{
  "best_metric": 0.43401163816452026,
  "best_model_checkpoint": "mikhail-panzo/fil_b64_le3_s4000/checkpoint-500",
  "epoch": 133.33333333333334,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 2.2609777450561523,
      "learning_rate": 2.5e-05,
      "loss": 0.7231,
      "step": 50
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 1.939948320388794,
      "learning_rate": 5e-05,
      "loss": 0.575,
      "step": 100
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 2.9677317142486572,
      "learning_rate": 7.5e-05,
      "loss": 0.5207,
      "step": 150
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 4.588547706604004,
      "learning_rate": 0.0001,
      "loss": 0.4996,
      "step": 200
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 1.8626277446746826,
      "learning_rate": 0.000125,
      "loss": 0.4881,
      "step": 250
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 3.4871737957000732,
      "learning_rate": 0.00015,
      "loss": 0.4729,
      "step": 300
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 3.5084946155548096,
      "learning_rate": 0.000175,
      "loss": 0.4661,
      "step": 350
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 7.88467264175415,
      "learning_rate": 0.0002,
      "loss": 0.4925,
      "step": 400
    },
    {
      "epoch": 20.0,
      "grad_norm": 4.796986103057861,
      "learning_rate": 0.00022500000000000002,
      "loss": 0.4899,
      "step": 450
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 6.419567584991455,
      "learning_rate": 0.00025,
      "loss": 0.4644,
      "step": 500
    },
    {
      "epoch": 22.22222222222222,
      "eval_loss": 0.43401163816452026,
      "eval_runtime": 17.3691,
      "eval_samples_per_second": 9.154,
      "eval_steps_per_second": 1.151,
      "step": 500
    },
    {
      "epoch": 24.444444444444443,
      "grad_norm": 2.6020901203155518,
      "learning_rate": 0.000275,
      "loss": 0.4492,
      "step": 550
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 7.353488445281982,
      "learning_rate": 0.0003,
      "loss": 0.4556,
      "step": 600
    },
    {
      "epoch": 28.88888888888889,
      "grad_norm": 4.736776351928711,
      "learning_rate": 0.00032500000000000004,
      "loss": 0.4669,
      "step": 650
    },
    {
      "epoch": 31.11111111111111,
      "grad_norm": 7.411691665649414,
      "learning_rate": 0.00035,
      "loss": 0.4892,
      "step": 700
    },
    {
      "epoch": 33.333333333333336,
      "grad_norm": 5.571273326873779,
      "learning_rate": 0.000375,
      "loss": 0.4858,
      "step": 750
    },
    {
      "epoch": 35.55555555555556,
      "grad_norm": 7.498672008514404,
      "learning_rate": 0.0004,
      "loss": 0.4929,
      "step": 800
    },
    {
      "epoch": 37.77777777777778,
      "grad_norm": 6.763402462005615,
      "learning_rate": 0.000425,
      "loss": 0.4874,
      "step": 850
    },
    {
      "epoch": 40.0,
      "grad_norm": 3.335564136505127,
      "learning_rate": 0.00045000000000000004,
      "loss": 0.513,
      "step": 900
    },
    {
      "epoch": 42.22222222222222,
      "grad_norm": 3.1004624366760254,
      "learning_rate": 0.000475,
      "loss": 0.4882,
      "step": 950
    },
    {
      "epoch": 44.44444444444444,
      "grad_norm": 10.330741882324219,
      "learning_rate": 0.0005,
      "loss": 0.6468,
      "step": 1000
    },
    {
      "epoch": 44.44444444444444,
      "eval_loss": 0.753709614276886,
      "eval_runtime": 16.7842,
      "eval_samples_per_second": 9.473,
      "eval_steps_per_second": 1.192,
      "step": 1000
    },
    {
      "epoch": 46.666666666666664,
      "grad_norm": 0.7254169583320618,
      "learning_rate": 0.0005245,
      "loss": 1.8403,
