{
  "best_metric": 0.8514555096626282,
  "best_model_checkpoint": "./output/training_results/C016_random_sample_llama3-8b-base_instruct_20240504_181744/checkpoint-40",
  "epoch": 4.0,
  "eval_steps": 20,
  "global_step": 192,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020833333333333332,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 0.9711,
      "step": 1
    },
    {
      "epoch": 0.10416666666666667,
      "grad_norm": 12.43094855680643,
      "learning_rate": 2.25e-06,
      "loss": 0.949,
      "step": 5
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 4.621787623957677,
      "learning_rate": 5.25e-06,
      "loss": 0.8756,
      "step": 10
    },
    {
      "epoch": 0.3125,
      "grad_norm": 5.102324743913745,
      "learning_rate": 9e-06,
      "loss": 0.8441,
      "step": 15
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 5.046721891827742,
      "learning_rate": 1.275e-05,
      "loss": 0.8263,
      "step": 20
    },
    {
      "epoch": 0.4166666666666667,
      "eval_loss": 0.8584555983543396,
      "eval_runtime": 2.0011,
      "eval_samples_per_second": 169.909,
      "eval_steps_per_second": 1.499,
      "step": 20
    },
    {
      "epoch": 0.5208333333333334,
      "grad_norm": 4.324316961991125,
      "learning_rate": 1.3195176200175283e-05,
      "loss": 0.8289,
      "step": 25
    },
    {
      "epoch": 0.625,
      "grad_norm": 3.9781224609845345,
      "learning_rate": 9.515676612044427e-06,
      "loss": 0.9015,
      "step": 30
    },
    {
      "epoch": 0.7291666666666666,
      "grad_norm": 3.851416633553399,
      "learning_rate": 6.797580677308734e-06,
      "loss": 0.823,
      "step": 35
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 4.17708940737461,
      "learning_rate": 4.808575415542887e-06,
      "loss": 0.8014,
      "step": 40
    },
    {
      "epoch": 0.8333333333333334,
      "eval_loss": 0.8514555096626282,
      "eval_runtime": 1.9698,
      "eval_samples_per_second": 172.604,
      "eval_steps_per_second": 1.523,
      "step": 40
    },
    {
      "epoch": 0.9375,
      "grad_norm": 3.865023549249989,
      "learning_rate": 3.3676619069852654e-06,
      "loss": 0.8814,
      "step": 45
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 3.6024413325254803,
      "learning_rate": 2.334947896124909e-06,
      "loss": 0.7239,
      "step": 50
    },
    {
      "epoch": 1.1458333333333333,
      "grad_norm": 2.9479655076003395,
      "learning_rate": 1.603233215095547e-06,
      "loss": 0.466,
      "step": 55
    },
    {
      "epoch": 1.25,
      "grad_norm": 3.8904458700460633,
      "learning_rate": 1.0911174606561334e-06,
      "loss": 0.4375,
      "step": 60
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.871760368347168,
      "eval_runtime": 1.9682,
      "eval_samples_per_second": 172.743,
      "eval_steps_per_second": 1.524,
      "step": 60
    },
    {
      "epoch": 1.3541666666666667,
      "grad_norm": 3.7825224108872644,
      "learning_rate": 7.373930741131784e-07,
      "loss": 0.3976,
      "step": 65
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 3.9946225754826044,
      "learning_rate": 4.965174334325768e-07,
      "loss": 0.4467,
      "step": 70
    },
    {
      "epoch": 1.5625,
      "grad_norm": 4.437633052742651,
      "learning_rate": 3.349849877937343e-07,
      "loss": 0.4382,
      "step": 75
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 3.8225073330864148,
      "learning_rate": 2.2844505627726646e-07,
      "loss": 0.4593,
      "step": 80
    },
    {
      "epoch": 1.6666666666666665,
      "eval_loss": 0.855841338634491,
      "eval_runtime": 1.9784,
      "eval_samples_per_second": 171.854,
      "eval_steps_per_second": 1.516,
      "step": 80
    },
    {
      "epoch": 1.7708333333333335,
      "grad_norm": 3.2403588491446373,
      "learning_rate": 1.594328760942437e-07,
      "loss": 0.4262,
      "step": 85
    },
    {
      "epoch": 1.875,
      "grad_norm": 3.3007958768222556,
      "learning_rate": 1.156010161291434e-07,
      "loss": 0.4236,
      "step": 90
    },
    {
      "epoch": 1.9791666666666665,
      "grad_norm": 3.7194840629840744,
      "learning_rate": 8.835555547373544e-08,
      "loss": 0.4859,
      "step": 95
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 3.4233566225496928,
      "learning_rate": 7.181664349277562e-08,
      "loss": 0.3969,
      "step": 100
