{
  "best_metric": 0.41453465819358826,
  "best_model_checkpoint": "mikhail-panzo/fil_b128_le3_s4000/checkpoint-500",
  "epoch": 222.22222222222223,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.444444444444445,
      "grad_norm": 1.5276468992233276,
      "learning_rate": 2.5e-05,
      "loss": 0.7204,
      "step": 50
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 1.4177956581115723,
      "learning_rate": 5e-05,
      "loss": 0.5494,
      "step": 100
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 3.0381362438201904,
      "learning_rate": 7.5e-05,
      "loss": 0.4948,
      "step": 150
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 1.7661292552947998,
      "learning_rate": 0.0001,
      "loss": 0.4734,
      "step": 200
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 1.4447826147079468,
      "learning_rate": 0.000125,
      "loss": 0.473,
      "step": 250
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 2.745311975479126,
      "learning_rate": 0.00015,
      "loss": 0.464,
      "step": 300
    },
    {
      "epoch": 31.11111111111111,
      "grad_norm": 2.4395411014556885,
      "learning_rate": 0.000175,
      "loss": 0.4593,
      "step": 350
    },
    {
      "epoch": 35.55555555555556,
      "grad_norm": 2.1870596408843994,
      "learning_rate": 0.0002,
      "loss": 0.4452,
      "step": 400
    },
    {
      "epoch": 40.0,
      "grad_norm": 2.9709341526031494,
      "learning_rate": 0.00022500000000000002,
      "loss": 0.4404,
      "step": 450
    },
    {
      "epoch": 44.44444444444444,
      "grad_norm": 3.981621265411377,
      "learning_rate": 0.00025,
      "loss": 0.4367,
      "step": 500
    },
    {
      "epoch": 44.44444444444444,
      "eval_loss": 0.41453465819358826,
      "eval_runtime": 9.8143,
      "eval_samples_per_second": 16.201,
      "eval_steps_per_second": 2.038,
      "step": 500
    },
    {
      "epoch": 48.888888888888886,
      "grad_norm": 9.63549518585205,
      "learning_rate": 0.000275,
      "loss": 0.4377,
      "step": 550
    },
    {
      "epoch": 53.333333333333336,
      "grad_norm": 3.0573575496673584,
      "learning_rate": 0.0003,
      "loss": 0.4377,
      "step": 600
    },
    {
      "epoch": 57.77777777777778,
      "grad_norm": 2.144299268722534,
      "learning_rate": 0.00032500000000000004,
      "loss": 0.4342,
      "step": 650
    },
    {
      "epoch": 62.22222222222222,
      "grad_norm": 3.523078203201294,
      "learning_rate": 0.00035,
      "loss": 0.4375,
      "step": 700
    },
    {
      "epoch": 66.66666666666667,
      "grad_norm": 3.0409202575683594,
      "learning_rate": 0.000375,
      "loss": 0.4382,
      "step": 750
    },
    {
      "epoch": 71.11111111111111,
      "grad_norm": 7.679955959320068,
      "learning_rate": 0.0004,
      "loss": 0.4435,
      "step": 800
    },
    {
      "epoch": 75.55555555555556,
      "grad_norm": 4.8538618087768555,
      "learning_rate": 0.000425,
      "loss": 0.4313,
      "step": 850
    },
    {
      "epoch": 80.0,
      "grad_norm": 3.889315128326416,
      "learning_rate": 0.00045000000000000004,
      "loss": 0.4242,
      "step": 900
    },
    {
      "epoch": 84.44444444444444,
      "grad_norm": 9.169000625610352,
      "learning_rate": 0.000475,
      "loss": 0.4703,
      "step": 950
    },
    {
      "epoch": 88.88888888888889,
      "grad_norm": 3.6193642616271973,
      "learning_rate": 0.0005,
      "loss": 0.4709,
      "step": 1000
    },
    {
      "epoch": 88.88888888888889,
      "eval_loss": 0.4963502883911133,
      "eval_runtime": 9.099,
      "eval_samples_per_second": 17.474,
      "eval_steps_per_second": 2.198,
      "step": 1000
    },
    {
      "epoch": 93.33333333333333,
      "grad_norm": 1.7234474420547485,
      "learning_rate": 0.0005250000000000001,
      "loss": 0.4563,
      "step": 1050
    },
    {
      "epoch": 97.77777777777777,
      "grad_norm": 2.933211326599121,
      "learning_rate": 0.00055,
      "loss": 0.4449,
      "step": 1100
    },
    {
      "epoch": 102.22222222222223,
      "grad_norm": 2.3853394985198975,
      "learning_rate": 0.000575,
      "loss": 0.4712,
      "step": 1150
    },
    {
      "epoch": 106.66666666666667,
      "grad_norm": 0.20497646927833557,
      "learning_rate": 0.000599,
      "loss": 1.534,
      "step": 1200
    },
    {
      "epoch": 111.11111111111111,
      "grad_norm": 0.0945451632142067,
      "learning_rate": 0.000624,
      "loss": 1.5935,
      "step": 1250
    },
    {
      "epoch": 115.55555555555556,
      "grad_norm": 0.13917936384677887,
      "learning_rate": 0.0006490000000000001,
      "loss": 1.5906,
      "step": 1300
    },
    {
