{
  "best_metric": 0.8584259152412415,
  "best_model_checkpoint": "data/Gemma-2-2B_task-2_180-samples_config-2_full_auto/checkpoint-76",
  "epoch": 16.0,
  "eval_steps": 500,
  "global_step": 136,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.4396458864212036,
      "learning_rate": 2.5e-06,
      "loss": 1.3119,
      "step": 1
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.46355336904525757,
      "learning_rate": 5e-06,
      "loss": 1.3371,
      "step": 2
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.42811518907546997,
      "learning_rate": 1e-05,
      "loss": 1.3332,
      "step": 4
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 0.4657036364078522,
      "learning_rate": 1.5e-05,
      "loss": 1.2805,
      "step": 6
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.39682307839393616,
      "learning_rate": 2e-05,
      "loss": 1.3006,
      "step": 8
    },
    {
      "epoch": 0.9411764705882353,
      "eval_loss": 1.2976187467575073,
      "eval_runtime": 14.3225,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 8
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.4378485083580017,
      "learning_rate": 2.5e-05,
      "loss": 1.3197,
      "step": 10
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 0.3689347505569458,
      "learning_rate": 3e-05,
      "loss": 1.2581,
      "step": 12
    },
    {
      "epoch": 1.6470588235294117,
      "grad_norm": 0.34312155842781067,
      "learning_rate": 3.5e-05,
      "loss": 1.2346,
      "step": 14
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 0.32624560594558716,
      "learning_rate": 4e-05,
      "loss": 1.2057,
      "step": 16
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.15829336643219,
      "eval_runtime": 14.3185,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 17
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 0.26776188611984253,
      "learning_rate": 4.5e-05,
      "loss": 1.1749,
      "step": 18
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.22213858366012573,
      "learning_rate": 5e-05,
      "loss": 1.1029,
      "step": 20
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 0.2549532949924469,
      "learning_rate": 5.500000000000001e-05,
      "loss": 1.1084,
      "step": 22
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.27232280373573303,
      "learning_rate": 6e-05,
      "loss": 1.1024,
      "step": 24
    },
    {
      "epoch": 2.9411764705882355,
      "eval_loss": 1.0637497901916504,
      "eval_runtime": 14.3197,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 25
    },
    {
      "epoch": 3.0588235294117645,
      "grad_norm": 0.2522396445274353,
      "learning_rate": 6.500000000000001e-05,
      "loss": 1.0387,
      "step": 26
    },
    {
      "epoch": 3.2941176470588234,
      "grad_norm": 0.21561956405639648,
      "learning_rate": 7e-05,
      "loss": 1.0127,
      "step": 28
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.23440401256084442,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.004,
      "step": 30
    },
    {
      "epoch": 3.764705882352941,
      "grad_norm": 0.2635060250759125,
      "learning_rate": 8e-05,
      "loss": 0.9906,
      "step": 32
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.25140380859375,
      "learning_rate": 8.5e-05,
      "loss": 0.9139,
      "step": 34
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.9544374942779541,
      "eval_runtime": 14.3239,
      "eval_samples_per_second": 2.513,
      "eval_steps_per_second": 2.513,
      "step": 34
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 0.20246362686157227,
      "learning_rate": 9e-05,
      "loss": 0.9272,
      "step": 36
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 0.1751442700624466,
      "learning_rate": 9.5e-05,
      "loss": 0.8889,
      "step": 38
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.1525234431028366,
      "learning_rate": 0.0001,
      "loss": 0.8834,
      "step": 40
    },
    {
      "epoch": 4.9411764705882355,
      "grad_norm": 0.16081523895263672,
      "learning_rate": 9.999238475781957e-05,
      "loss": 0.8515,
      "step": 42
    },
    {
      "epoch": 4.9411764705882355,
      "eval_loss": 0.9081065058708191,
      "eval_runtime": 14.3093,
      "eval_samples_per_second": 2.516,
      "eval_steps_per_second": 2.516,
      "step": 42
    },
    {
      "epoch": 5.176470588235294,
      "grad_norm": 0.15286272764205933,
      "learning_rate": 9.99695413509548e-05,
      "loss": 0.8256,
      "step": 44
    },
    {
      "epoch": 5.411764705882353,
      "grad_norm": 0.134580597281456,
      "learning_rate": 9.99314767377287e-05,
      "loss": 0.861,
      "step": 46
    },
    {
      "epoch": 5.647058823529412,
      "grad_norm": 0.1375444233417511,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.8197,
      "step": 48
    },
    {
      "epoch": 5.882352941176471,
      "grad_norm": 0.15164963901042938,