      "step": 1050
    },
    {
      "epoch": 48.888888888888886,
      "grad_norm": 0.14719372987747192,
      "learning_rate": 0.0005495,
      "loss": 1.6103,
      "step": 1100
    },
    {
      "epoch": 51.111111111111114,
      "grad_norm": 0.2177058607339859,
      "learning_rate": 0.0005745,
      "loss": 1.5878,
      "step": 1150
    },
    {
      "epoch": 53.333333333333336,
      "grad_norm": 0.20293082296848297,
      "learning_rate": 0.0005995000000000001,
      "loss": 1.5825,
      "step": 1200
    },
    {
      "epoch": 55.55555555555556,
      "grad_norm": 0.07884251326322556,
      "learning_rate": 0.0006245000000000001,
      "loss": 1.5827,
      "step": 1250
    },
    {
      "epoch": 57.77777777777778,
      "grad_norm": 0.2063368558883667,
      "learning_rate": 0.0006495,
      "loss": 1.5858,
      "step": 1300
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.21431827545166016,
      "learning_rate": 0.0006745,
      "loss": 1.5811,
      "step": 1350
    },
    {
      "epoch": 62.22222222222222,
      "grad_norm": 0.23930463194847107,
      "learning_rate": 0.0006995,
      "loss": 1.588,
      "step": 1400
    },
    {
      "epoch": 64.44444444444444,
      "grad_norm": 0.1440805047750473,
      "learning_rate": 0.0007245000000000001,
      "loss": 1.5773,
      "step": 1450
    },
    {
      "epoch": 66.66666666666667,
      "grad_norm": 0.18511304259300232,
      "learning_rate": 0.0007495000000000001,
      "loss": 1.5805,
      "step": 1500
    },
    {
      "epoch": 66.66666666666667,
      "eval_loss": 1.5452969074249268,
      "eval_runtime": 9.5395,
      "eval_samples_per_second": 16.668,
      "eval_steps_per_second": 2.097,
      "step": 1500
    },
    {
      "epoch": 68.88888888888889,
      "grad_norm": 0.1333579570055008,
      "learning_rate": 0.0007745,
      "loss": 1.5895,
      "step": 1550
    },
    {
      "epoch": 71.11111111111111,
      "grad_norm": 0.24500781297683716,
      "learning_rate": 0.0007995,
      "loss": 1.5823,
      "step": 1600
    },
    {
      "epoch": 73.33333333333333,
      "grad_norm": 0.16170234978199005,
      "learning_rate": 0.0008245,
      "loss": 1.5798,
      "step": 1650
    },
    {
      "epoch": 75.55555555555556,
      "grad_norm": 0.14667277038097382,
      "learning_rate": 0.0008495000000000001,
      "loss": 1.5756,
      "step": 1700
    },
    {
      "epoch": 77.77777777777777,
      "grad_norm": 0.12401594966650009,
      "learning_rate": 0.0008745000000000001,
      "loss": 1.5832,
      "step": 1750
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.10182458907365799,
      "learning_rate": 0.0008995,
      "loss": 1.5767,
      "step": 1800
    },
    {
      "epoch": 82.22222222222223,
      "grad_norm": 0.11051663756370544,
      "learning_rate": 0.0009245,
      "loss": 1.5745,
      "step": 1850
    },
    {
      "epoch": 84.44444444444444,
      "grad_norm": 0.2343149483203888,
      "learning_rate": 0.0009495,
      "loss": 1.5828,
      "step": 1900
    },
    {
      "epoch": 86.66666666666667,
      "grad_norm": 0.07052835822105408,
      "learning_rate": 0.0009745000000000001,
      "loss": 1.5747,
      "step": 1950
    },
    {
      "epoch": 88.88888888888889,
      "grad_norm": 0.044963087886571884,
      "learning_rate": 0.0009995000000000002,
      "loss": 1.5766,
      "step": 2000
    },
    {
      "epoch": 88.88888888888889,
      "eval_loss": 1.5454070568084717,
      "eval_runtime": 9.1426,
      "eval_samples_per_second": 17.391,
      "eval_steps_per_second": 2.188,
      "step": 2000
    },
    {
      "epoch": 91.11111111111111,
      "grad_norm": 0.13770775496959686,
      "learning_rate": 0.0009755,
      "loss": 1.583,
      "step": 2050
    },
    {
