424,
      "learning_rate": 2.1182352941176473e-05,
      "loss": 0.1616,
      "step": 4900
    },
    {
      "epoch": 5.882352941176471,
      "grad_norm": 19.01175880432129,
      "learning_rate": 2.0594117647058826e-05,
      "loss": 0.2451,
      "step": 5000
    },
    {
      "epoch": 6.0,
      "grad_norm": 9.680488586425781,
      "learning_rate": 2.000588235294118e-05,
      "loss": 0.1954,
      "step": 5100
    },
    {
      "epoch": 6.117647058823529,
      "grad_norm": 13.620891571044922,
      "learning_rate": 1.9417647058823528e-05,
      "loss": 0.1739,
      "step": 5200
    },
    {
      "epoch": 6.235294117647059,
      "grad_norm": 23.215810775756836,
      "learning_rate": 1.8829411764705885e-05,
      "loss": 0.1989,
      "step": 5300
    },
    {
      "epoch": 6.352941176470588,
      "grad_norm": 34.69459915161133,
      "learning_rate": 1.8241176470588238e-05,
      "loss": 0.1845,
      "step": 5400
    },
    {
      "epoch": 6.470588235294118,
      "grad_norm": 0.17689703404903412,
      "learning_rate": 1.7652941176470587e-05,
      "loss": 0.1888,
      "step": 5500
    },
    {
      "epoch": 6.588235294117647,
      "grad_norm": 0.1681346297264099,
      "learning_rate": 1.706470588235294e-05,
      "loss": 0.1808,
      "step": 5600
    },
    {
      "epoch": 6.705882352941177,
      "grad_norm": 0.21678882837295532,
      "learning_rate": 1.6476470588235297e-05,
      "loss": 0.1527,
      "step": 5700
    },
    {
      "epoch": 6.823529411764706,
      "grad_norm": 6.419234752655029,
      "learning_rate": 1.588823529411765e-05,
      "loss": 0.2019,
      "step": 5800
    },
    {
      "epoch": 6.9411764705882355,
      "grad_norm": 0.09294360131025314,
      "learning_rate": 1.53e-05,
      "loss": 0.1881,
      "step": 5900
    },
    {
      "epoch": 7.0588235294117645,
      "grad_norm": 14.81273365020752,
      "learning_rate": 1.4711764705882352e-05,
      "loss": 0.1606,
      "step": 6000
    },
    {
      "epoch": 7.176470588235294,
      "grad_norm": 64.77528381347656,
      "learning_rate": 1.4123529411764707e-05,
      "loss": 0.103,
      "step": 6100
    },
    {
      "epoch": 7.294117647058823,
      "grad_norm": 0.01647937297821045,
      "learning_rate": 1.353529411764706e-05,
      "loss": 0.144,
      "step": 6200
    },
    {
      "epoch": 7.411764705882353,
      "grad_norm": 2.1152279376983643,
      "learning_rate": 1.2947058823529413e-05,
      "loss": 0.1687,
      "step": 6300
    },
    {
      "epoch": 7.529411764705882,
      "grad_norm": 5.980208396911621,
      "learning_rate": 1.2358823529411766e-05,
      "loss": 0.134,
      "step": 6400
    },
    {
      "epoch": 7.647058823529412,
      "grad_norm": 20.869848251342773,
      "learning_rate": 1.1770588235294117e-05,
      "loss": 0.1346,
      "step": 6500
    },
    {
      "epoch": 7.764705882352941,
      "grad_norm": 46.88475799560547,
      "learning_rate": 1.1182352941176472e-05,
      "loss": 0.1871,
      "step": 6600
    },
    {
      "epoch": 7.882352941176471,
      "grad_norm": 0.02130250073969364,
      "learning_rate": 1.0594117647058823e-05,
      "loss": 0.1287,
      "step": 6700
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.03343911096453667,
      "learning_rate": 1.0005882352941178e-05,
      "loss": 0.1196,
      "step": 6800
    },
    {
      "epoch": 8.117647058823529,