      "learning_rate": 9.980973490458728e-05,
      "loss": 0.863,
      "step": 50
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.8835119605064392,
      "eval_runtime": 14.3145,
      "eval_samples_per_second": 2.515,
      "eval_steps_per_second": 2.515,
      "step": 51
    },
    {
      "epoch": 6.117647058823529,
      "grad_norm": 0.1275659203529358,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.8009,
      "step": 52
    },
    {
      "epoch": 6.352941176470588,
      "grad_norm": 0.12862518429756165,
      "learning_rate": 9.962730758206611e-05,
      "loss": 0.7995,
      "step": 54
    },
    {
      "epoch": 6.588235294117647,
      "grad_norm": 0.13512417674064636,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.8078,
      "step": 56
    },
    {
      "epoch": 6.823529411764706,
      "grad_norm": 0.12700602412223816,
      "learning_rate": 9.938441702975689e-05,
      "loss": 0.8439,
      "step": 58
    },
    {
      "epoch": 6.9411764705882355,
      "eval_loss": 0.8702912926673889,
      "eval_runtime": 14.3215,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 59
    },
    {
      "epoch": 7.0588235294117645,
      "grad_norm": 0.1407477855682373,
      "learning_rate": 9.924038765061042e-05,
      "loss": 0.8221,
      "step": 60
    },
    {
      "epoch": 7.294117647058823,
      "grad_norm": 0.13812792301177979,
      "learning_rate": 9.908135917238321e-05,
      "loss": 0.7923,
      "step": 62
    },
    {
      "epoch": 7.529411764705882,
      "grad_norm": 0.1376505196094513,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.7691,
      "step": 64
    },
    {
      "epoch": 7.764705882352941,
      "grad_norm": 0.14402402937412262,
      "learning_rate": 9.871850323926177e-05,
      "loss": 0.7609,
      "step": 66
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.15684427320957184,
      "learning_rate": 9.851478631379982e-05,
      "loss": 0.7927,
      "step": 68
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.8620401620864868,
      "eval_runtime": 14.314,
      "eval_samples_per_second": 2.515,
      "eval_steps_per_second": 2.515,
      "step": 68
    },
    {
      "epoch": 8.235294117647058,
      "grad_norm": 0.14633484184741974,
      "learning_rate": 9.829629131445342e-05,
      "loss": 0.7506,
      "step": 70
    },
    {
      "epoch": 8.470588235294118,
      "grad_norm": 0.14848043024539948,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.7807,
      "step": 72
    },
    {
      "epoch": 8.705882352941176,
      "grad_norm": 0.14178834855556488,
      "learning_rate": 9.781523779815179e-05,
      "loss": 0.7781,
      "step": 74
    },
    {
      "epoch": 8.941176470588236,
      "grad_norm": 0.1646076738834381,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.7343,
      "step": 76
    },
    {
      "epoch": 8.941176470588236,
      "eval_loss": 0.8584259152412415,
      "eval_runtime": 14.3158,
      "eval_samples_per_second": 2.515,
      "eval_steps_per_second": 2.515,
      "step": 76
    },
    {
      "epoch": 9.176470588235293,
      "grad_norm": 0.15060319006443024,
      "learning_rate": 9.727592877996585e-05,
      "loss": 0.7412,
      "step": 78
    },
    {
      "epoch": 9.411764705882353,
      "grad_norm": 0.1598275899887085,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.7429,
      "step": 80
    },
    {
      "epoch": 9.647058823529411,
      "grad_norm": 0.17243456840515137,
      "learning_rate": 9.667902132486009e-05,
      "loss": 0.7363,
      "step": 82
    },
    {
      "epoch": 9.882352941176471,
      "grad_norm": 0.18578656017780304,
      "learning_rate": 9.635919272833938e-05,
      "loss": 0.7243,
      "step": 84
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.8590600490570068,
      "eval_runtime": 14.313,
      "eval_samples_per_second": 2.515,
      "eval_steps_per_second": 2.515,
      "step": 85
    },
    {
      "epoch": 10.117647058823529,
      "grad_norm": 0.18428955972194672,
      "learning_rate": 9.602524267262203e-05,
      "loss": 0.7048,
      "step": 86
    },
    {
      "epoch": 10.352941176470589,
      "grad_norm": 0.2186216115951538,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.7212,
      "step": 88
    },
    {
      "epoch": 10.588235294117647,
      "grad_norm": 0.19397258758544922,
      "learning_rate": 9.53153893518325e-05,
      "loss": 0.6943,
      "step": 90
    },
    {
      "epoch": 10.823529411764707,
      "grad_norm": 0.2317022830247879,
      "learning_rate": 9.493970231495835e-05,
      "loss": 0.6957,
      "step": 92
    },
    {
      "epoch": 10.941176470588236,
      "eval_loss": 0.8660648465156555,
      "eval_runtime": 14.3176,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 93
    },
    {
      "epoch": 11.058823529411764,
      "grad_norm": 0.209140807390213,
      "learning_rate": 9.45503262094184e-05,
      "loss": 0.7243,
      "step": 94