      "epoch": 93.33333333333333,
      "grad_norm": 0.07407082617282867,
      "learning_rate": 0.0009505000000000001,
      "loss": 1.5789,
      "step": 2100
    },
    {
      "epoch": 95.55555555555556,
      "grad_norm": 0.13063107430934906,
      "learning_rate": 0.0009255,
      "loss": 1.5762,
      "step": 2150
    },
    {
      "epoch": 97.77777777777777,
      "grad_norm": 0.09662608057260513,
      "learning_rate": 0.0009004999999999999,
      "loss": 1.5788,
      "step": 2200
    },
    {
      "epoch": 100.0,
      "grad_norm": 0.09555140137672424,
      "learning_rate": 0.0008755,
      "loss": 1.5794,
      "step": 2250
    },
    {
      "epoch": 102.22222222222223,
      "grad_norm": 0.08856567740440369,
      "learning_rate": 0.0008505,
      "loss": 1.5783,
      "step": 2300
    },
    {
      "epoch": 104.44444444444444,
      "grad_norm": 0.0869736298918724,
      "learning_rate": 0.0008255000000000001,
      "loss": 1.5769,
      "step": 2350
    },
    {
      "epoch": 106.66666666666667,
      "grad_norm": 0.19869808852672577,
      "learning_rate": 0.0008005,
      "loss": 1.5841,
      "step": 2400
    },
    {
      "epoch": 108.88888888888889,
      "grad_norm": 0.21654528379440308,
      "learning_rate": 0.0007754999999999999,
      "loss": 1.5781,
      "step": 2450
    },
    {
      "epoch": 111.11111111111111,
      "grad_norm": 0.06469714641571045,
      "learning_rate": 0.0007505,
      "loss": 1.5747,
      "step": 2500
    },
    {
      "epoch": 111.11111111111111,
      "eval_loss": 1.542811632156372,
      "eval_runtime": 9.2318,
      "eval_samples_per_second": 17.223,
      "eval_steps_per_second": 2.166,
      "step": 2500
    },
    {
      "epoch": 113.33333333333333,
      "grad_norm": 0.07496875524520874,
      "learning_rate": 0.0007255,
      "loss": 1.5798,
      "step": 2550
    },
    {
      "epoch": 115.55555555555556,
      "grad_norm": 0.04544850066304207,
      "learning_rate": 0.0007005000000000001,
      "loss": 1.575,
      "step": 2600
    },
    {
      "epoch": 117.77777777777777,
      "grad_norm": 0.09091202169656754,
      "learning_rate": 0.0006755,
      "loss": 1.5782,
      "step": 2650
    },
    {
      "epoch": 120.0,
      "grad_norm": 0.13279400765895844,
      "learning_rate": 0.0006504999999999999,
      "loss": 1.5919,
      "step": 2700
    },
    {
      "epoch": 122.22222222222223,
      "grad_norm": 0.11825087666511536,
      "learning_rate": 0.0006255,
      "loss": 1.5883,
      "step": 2750
    },
    {
      "epoch": 124.44444444444444,
      "grad_norm": 0.07335207611322403,
      "learning_rate": 0.0006005,
      "loss": 1.5781,
      "step": 2800
    },
    {
      "epoch": 126.66666666666667,
      "grad_norm": 0.10626611858606339,
      "learning_rate": 0.0005755000000000001,
      "loss": 1.5745,
      "step": 2850
    },
    {
      "epoch": 128.88888888888889,
      "grad_norm": 0.11508317291736603,
      "learning_rate": 0.0005505,
      "loss": 1.6894,
      "step": 2900
    },
    {
      "epoch": 131.11111111111111,
      "grad_norm": 0.10172197967767715,
      "learning_rate": 0.0005254999999999999,
      "loss": 1.5783,
      "step": 2950
    },
    {
      "epoch": 133.33333333333334,
      "grad_norm": 0.10468525439500809,
      "learning_rate": 0.0005005,
      "loss": 1.578,
      "step": 3000
    },
    {
      "epoch": 133.33333333333334,
      "eval_loss": 1.5455601215362549,
      "eval_runtime": 9.7139,
      "eval_samples_per_second": 16.368,
      "eval_steps_per_second": 2.059,
      "step": 3000
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 182,
  "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": 4.128668597841638e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}