      "grad_norm": 0.6227496862411499,
      "learning_rate": 9.417647058823529e-06,
      "loss": 0.1297,
      "step": 6900
    },
    {
      "epoch": 8.235294117647058,
      "grad_norm": 0.6179134249687195,
      "learning_rate": 8.829411764705884e-06,
      "loss": 0.1624,
      "step": 7000
    },
    {
      "epoch": 8.352941176470589,
      "grad_norm": 0.015830175951123238,
      "learning_rate": 8.241176470588235e-06,
      "loss": 0.1167,
      "step": 7100
    },
    {
      "epoch": 8.470588235294118,
      "grad_norm": 0.9566899538040161,
      "learning_rate": 7.65294117647059e-06,
      "loss": 0.1552,
      "step": 7200
    },
    {
      "epoch": 8.588235294117647,
      "grad_norm": 0.0542168989777565,
      "learning_rate": 7.064705882352941e-06,
      "loss": 0.1505,
      "step": 7300
    },
    {
      "epoch": 8.705882352941176,
      "grad_norm": 20.501676559448242,
      "learning_rate": 6.476470588235295e-06,
      "loss": 0.1435,
      "step": 7400
    },
    {
      "epoch": 8.823529411764707,
      "grad_norm": 24.469806671142578,
      "learning_rate": 5.888235294117647e-06,
      "loss": 0.1411,
      "step": 7500
    },
    {
      "epoch": 8.941176470588236,
      "grad_norm": 0.4506582021713257,
      "learning_rate": 5.3e-06,
      "loss": 0.1091,
      "step": 7600
    },
    {
      "epoch": 9.058823529411764,
      "grad_norm": 0.023853685706853867,
      "learning_rate": 4.711764705882353e-06,
      "loss": 0.1062,
      "step": 7700
    },
    {
      "epoch": 9.176470588235293,
      "grad_norm": 0.06553351134061813,
      "learning_rate": 4.123529411764706e-06,
      "loss": 0.0913,
      "step": 7800
    },
    {
      "epoch": 9.294117647058824,
      "grad_norm": 0.1247912049293518,
      "learning_rate": 3.535294117647059e-06,
      "loss": 0.0952,
      "step": 7900
    },
    {
      "epoch": 9.411764705882353,
      "grad_norm": 0.03167188912630081,
      "learning_rate": 2.9470588235294116e-06,
      "loss": 0.1452,
      "step": 8000
    },
    {
      "epoch": 9.529411764705882,
      "grad_norm": 0.014582249335944653,
      "learning_rate": 2.3588235294117646e-06,
      "loss": 0.0852,
      "step": 8100
    },
    {
      "epoch": 9.647058823529411,
      "grad_norm": 0.022731494158506393,
      "learning_rate": 1.7705882352941176e-06,
      "loss": 0.1289,
      "step": 8200
    },
    {
      "epoch": 9.764705882352942,
      "grad_norm": 0.9212973713874817,
      "learning_rate": 1.1823529411764708e-06,
      "loss": 0.1184,
      "step": 8300
    },
    {
      "epoch": 9.882352941176471,
      "grad_norm": 0.4289371967315674,
      "learning_rate": 5.941176470588236e-07,
      "loss": 0.0689,
      "step": 8400
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.018015075474977493,
      "learning_rate": 5.882352941176471e-09,
      "loss": 0.119,
      "step": 8500
    },
    {
      "epoch": 10.0,
      "step": 8500,
      "total_flos": 5.269455293792256e+18,
      "train_loss": 0.24702390199549057,
      "train_runtime": 4430.2336,
      "train_samples_per_second": 15.349,
      "train_steps_per_second": 1.919
    }
  ],
  "logging_steps": 100,
  "max_steps": 8500,
  "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": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.269455293792256e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}