{
  "best_metric": 0.4078540802001953,
  "best_model_checkpoint": "mikhail-panzo/fil_b128_le4_s8000/checkpoint-1000",
  "epoch": 220.99447513812154,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.419889502762431,
      "grad_norm": 2.1200406551361084,
      "learning_rate": 2.5e-06,
      "loss": 0.77,
      "step": 50
    },
    {
      "epoch": 8.839779005524862,
      "grad_norm": 0.8919622302055359,
      "learning_rate": 5e-06,
      "loss": 0.695,
      "step": 100
    },
    {
      "epoch": 13.259668508287293,
      "grad_norm": 0.9864558577537537,
      "learning_rate": 7.5e-06,
      "loss": 0.6278,
      "step": 150
    },
    {
      "epoch": 17.679558011049725,
      "grad_norm": 0.9195986986160278,
      "learning_rate": 1e-05,
      "loss": 0.5431,
      "step": 200
    },
    {
      "epoch": 22.099447513812155,
      "grad_norm": 1.0175281763076782,
      "learning_rate": 1.25e-05,
      "loss": 0.5155,
      "step": 250
    },
    {
      "epoch": 26.519337016574585,
      "grad_norm": 1.0303566455841064,
      "learning_rate": 1.5e-05,
      "loss": 0.502,
      "step": 300
    },
    {
      "epoch": 30.939226519337016,
      "grad_norm": 1.005897879600525,
      "learning_rate": 1.75e-05,
      "loss": 0.4889,
      "step": 350
    },
    {
      "epoch": 35.35911602209945,
      "grad_norm": 1.0381360054016113,
      "learning_rate": 2e-05,
      "loss": 0.477,
      "step": 400
    },
    {
      "epoch": 39.77900552486188,
      "grad_norm": 1.2148911952972412,
      "learning_rate": 2.25e-05,
      "loss": 0.4711,
      "step": 450
    },
    {
      "epoch": 44.19889502762431,
      "grad_norm": 1.5404785871505737,
      "learning_rate": 2.5e-05,
      "loss": 0.463,
      "step": 500
    },
    {
      "epoch": 44.19889502762431,
      "eval_loss": 0.42523157596588135,
      "eval_runtime": 15.6706,
      "eval_samples_per_second": 10.274,
      "eval_steps_per_second": 1.34,
      "step": 500
    },
    {
      "epoch": 48.61878453038674,
      "grad_norm": 1.1097612380981445,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.4601,
      "step": 550
    },
    {
      "epoch": 53.03867403314917,
      "grad_norm": 1.2796109914779663,
      "learning_rate": 3e-05,
      "loss": 0.454,
      "step": 600
    },
    {
      "epoch": 57.4585635359116,
      "grad_norm": 1.0656301975250244,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.4498,
      "step": 650
    },
    {
      "epoch": 61.87845303867403,
      "grad_norm": 1.3971511125564575,
      "learning_rate": 3.5e-05,
      "loss": 0.4439,
      "step": 700
    },
    {
      "epoch": 66.29834254143647,
      "grad_norm": 1.9344899654388428,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4418,
      "step": 750
    },
    {
      "epoch": 70.7182320441989,
      "grad_norm": 1.080334186553955,
      "learning_rate": 4e-05,
      "loss": 0.4379,
      "step": 800
    },
    {
      "epoch": 75.13812154696133,
      "grad_norm": 1.9891581535339355,
      "learning_rate": 4.25e-05,
      "loss": 0.4334,
      "step": 850
    },
    {
      "epoch": 79.55801104972376,
      "grad_norm": 1.3854402303695679,
      "learning_rate": 4.5e-05,
      "loss": 0.4331,
      "step": 900
    },
    {
      "epoch": 83.97790055248619,
      "grad_norm": 1.5818183422088623,
      "learning_rate": 4.75e-05,
      "loss": 0.428,
      "step": 950
    },
    {
      "epoch": 88.39779005524862,
      "grad_norm": 1.8474106788635254,
      "learning_rate": 5e-05,
      "loss": 0.4238,
      "step": 1000
    },
    {
      "epoch": 88.39779005524862,
      "eval_loss": 0.4078540802001953,
      "eval_runtime": 14.8436,
      "eval_samples_per_second": 10.846,
      "eval_steps_per_second": 1.415,
      "step": 1000
    },
    {
      "epoch": 92.81767955801105,
      "grad_norm": 3.1851751804351807,
      "learning_rate": 5.25e-05,
      "loss": 0.4234,
      "step": 1050
    },
    {
      "epoch": 97.23756906077348,
      "grad_norm": 1.3754140138626099,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.4209,
      "step": 1100
    },
    {
      "epoch": 101.65745856353591,
      "grad_norm": 1.6143475770950317,
      "learning_rate": 5.7499999999999995e-05,
      "loss": 0.4228,
      "step": 1150
    },
    {
      "epoch": 106.07734806629834,
      "grad_norm": 1.9961720705032349,
      "learning_rate": 6e-05,
      "loss": 0.4162,
      "step": 1200
    },
    {
      "epoch": 110.49723756906077,
      "grad_norm": 1.9125347137451172,
      "learning_rate": 6.25e-05,
      "loss": 0.4178,
      "step": 1250
    },
    {
      "epoch": 114.9171270718232,
      "grad_norm": 3.1913959980010986,
      "learning_rate": 6.500000000000001e-05,
      "loss": 0.4143,
      "step": 1300
    },
    {
      "epoch": 119.33701657458563,