    },
    {
      "epoch": 2.0833333333333335,
      "eval_loss": 0.8527613282203674,
      "eval_runtime": 1.973,
      "eval_samples_per_second": 172.329,
      "eval_steps_per_second": 1.521,
      "step": 100
    },
    {
      "epoch": 2.1875,
      "grad_norm": 3.255038919801069,
      "learning_rate": 6.203637972657601e-08,
      "loss": 0.3693,
      "step": 105
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 3.1973516190995497,
      "learning_rate": 5.6418543066491835e-08,
      "loss": 0.4039,
      "step": 110
    },
    {
      "epoch": 2.3958333333333335,
      "grad_norm": 3.463654018617807,
      "learning_rate": 5.329471712759216e-08,
      "loss": 0.3573,
      "step": 115
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.418086411896532,
      "learning_rate": 5.161995210302015e-08,
      "loss": 0.3982,
      "step": 120
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.8576084971427917,
      "eval_runtime": 1.9748,
      "eval_samples_per_second": 172.17,
      "eval_steps_per_second": 1.519,
      "step": 120
    },
    {
      "epoch": 2.6041666666666665,
      "grad_norm": 3.4030059443316723,
      "learning_rate": 5.075841465580837e-08,
      "loss": 0.3815,
      "step": 125
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 3.7277312294575444,
      "learning_rate": 5.033564114946932e-08,
      "loss": 0.3762,
      "step": 130
    },
    {
      "epoch": 2.8125,
      "grad_norm": 3.3451733690112104,
      "learning_rate": 5.013915282607116e-08,
      "loss": 0.3809,
      "step": 135
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 3.217687429197331,
      "learning_rate": 5.005343402153039e-08,
      "loss": 0.3742,
      "step": 140
    },
    {
      "epoch": 2.9166666666666665,
      "eval_loss": 0.8624204993247986,
      "eval_runtime": 1.974,
      "eval_samples_per_second": 172.243,
      "eval_steps_per_second": 1.52,
      "step": 140
    },
    {
      "epoch": 3.0208333333333335,
      "grad_norm": 3.0189998840641734,
      "learning_rate": 5.001872829857116e-08,
      "loss": 0.3793,
      "step": 145
    },
    {
      "epoch": 3.125,
      "grad_norm": 3.1862692972750577,
      "learning_rate": 5.000587713853837e-08,
      "loss": 0.3734,
      "step": 150
    },
    {
      "epoch": 3.2291666666666665,
      "grad_norm": 3.179336124252825,
      "learning_rate": 5.0001608748597456e-08,
      "loss": 0.3713,
      "step": 155
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.66026630563327,
      "learning_rate": 5.0000370319656156e-08,
      "loss": 0.3692,
      "step": 160
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 0.8662496209144592,
      "eval_runtime": 1.9702,
      "eval_samples_per_second": 172.572,
      "eval_steps_per_second": 1.523,
      "step": 160
    },
    {
      "epoch": 3.4375,
      "grad_norm": 3.286706597182968,
      "learning_rate": 5.0000067945715855e-08,
      "loss": 0.3657,
      "step": 165
    },
    {
      "epoch": 3.5416666666666665,
      "grad_norm": 3.4539000739999817,
      "learning_rate": 5.0000009144677036e-08,
      "loss": 0.3513,
      "step": 170
    },
    {
      "epoch": 3.6458333333333335,
      "grad_norm": 3.3076231790188904,
      "learning_rate": 5.0000000785521776e-08,
      "loss": 0.3683,
      "step": 175
    },
    {
      "epoch": 3.75,
      "grad_norm": 3.4401672686211464,
      "learning_rate": 5.000000003317662e-08,
      "loss": 0.3667,
      "step": 180
    },
    {
      "epoch": 3.75,
      "eval_loss": 0.8689968585968018,
      "eval_runtime": 1.9682,
      "eval_samples_per_second": 172.746,
      "eval_steps_per_second": 1.524,
      "step": 180
    },
    {
      "epoch": 3.8541666666666665,
      "grad_norm": 3.709318957772044,
      "learning_rate": 5.000000000038355e-08,
      "loss": 0.3741,
      "step": 185
    },
    {
      "epoch": 3.9583333333333335,
      "grad_norm": 3.1576885454409456,
      "learning_rate": 5.000000000000018e-08,
      "loss": 0.3694,
      "step": 190
    },
    {
      "epoch": 4.0,
      "step": 192,
      "total_flos": 5362900008960.0,
      "train_loss": 0.5138544160872698,
      "train_runtime": 1043.9459,
      "train_samples_per_second": 11.702,
      "train_steps_per_second": 0.184
    }
  ],
  "logging_steps": 5,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 20,
  "total_flos": 5362900008960.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}