      "epoch": 120.0,
      "grad_norm": 0.21496649086475372,
      "learning_rate": 0.000674,
      "loss": 1.5817,
      "step": 1350
    },
    {
      "epoch": 124.44444444444444,
      "grad_norm": 0.7981905341148376,
      "learning_rate": 0.000699,
      "loss": 1.5851,
      "step": 1400
    },
    {
      "epoch": 128.88888888888889,
      "grad_norm": 0.3589198589324951,
      "learning_rate": 0.000724,
      "loss": 1.5778,
      "step": 1450
    },
    {
      "epoch": 133.33333333333334,
      "grad_norm": 0.15568029880523682,
      "learning_rate": 0.000749,
      "loss": 1.5891,
      "step": 1500
    },
    {
      "epoch": 133.33333333333334,
      "eval_loss": 1.5470129251480103,
      "eval_runtime": 10.6543,
      "eval_samples_per_second": 14.924,
      "eval_steps_per_second": 1.877,
      "step": 1500
    },
    {
      "epoch": 137.77777777777777,
      "grad_norm": 0.12888091802597046,
      "learning_rate": 0.0007740000000000001,
      "loss": 1.5836,
      "step": 1550
    },
    {
      "epoch": 142.22222222222223,
      "grad_norm": 0.17486007511615753,
      "learning_rate": 0.000799,
      "loss": 1.5765,
      "step": 1600
    },
    {
      "epoch": 146.66666666666666,
      "grad_norm": 0.07815095037221909,
      "learning_rate": 0.000824,
      "loss": 1.5771,
      "step": 1650
    },
    {
      "epoch": 151.11111111111111,
      "grad_norm": 0.8746868371963501,
      "learning_rate": 0.000849,
      "loss": 1.5796,
      "step": 1700
    },
    {
      "epoch": 155.55555555555554,
      "grad_norm": 0.30329012870788574,
      "learning_rate": 0.000874,
      "loss": 1.5779,
      "step": 1750
    },
    {
      "epoch": 160.0,
      "grad_norm": 0.3016453981399536,
      "learning_rate": 0.0008990000000000001,
      "loss": 1.5766,
      "step": 1800
    },
    {
      "epoch": 164.44444444444446,
      "grad_norm": 0.12022041529417038,
      "learning_rate": 0.000924,
      "loss": 1.5775,
      "step": 1850
    },
    {
      "epoch": 168.88888888888889,
      "grad_norm": 0.20793229341506958,
      "learning_rate": 0.000949,
      "loss": 1.5768,
      "step": 1900
    },
    {
      "epoch": 173.33333333333334,
      "grad_norm": 0.03221210092306137,
      "learning_rate": 0.000974,
      "loss": 1.5779,
      "step": 1950
    },
    {
      "epoch": 177.77777777777777,
      "grad_norm": 0.22706760466098785,
      "learning_rate": 0.000999,
      "loss": 1.5787,
      "step": 2000
    },
    {
      "epoch": 177.77777777777777,
      "eval_loss": 1.5496832132339478,
      "eval_runtime": 8.1644,
      "eval_samples_per_second": 19.475,
      "eval_steps_per_second": 2.45,
      "step": 2000
    },
    {
      "epoch": 182.22222222222223,
      "grad_norm": 0.13995479047298431,
      "learning_rate": 0.000976,
      "loss": 1.5783,
      "step": 2050
    },
    {
      "epoch": 186.66666666666666,
      "grad_norm": 0.19275836646556854,
      "learning_rate": 0.000951,
      "loss": 1.5748,
      "step": 2100
    },
    {
      "epoch": 191.11111111111111,
      "grad_norm": 0.20744530856609344,
      "learning_rate": 0.0009260000000000001,
      "loss": 1.578,
      "step": 2150
    },
    {
      "epoch": 195.55555555555554,
      "grad_norm": 0.08092151582241058,
      "learning_rate": 0.000901,
      "loss": 1.5765,
      "step": 2200
    },
    {
      "epoch": 200.0,
      "grad_norm": 0.1601410210132599,
      "learning_rate": 0.000876,
      "loss": 1.5751,
      "step": 2250
    },
    {
      "epoch": 204.44444444444446,
      "grad_norm": 0.20833778381347656,
      "learning_rate": 0.000851,
      "loss": 1.5723,
      "step": 2300
    },
    {
      "epoch": 208.88888888888889,
      "grad_norm": 0.3090430498123169,
      "learning_rate": 0.000826,
      "loss": 1.5741,
      "step": 2350
    },
    {
      "epoch": 213.33333333333334,
      "grad_norm": 0.049218811094760895,
      "learning_rate": 0.0008010000000000001,
      "loss": 1.5743,
      "step": 2400
    },
    {
      "epoch": 217.77777777777777,
      "grad_norm": 0.10529515147209167,
      "learning_rate": 0.000776,
      "loss": 1.5721,
      "step": 2450
    },
    {
      "epoch": 222.22222222222223,
      "grad_norm": 0.21685709059238434,
      "learning_rate": 0.000751,
      "loss": 1.5714,
      "step": 2500
    },
    {
      "epoch": 222.22222222222223,
      "eval_loss": 1.5523545742034912,
      "eval_runtime": 10.3314,
      "eval_samples_per_second": 15.39,
      "eval_steps_per_second": 1.936,
      "step": 2500
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 364,
  "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": 6.882220043045424e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}