{
  "best_metric": 0.41326722502708435,
  "best_model_checkpoint": "mikhail-panzo/fil_b64_le5_s8000/checkpoint-3500",
  "epoch": 152.17391304347825,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 4.140745639801025,
      "learning_rate": 2.4500000000000004e-07,
      "loss": 0.8272,
      "step": 50
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 1.78262197971344,
      "learning_rate": 4.95e-07,
      "loss": 0.7463,
      "step": 100
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 4.001489639282227,
      "learning_rate": 7.450000000000001e-07,
      "loss": 0.7504,
      "step": 150
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": 1.0013818740844727,
      "learning_rate": 9.950000000000002e-07,
      "loss": 0.7295,
      "step": 200
    },
    {
      "epoch": 10.869565217391305,
      "grad_norm": 1.9014511108398438,
      "learning_rate": 1.2450000000000002e-06,
      "loss": 0.7182,
      "step": 250
    },
    {
      "epoch": 13.043478260869565,
      "grad_norm": 13.358899116516113,
      "learning_rate": 1.495e-06,
      "loss": 0.7037,
      "step": 300
    },
    {
      "epoch": 15.217391304347826,
      "grad_norm": 1.4039109945297241,
      "learning_rate": 1.745e-06,
      "loss": 0.6966,
      "step": 350
    },
    {
      "epoch": 17.391304347826086,
      "grad_norm": 3.0226972103118896,
      "learning_rate": 1.9950000000000004e-06,
      "loss": 0.6702,
      "step": 400
    },
    {
      "epoch": 19.565217391304348,
      "grad_norm": 1.716300129890442,
      "learning_rate": 2.24e-06,
      "loss": 0.6487,
      "step": 450
    },
    {
      "epoch": 21.73913043478261,
      "grad_norm": 2.9351744651794434,
      "learning_rate": 2.4900000000000003e-06,
      "loss": 0.6071,
      "step": 500
    },
    {
      "epoch": 21.73913043478261,
      "eval_loss": 0.521294116973877,
      "eval_runtime": 6.6451,
      "eval_samples_per_second": 23.927,
      "eval_steps_per_second": 3.01,
      "step": 500
    },
    {
      "epoch": 23.91304347826087,
      "grad_norm": 1.9992552995681763,
      "learning_rate": 2.7400000000000004e-06,
      "loss": 0.5759,
      "step": 550
    },
    {
      "epoch": 26.08695652173913,
      "grad_norm": 1.2594470977783203,
      "learning_rate": 2.99e-06,
      "loss": 0.56,
      "step": 600
    },
    {
      "epoch": 28.26086956521739,
      "grad_norm": 1.0602363348007202,
      "learning_rate": 3.2400000000000003e-06,
      "loss": 0.5709,
      "step": 650
    },
    {
      "epoch": 30.434782608695652,
      "grad_norm": 1.116231083869934,
      "learning_rate": 3.49e-06,
      "loss": 0.5322,
      "step": 700
    },
    {
      "epoch": 32.608695652173914,
      "grad_norm": 1.5203620195388794,
      "learning_rate": 3.74e-06,
      "loss": 0.5264,
      "step": 750
    },
    {
      "epoch": 34.78260869565217,
      "grad_norm": 1.6139357089996338,
      "learning_rate": 3.990000000000001e-06,
      "loss": 0.5191,
      "step": 800
    },
    {
      "epoch": 36.95652173913044,
      "grad_norm": 1.436273217201233,
      "learning_rate": 4.24e-06,
      "loss": 0.5272,
      "step": 850
    },
    {
      "epoch": 39.130434782608695,
      "grad_norm": 1.757323145866394,
      "learning_rate": 4.49e-06,
      "loss": 0.5016,
      "step": 900
    },
    {
      "epoch": 41.30434782608695,
      "grad_norm": 0.9879368543624878,
