{
  "best_metric": 0.4118640422821045,
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le4_s4000/checkpoint-3000",
  "epoch": 66.66666666666667,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 1.5478250980377197,
      "learning_rate": 2.5e-06,
      "loss": 0.7933,
      "step": 50
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 2.5992791652679443,
      "learning_rate": 5e-06,
      "loss": 0.7126,
      "step": 100
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.512153625488281,
      "learning_rate": 7.45e-06,
      "loss": 0.6643,
      "step": 150
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 5.529618740081787,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.6161,
      "step": 200
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 1.547269344329834,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.5572,
      "step": 250
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 1.7310105562210083,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.5307,
      "step": 300
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 4.040494441986084,
      "learning_rate": 1.745e-05,
      "loss": 0.5259,
      "step": 350
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 1.6819641590118408,
      "learning_rate": 1.995e-05,
      "loss": 0.4981,
      "step": 400
    },
    {
      "epoch": 10.0,
      "grad_norm": 3.428649663925171,
      "learning_rate": 2.245e-05,
      "loss": 0.5008,
      "step": 450
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 2.2339069843292236,
      "learning_rate": 2.495e-05,
      "loss": 0.4923,
      "step": 500
    },
    {
      "epoch": 11.11111111111111,
      "eval_loss": 0.445181667804718,
      "eval_runtime": 8.1426,
      "eval_samples_per_second": 19.527,
      "eval_steps_per_second": 2.456,
      "step": 500
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 3.6943907737731934,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.4999,
      "step": 550
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 2.8963587284088135,
      "learning_rate": 2.995e-05,
      "loss": 0.4723,
      "step": 600
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 2.157615900039673,
      "learning_rate": 3.245e-05,
      "loss": 0.4771,
      "step": 650
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 4.2828450202941895,
      "learning_rate": 3.495e-05,
      "loss": 0.4656,
      "step": 700
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 2.30914044380188,
      "learning_rate": 3.745e-05,
      "loss": 0.4681,
      "step": 750
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 1.7412217855453491,
      "learning_rate": 3.995e-05,
      "loss": 0.4639,
      "step": 800
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 2.457113742828369,
      "learning_rate": 4.245e-05,
      "loss": 0.4691,
      "step": 850
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.960203170776367,
      "learning_rate": 4.495e-05,
      "loss": 0.4569,
      "step": 900
    },
    {
      "epoch": 21.11111111111111,
      "grad_norm": 3.9787750244140625,
      "learning_rate": 4.745e-05,
      "loss": 0.4547,
      "step": 950
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 1.6046252250671387,
      "learning_rate": 4.995e-05,
      "loss": 0.4488,
      "step": 1000
    },
    {
      "epoch": 22.22222222222222,
      "eval_loss": 0.42208045721054077,
      "eval_runtime": 7.5792,
      "eval_samples_per_second": 20.978,
      "eval_steps_per_second": 2.639,
      "step": 1000
    },
    {
      "epoch": 23.333333333333332,
      "grad_norm": 2.2585225105285645,
      "learning_rate": 5.245e-05,
      "loss": 0.4537,
      "step": 1050
    },
    {