      "grad_norm": 2.1010069847106934,
      "learning_rate": 6.750000000000001e-05,
      "loss": 0.4118,
      "step": 1350
    },
    {
      "epoch": 123.75690607734806,
      "grad_norm": 1.3079431056976318,
      "learning_rate": 7e-05,
      "loss": 0.4091,
      "step": 1400
    },
    {
      "epoch": 128.1767955801105,
      "grad_norm": 2.3663132190704346,
      "learning_rate": 7.25e-05,
      "loss": 0.4092,
      "step": 1450
    },
    {
      "epoch": 132.59668508287294,
      "grad_norm": 1.031698226928711,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.4083,
      "step": 1500
    },
    {
      "epoch": 132.59668508287294,
      "eval_loss": 0.4082767069339752,
      "eval_runtime": 15.1068,
      "eval_samples_per_second": 10.657,
      "eval_steps_per_second": 1.39,
      "step": 1500
    },
    {
      "epoch": 137.01657458563537,
      "grad_norm": 2.7781426906585693,
      "learning_rate": 7.75e-05,
      "loss": 0.4079,
      "step": 1550
    },
    {
      "epoch": 141.4364640883978,
      "grad_norm": 1.136169672012329,
      "learning_rate": 8e-05,
      "loss": 0.4028,
      "step": 1600
    },
    {
      "epoch": 145.85635359116023,
      "grad_norm": 1.7299158573150635,
      "learning_rate": 8.25e-05,
      "loss": 0.4034,
      "step": 1650
    },
    {
      "epoch": 150.27624309392266,
      "grad_norm": 2.4281983375549316,
      "learning_rate": 8.5e-05,
      "loss": 0.3992,
      "step": 1700
    },
    {
      "epoch": 154.6961325966851,
      "grad_norm": 3.6834373474121094,
      "learning_rate": 8.75e-05,
      "loss": 0.3997,
      "step": 1750
    },
    {
      "epoch": 159.11602209944752,
      "grad_norm": 3.1484293937683105,
      "learning_rate": 9e-05,
      "loss": 0.396,
      "step": 1800
    },
    {
      "epoch": 163.53591160220995,
      "grad_norm": 3.978259563446045,
      "learning_rate": 9.250000000000001e-05,
      "loss": 0.3969,
      "step": 1850
    },
    {
      "epoch": 167.95580110497238,
      "grad_norm": 2.906266212463379,
      "learning_rate": 9.5e-05,
      "loss": 0.4017,
      "step": 1900
    },
    {
      "epoch": 172.3756906077348,
      "grad_norm": 1.1569880247116089,
      "learning_rate": 9.75e-05,
      "loss": 0.396,
      "step": 1950
    },
    {
      "epoch": 176.79558011049724,
      "grad_norm": 2.2872982025146484,
      "learning_rate": 0.0001,
      "loss": 0.3928,
      "step": 2000
    },
    {
      "epoch": 176.79558011049724,
      "eval_loss": 0.414584219455719,
      "eval_runtime": 15.2895,
      "eval_samples_per_second": 10.53,
      "eval_steps_per_second": 1.373,
      "step": 2000
    },
    {
      "epoch": 181.21546961325967,
      "grad_norm": 1.1768990755081177,
      "learning_rate": 9.916666666666667e-05,
      "loss": 0.3928,
      "step": 2050
    },
    {
      "epoch": 185.6353591160221,
      "grad_norm": 1.2087491750717163,
      "learning_rate": 9.833333333333333e-05,
      "loss": 0.3901,
      "step": 2100
    },
    {
      "epoch": 190.05524861878453,
      "grad_norm": 1.17588210105896,
      "learning_rate": 9.75e-05,
      "loss": 0.3877,
      "step": 2150
    },
    {
      "epoch": 194.47513812154696,
      "grad_norm": 1.5979018211364746,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.3879,
      "step": 2200
    },
    {
      "epoch": 198.8950276243094,
      "grad_norm": 1.0489567518234253,
      "learning_rate": 9.583333333333334e-05,
      "loss": 0.3869,
      "step": 2250
    },
    {
      "epoch": 203.31491712707182,
      "grad_norm": 1.833038091659546,
      "learning_rate": 9.501666666666668e-05,
      "loss": 0.3859,
      "step": 2300
    },
    {
      "epoch": 207.73480662983425,
      "grad_norm": 2.075849771499634,
      "learning_rate": 9.418333333333334e-05,
      "loss": 0.3843,
      "step": 2350
    },
    {
      "epoch": 212.15469613259668,
      "grad_norm": 2.8613317012786865,
      "learning_rate": 9.335e-05,
      "loss": 0.381,
      "step": 2400
    },
    {
      "epoch": 216.5745856353591,
      "grad_norm": 1.6291433572769165,
      "learning_rate": 9.251666666666667e-05,
      "loss": 0.3802,
      "step": 2450
    },
    {
      "epoch": 220.99447513812154,
      "grad_norm": 1.1548006534576416,
      "learning_rate": 9.168333333333333e-05,
      "loss": 0.381,
      "step": 2500
    },
    {
      "epoch": 220.99447513812154,
      "eval_loss": 0.4127160906791687,
      "eval_runtime": 15.1026,
      "eval_samples_per_second": 10.66,
      "eval_steps_per_second": 1.39,
      "step": 2500
    }
  ],
  "logging_steps": 50,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 728,
  "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": 5.660506148311872e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}