      "learning_rate": 4.74e-06,
      "loss": 0.4975,
      "step": 950
    },
    {
      "epoch": 43.47826086956522,
      "grad_norm": 1.3402222394943237,
      "learning_rate": 4.9900000000000005e-06,
      "loss": 0.5126,
      "step": 1000
    },
    {
      "epoch": 43.47826086956522,
      "eval_loss": 0.4506620466709137,
      "eval_runtime": 6.4644,
      "eval_samples_per_second": 24.596,
      "eval_steps_per_second": 3.094,
      "step": 1000
    },
    {
      "epoch": 45.65217391304348,
      "grad_norm": 1.9366440773010254,
      "learning_rate": 5.240000000000001e-06,
      "loss": 0.5043,
      "step": 1050
    },
    {
      "epoch": 47.82608695652174,
      "grad_norm": 1.5358505249023438,
      "learning_rate": 5.490000000000001e-06,
      "loss": 0.5034,
      "step": 1100
    },
    {
      "epoch": 50.0,
      "grad_norm": 1.6836323738098145,
      "learning_rate": 5.74e-06,
      "loss": 0.499,
      "step": 1150
    },
    {
      "epoch": 52.17391304347826,
      "grad_norm": 1.5092076063156128,
      "learning_rate": 5.99e-06,
      "loss": 0.487,
      "step": 1200
    },
    {
      "epoch": 54.34782608695652,
      "grad_norm": 1.0751235485076904,
      "learning_rate": 6.24e-06,
      "loss": 0.4871,
      "step": 1250
    },
    {
      "epoch": 56.52173913043478,
      "grad_norm": 1.327528476715088,
      "learning_rate": 6.4900000000000005e-06,
      "loss": 0.4911,
      "step": 1300
    },
    {
      "epoch": 58.69565217391305,
      "grad_norm": 1.7827174663543701,
      "learning_rate": 6.740000000000001e-06,
      "loss": 0.4772,
      "step": 1350
    },
    {
      "epoch": 60.869565217391305,
      "grad_norm": 1.2707558870315552,
      "learning_rate": 6.99e-06,
      "loss": 0.4786,
      "step": 1400
    },
    {
      "epoch": 63.04347826086956,
      "grad_norm": 1.0794321298599243,
      "learning_rate": 7.24e-06,
      "loss": 0.4741,
      "step": 1450
    },
    {
      "epoch": 65.21739130434783,
      "grad_norm": 2.0180580615997314,
      "learning_rate": 7.49e-06,
      "loss": 0.4749,
      "step": 1500
    },
    {
      "epoch": 65.21739130434783,
      "eval_loss": 0.4311259388923645,
      "eval_runtime": 6.4494,
      "eval_samples_per_second": 24.653,
      "eval_steps_per_second": 3.101,
      "step": 1500
    },
    {
      "epoch": 67.3913043478261,
      "grad_norm": 1.2938276529312134,
      "learning_rate": 7.74e-06,
      "loss": 0.4684,
      "step": 1550
    },
    {
      "epoch": 69.56521739130434,
      "grad_norm": 1.8100048303604126,
      "learning_rate": 7.990000000000001e-06,
      "loss": 0.4797,
      "step": 1600
    },
    {
      "epoch": 71.73913043478261,
      "grad_norm": 1.8357532024383545,
      "learning_rate": 8.24e-06,
      "loss": 0.4813,
      "step": 1650
    },
    {
      "epoch": 73.91304347826087,
      "grad_norm": 1.4251928329467773,
      "learning_rate": 8.49e-06,
      "loss": 0.4622,
      "step": 1700
    },
    {
      "epoch": 76.08695652173913,
      "grad_norm": 1.1764353513717651,
      "learning_rate": 8.740000000000001e-06,
      "loss": 0.4642,
      "step": 1750
    },
    {
      "epoch": 78.26086956521739,
      "grad_norm": 1.6363924741744995,
      "learning_rate": 8.99e-06,
      "loss": 0.4669,
      "step": 1800
    },
    {
      "epoch": 80.43478260869566,
      "grad_norm": 1.485127568244934,