    },
    {
      "epoch": 11.294117647058824,
      "grad_norm": 0.22573913633823395,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.7136,
      "step": 96
    },
    {
      "epoch": 11.529411764705882,
      "grad_norm": 0.23610715568065643,
      "learning_rate": 9.373098535696979e-05,
      "loss": 0.649,
      "step": 98
    },
    {
      "epoch": 11.764705882352942,
      "grad_norm": 0.26594263315200806,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.6808,
      "step": 100
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.272132009267807,
      "learning_rate": 9.285836503510562e-05,
      "loss": 0.6372,
      "step": 102
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.8780656456947327,
      "eval_runtime": 14.3181,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 102
    },
    {
      "epoch": 12.235294117647058,
      "grad_norm": 0.29550257325172424,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.6356,
      "step": 104
    },
    {
      "epoch": 12.470588235294118,
      "grad_norm": 0.3006047308444977,
      "learning_rate": 9.193352839727121e-05,
      "loss": 0.6283,
      "step": 106
    },
    {
      "epoch": 12.705882352941176,
      "grad_norm": 0.3633236587047577,
      "learning_rate": 9.145187862775209e-05,
      "loss": 0.6436,
      "step": 108
    },
    {
      "epoch": 12.941176470588236,
      "grad_norm": 0.3862960636615753,
      "learning_rate": 9.09576022144496e-05,
      "loss": 0.6342,
      "step": 110
    },
    {
      "epoch": 12.941176470588236,
      "eval_loss": 0.8992895483970642,
      "eval_runtime": 14.321,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 110
    },
    {
      "epoch": 13.176470588235293,
      "grad_norm": 0.3040759861469269,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.6252,
      "step": 112
    },
    {
      "epoch": 13.411764705882353,
      "grad_norm": 0.37966129183769226,
      "learning_rate": 8.993177550236464e-05,
      "loss": 0.5624,
      "step": 114
    },
    {
      "epoch": 13.647058823529411,
      "grad_norm": 0.4189399182796478,
      "learning_rate": 8.940053768033609e-05,
      "loss": 0.5971,
      "step": 116
    },
    {
      "epoch": 13.882352941176471,
      "grad_norm": 0.4090847969055176,
      "learning_rate": 8.885729807284856e-05,
      "loss": 0.6126,
      "step": 118
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.9277560114860535,
      "eval_runtime": 14.3205,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 119
    },
    {
      "epoch": 14.117647058823529,
      "grad_norm": 0.3958241939544678,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.6097,
      "step": 120
    },
    {
      "epoch": 14.352941176470589,
      "grad_norm": 0.5418867468833923,
      "learning_rate": 8.773547901113862e-05,
      "loss": 0.514,
      "step": 122
    },
    {
      "epoch": 14.588235294117647,
      "grad_norm": 0.49185261130332947,
      "learning_rate": 8.715724127386972e-05,
      "loss": 0.579,
      "step": 124
    },
    {
      "epoch": 14.823529411764707,
      "grad_norm": 0.51872718334198,
      "learning_rate": 8.656768508095853e-05,
      "loss": 0.5304,
      "step": 126
    },
    {
      "epoch": 14.941176470588236,
      "eval_loss": 0.9825281500816345,
      "eval_runtime": 14.319,
      "eval_samples_per_second": 2.514,
      "eval_steps_per_second": 2.514,
      "step": 127
    },
    {
      "epoch": 15.058823529411764,
      "grad_norm": 0.5219426155090332,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.5779,
      "step": 128
    },
    {
      "epoch": 15.294117647058824,
      "grad_norm": 0.5088827013969421,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.5208,
      "step": 130
    },
    {
      "epoch": 15.529411764705882,
      "grad_norm": 0.5148976445198059,
      "learning_rate": 8.473291852294987e-05,
      "loss": 0.5362,
      "step": 132
    },
    {
      "epoch": 15.764705882352942,
      "grad_norm": 0.5440108180046082,
      "learning_rate": 8.409991800312493e-05,
      "loss": 0.5373,
      "step": 134
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.6404002904891968,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.4754,
      "step": 136
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.0304640531539917,
      "eval_runtime": 14.3147,
      "eval_samples_per_second": 2.515,
      "eval_steps_per_second": 2.515,
      "step": 136
    },
    {
      "epoch": 16.0,
      "step": 136,
      "total_flos": 7.332060624584704e+16,
      "train_loss": 0.8212331099545255,
      "train_runtime": 3120.3597,
      "train_samples_per_second": 2.179,
      "train_steps_per_second": 0.128
    }
  ],
  "logging_steps": 2,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 25,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 7,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.332060624584704e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}