      "epoch": 24.444444444444443,
      "grad_norm": 2.7472567558288574,
      "learning_rate": 5.495e-05,
      "loss": 0.4438,
      "step": 1100
    },
    {
      "epoch": 25.555555555555557,
      "grad_norm": 2.1399290561676025,
      "learning_rate": 5.745e-05,
      "loss": 0.4474,
      "step": 1150
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 2.671912670135498,
      "learning_rate": 5.995000000000001e-05,
      "loss": 0.4488,
      "step": 1200
    },
    {
      "epoch": 27.77777777777778,
      "grad_norm": 4.835938453674316,
      "learning_rate": 6.245000000000001e-05,
      "loss": 0.4472,
      "step": 1250
    },
    {
      "epoch": 28.88888888888889,
      "grad_norm": 5.086827754974365,
      "learning_rate": 6.494999999999999e-05,
      "loss": 0.4526,
      "step": 1300
    },
    {
      "epoch": 30.0,
      "grad_norm": 2.293753147125244,
      "learning_rate": 6.745e-05,
      "loss": 0.4448,
      "step": 1350
    },
    {
      "epoch": 31.11111111111111,
      "grad_norm": 3.609741449356079,
      "learning_rate": 6.995e-05,
      "loss": 0.438,
      "step": 1400
    },
    {
      "epoch": 32.22222222222222,
      "grad_norm": 1.580810546875,
      "learning_rate": 7.245000000000001e-05,
      "loss": 0.4446,
      "step": 1450
    },
    {
      "epoch": 33.333333333333336,
      "grad_norm": 2.7843821048736572,
      "learning_rate": 7.495e-05,
      "loss": 0.4413,
      "step": 1500
    },
    {
      "epoch": 33.333333333333336,
      "eval_loss": 0.41924577951431274,
      "eval_runtime": 7.8331,
      "eval_samples_per_second": 20.298,
      "eval_steps_per_second": 2.553,
      "step": 1500
    },
    {
      "epoch": 34.44444444444444,
      "grad_norm": 5.7151875495910645,
      "learning_rate": 7.745e-05,
      "loss": 0.4369,
      "step": 1550
    },
    {
      "epoch": 35.55555555555556,
      "grad_norm": 2.420909881591797,
      "learning_rate": 7.995e-05,
      "loss": 0.4491,
      "step": 1600
    },
    {
      "epoch": 36.666666666666664,
      "grad_norm": 3.388465642929077,
      "learning_rate": 8.245e-05,
      "loss": 0.4381,
      "step": 1650
    },
    {
      "epoch": 37.77777777777778,
      "grad_norm": 1.8416664600372314,
      "learning_rate": 8.495e-05,
      "loss": 0.4332,
      "step": 1700
    },
    {
      "epoch": 38.888888888888886,
      "grad_norm": 4.991604328155518,
      "learning_rate": 8.745000000000001e-05,
      "loss": 0.4363,
      "step": 1750
    },
    {
      "epoch": 40.0,
      "grad_norm": 1.6002848148345947,
      "learning_rate": 8.995e-05,
      "loss": 0.4296,
      "step": 1800
    },
    {
      "epoch": 41.111111111111114,
      "grad_norm": 2.4206886291503906,
      "learning_rate": 9.245e-05,
      "loss": 0.4273,
      "step": 1850
    },
    {
      "epoch": 42.22222222222222,
      "grad_norm": 3.848668098449707,
      "learning_rate": 9.495e-05,
      "loss": 0.4244,
      "step": 1900
    },
    {
      "epoch": 43.333333333333336,
      "grad_norm": 3.4843313694000244,
      "learning_rate": 9.745000000000001e-05,
      "loss": 0.4213,
      "step": 1950
    },
    {
      "epoch": 44.44444444444444,
      "grad_norm": 7.2384772300720215,
      "learning_rate": 9.995e-05,
      "loss": 0.4318,
      "step": 2000
    },
    {
      "epoch": 44.44444444444444,
      "eval_loss": 0.42309698462486267,
      "eval_runtime": 7.2889,
      "eval_samples_per_second": 21.814,
      "eval_steps_per_second": 2.744,
      "step": 2000
    },
    {
      "epoch": 45.55555555555556,
      "grad_norm": 1.936278223991394,
      "learning_rate": 9.755000000000001e-05,
      "loss": 0.4279,
      "step": 2050
    },
    {
      "epoch": 46.666666666666664,