      "learning_rate": 9.240000000000001e-06,
      "loss": 0.4595,
      "step": 1850
    },
    {
      "epoch": 82.6086956521739,
      "grad_norm": 1.2844778299331665,
      "learning_rate": 9.49e-06,
      "loss": 0.463,
      "step": 1900
    },
    {
      "epoch": 84.78260869565217,
      "grad_norm": 1.0181550979614258,
      "learning_rate": 9.74e-06,
      "loss": 0.4529,
      "step": 1950
    },
    {
      "epoch": 86.95652173913044,
      "grad_norm": 1.3658286333084106,
      "learning_rate": 9.990000000000001e-06,
      "loss": 0.454,
      "step": 2000
    },
    {
      "epoch": 86.95652173913044,
      "eval_loss": 0.4230838716030121,
      "eval_runtime": 6.4713,
      "eval_samples_per_second": 24.57,
      "eval_steps_per_second": 3.091,
      "step": 2000
    },
    {
      "epoch": 89.1304347826087,
      "grad_norm": 2.8028907775878906,
      "learning_rate": 9.920000000000002e-06,
      "loss": 0.4694,
      "step": 2050
    },
    {
      "epoch": 91.30434782608695,
      "grad_norm": 0.8982108235359192,
      "learning_rate": 9.836666666666668e-06,
      "loss": 0.4495,
      "step": 2100
    },
    {
      "epoch": 93.47826086956522,
      "grad_norm": 2.500758409500122,
      "learning_rate": 9.753333333333335e-06,
      "loss": 0.4456,
      "step": 2150
    },
    {
      "epoch": 95.65217391304348,
      "grad_norm": 1.7864420413970947,
      "learning_rate": 9.67e-06,
      "loss": 0.4816,
      "step": 2200
    },
    {
      "epoch": 97.82608695652173,
      "grad_norm": 1.3289756774902344,
      "learning_rate": 9.586666666666667e-06,
      "loss": 0.4512,
      "step": 2250
    },
    {
      "epoch": 100.0,
      "grad_norm": 1.5230562686920166,
      "learning_rate": 9.503333333333334e-06,
      "loss": 0.4564,
      "step": 2300
    },
    {
      "epoch": 102.17391304347827,
      "grad_norm": 2.709244728088379,
      "learning_rate": 9.42e-06,
      "loss": 0.4543,
      "step": 2350
    },
    {
      "epoch": 104.34782608695652,
      "grad_norm": 2.3624541759490967,
      "learning_rate": 9.336666666666666e-06,
      "loss": 0.4505,
      "step": 2400
    },
    {
      "epoch": 106.52173913043478,
      "grad_norm": 1.4635056257247925,
      "learning_rate": 9.253333333333333e-06,
      "loss": 0.4558,
      "step": 2450
    },
    {
      "epoch": 108.69565217391305,
      "grad_norm": 1.9029183387756348,
      "learning_rate": 9.17e-06,
      "loss": 0.443,
      "step": 2500
    },
    {
      "epoch": 108.69565217391305,
      "eval_loss": 0.41728243231773376,
      "eval_runtime": 6.4572,
      "eval_samples_per_second": 24.624,
      "eval_steps_per_second": 3.097,
      "step": 2500
    },
    {
      "epoch": 110.8695652173913,
      "grad_norm": 1.4863485097885132,
      "learning_rate": 9.088333333333334e-06,
      "loss": 0.4562,
      "step": 2550
    },
    {
      "epoch": 113.04347826086956,
      "grad_norm": 1.1933757066726685,
      "learning_rate": 9.005000000000001e-06,
      "loss": 0.4438,
      "step": 2600
    },
    {
      "epoch": 115.21739130434783,
      "grad_norm": 2.2257957458496094,
      "learning_rate": 8.921666666666668e-06,
      "loss": 0.4511,
      "step": 2650
    },
    {
      "epoch": 117.3913043478261,
      "grad_norm": 1.382228970527649,
      "learning_rate": 8.838333333333335e-06,
      "loss": 0.4442,
      "step": 2700
    },
    {
      "epoch": 119.56521739130434,
      "grad_norm": 1.1577272415161133,
      "learning_rate": 8.755e-06,
      "loss": 0.4376,
      "step": 2750
    },
    {
      "epoch": 121.73913043478261,
      "grad_norm": 1.16239333152771,
      "learning_rate": 8.671666666666667e-06,
      "loss": 0.4378,
      "step": 2800
    },
    {
      "epoch": 123.91304347826087,
      "grad_norm": 1.3845179080963135,
      "learning_rate": 8.588333333333334e-06,
      "loss": 0.4368,
      "step": 2850
    },
    {
      "epoch": 126.08695652173913,
      "grad_norm": 1.4513936042785645,
      "learning_rate": 8.505e-06,
      "loss": 0.4362,
      "step": 2900
    },
    {
      "epoch": 128.2608695652174,
      "grad_norm": 2.2269773483276367,
      "learning_rate": 8.421666666666668e-06,
      "loss": 0.4498,
      "step": 2950
    },
    {
      "epoch": 130.43478260869566,
      "grad_norm": 2.3055992126464844,
      "learning_rate": 8.338333333333335e-06,
      "loss": 0.4376,
      "step": 3000
    },
    {
      "epoch": 130.43478260869566,
      "eval_loss": 0.4168698191642761,
      "eval_runtime": 6.612,
      "eval_samples_per_second": 24.047,
      "eval_steps_per_second": 3.025,
      "step": 3000
    },
    {
      "epoch": 132.6086956521739,
      "grad_norm": 1.1443853378295898,
      "learning_rate": 8.255000000000001e-06,
      "loss": 0.4369,
      "step": 3050
    },
    {
      "epoch": 134.7826086956522,
      "grad_norm": 1.3519920110702515,
      "learning_rate": 8.171666666666668e-06,
      "loss": 0.4322,
      "step": 3100
    },
    {
      "epoch": 136.95652173913044,
      "grad_norm": 1.4229977130889893,
      "learning_rate": 8.088333333333334e-06,
      "loss": 0.436,
      "step": 3150
    },
    {
      "epoch": 139.1304347826087,
      "grad_norm": 3.4992916584014893,
      "learning_rate": 8.005e-06,
      "loss": 0.4275,
      "step": 3200
    },
    {
      "epoch": 141.30434782608697,
      "grad_norm": 0.9590508937835693,
      "learning_rate": 7.921666666666667e-06,
      "loss": 0.4366,
      "step": 3250
    },
    {
      "epoch": 143.47826086956522,
      "grad_norm": 4.028496265411377,
      "learning_rate": 7.838333333333334e-06,
      "loss": 0.437,
      "step": 3300
    },
    {
      "epoch": 145.65217391304347,
      "grad_norm": 0.9438303709030151,
      "learning_rate": 7.755000000000001e-06,
      "loss": 0.4317,
      "step": 3350
    },
    {
      "epoch": 147.82608695652175,
      "grad_norm": 1.5847684144973755,
      "learning_rate": 7.671666666666668e-06,
      "loss": 0.4303,
      "step": 3400
    },
    {
      "epoch": 150.0,
      "grad_norm": 1.676293969154358,
      "learning_rate": 7.588333333333334e-06,
      "loss": 0.4312,
      "step": 3450
    },
    {
      "epoch": 152.17391304347825,
      "grad_norm": 0.9531082510948181,
      "learning_rate": 7.505e-06,
      "loss": 0.4287,
      "step": 3500
    },
    {
      "epoch": 152.17391304347825,
      "eval_loss": 0.41326722502708435,
      "eval_runtime": 6.485,
      "eval_samples_per_second": 24.518,
      "eval_steps_per_second": 3.084,
      "step": 3500
    }
  ],
  "logging_steps": 50,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 348,
  "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": 5.194546896348691e+16,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}