      "grad_norm": 6.694615840911865,
      "learning_rate": 9.505e-05,
      "loss": 0.4326,
      "step": 2100
    },
    {
      "epoch": 47.77777777777778,
      "grad_norm": 3.927391529083252,
      "learning_rate": 9.260000000000001e-05,
      "loss": 0.434,
      "step": 2150
    },
    {
      "epoch": 48.888888888888886,
      "grad_norm": 2.543116569519043,
      "learning_rate": 9.010000000000001e-05,
      "loss": 0.4338,
      "step": 2200
    },
    {
      "epoch": 50.0,
      "grad_norm": 3.7880661487579346,
      "learning_rate": 8.76e-05,
      "loss": 0.4239,
      "step": 2250
    },
    {
      "epoch": 51.111111111111114,
      "grad_norm": 6.448720455169678,
      "learning_rate": 8.510000000000001e-05,
      "loss": 0.423,
      "step": 2300
    },
    {
      "epoch": 52.22222222222222,
      "grad_norm": 2.7751946449279785,
      "learning_rate": 8.26e-05,
      "loss": 0.4182,
      "step": 2350
    },
    {
      "epoch": 53.333333333333336,
      "grad_norm": 3.261136054992676,
      "learning_rate": 8.010000000000001e-05,
      "loss": 0.4133,
      "step": 2400
    },
    {
      "epoch": 54.44444444444444,
      "grad_norm": 3.6887683868408203,
      "learning_rate": 7.76e-05,
      "loss": 0.4096,
      "step": 2450
    },
    {
      "epoch": 55.55555555555556,
      "grad_norm": 2.9640533924102783,
      "learning_rate": 7.510000000000001e-05,
      "loss": 0.4155,
      "step": 2500
    },
    {
      "epoch": 55.55555555555556,
      "eval_loss": 0.41808074712753296,
      "eval_runtime": 7.7845,
      "eval_samples_per_second": 20.425,
      "eval_steps_per_second": 2.569,
      "step": 2500
    },
    {
      "epoch": 56.666666666666664,
      "grad_norm": 2.4800713062286377,
      "learning_rate": 7.26e-05,
      "loss": 0.4101,
      "step": 2550
    },
    {
      "epoch": 57.77777777777778,
      "grad_norm": 4.109247207641602,
      "learning_rate": 7.01e-05,
      "loss": 0.4087,
      "step": 2600
    },
    {
      "epoch": 58.888888888888886,
      "grad_norm": 2.4882519245147705,
      "learning_rate": 6.76e-05,
      "loss": 0.4022,
      "step": 2650
    },
    {
      "epoch": 60.0,
      "grad_norm": 2.2635858058929443,
      "learning_rate": 6.510000000000001e-05,
      "loss": 0.4176,
      "step": 2700
    },
    {
      "epoch": 61.111111111111114,
      "grad_norm": 5.780401706695557,
      "learning_rate": 6.26e-05,
      "loss": 0.4194,
      "step": 2750
    },
    {
      "epoch": 62.22222222222222,
      "grad_norm": 1.2802194356918335,
      "learning_rate": 6.0100000000000004e-05,
      "loss": 0.3982,
      "step": 2800
    },
    {
      "epoch": 63.333333333333336,
      "grad_norm": 1.1778165102005005,
      "learning_rate": 5.76e-05,
      "loss": 0.3994,
      "step": 2850
    },
    {
      "epoch": 64.44444444444444,
      "grad_norm": 2.8410191535949707,
      "learning_rate": 5.5100000000000004e-05,
      "loss": 0.3988,
      "step": 2900
    },
    {
      "epoch": 65.55555555555556,
      "grad_norm": 1.2655699253082275,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 0.4005,
      "step": 2950
    },
    {
      "epoch": 66.66666666666667,
      "grad_norm": 2.0291759967803955,
      "learning_rate": 5.0100000000000005e-05,
      "loss": 0.3961,
      "step": 3000
    },
    {
      "epoch": 66.66666666666667,
      "eval_loss": 0.4118640422821045,
      "eval_runtime": 6.7013,
      "eval_samples_per_second": 23.727,
      "eval_steps_per_second": 2.984,
      "step": 3000
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 89,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.066579214506611e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}