{
  "best_metric": 0.4111216962337494,
  "best_model_checkpoint": "mikhail_panzo/zlm-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": 5.992666721343994,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.9248,
      "step": 50
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 3.4944448471069336,
      "learning_rate": 4.95e-07,
      "loss": 0.8709,
      "step": 100
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 1.9902064800262451,
      "learning_rate": 7.450000000000001e-07,
      "loss": 0.8005,
      "step": 150
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": 1.4649031162261963,
      "learning_rate": 9.950000000000002e-07,
      "loss": 0.733,
      "step": 200
    },
    {
      "epoch": 10.869565217391305,
      "grad_norm": 1.4645055532455444,
      "learning_rate": 1.2450000000000002e-06,
      "loss": 0.6758,
      "step": 250
    },
    {
      "epoch": 13.043478260869565,
      "grad_norm": 1.5862590074539185,
      "learning_rate": 1.495e-06,
      "loss": 0.6246,
      "step": 300
    },
    {
      "epoch": 15.217391304347826,
      "grad_norm": 1.198890209197998,
      "learning_rate": 1.745e-06,
      "loss": 0.5917,
      "step": 350
    },
    {
      "epoch": 17.391304347826086,
      "grad_norm": 1.324212908744812,
      "learning_rate": 1.9950000000000004e-06,
      "loss": 0.5721,
      "step": 400
    },
    {
      "epoch": 19.565217391304348,
      "grad_norm": 1.1638662815093994,
      "learning_rate": 2.245e-06,
      "loss": 0.5635,
      "step": 450
    },
    {
      "epoch": 21.73913043478261,
      "grad_norm": 0.9293428063392639,
      "learning_rate": 2.4950000000000003e-06,
      "loss": 0.5541,
      "step": 500
    },
    {
      "epoch": 21.73913043478261,
      "eval_loss": 0.49769046902656555,
      "eval_runtime": 12.8521,
      "eval_samples_per_second": 12.527,
      "eval_steps_per_second": 1.634,
      "step": 500
    },
    {
      "epoch": 23.91304347826087,
      "grad_norm": 0.7982977628707886,
      "learning_rate": 2.7450000000000004e-06,
      "loss": 0.5487,
      "step": 550
    },
    {
      "epoch": 26.08695652173913,
      "grad_norm": 0.8491495251655579,
      "learning_rate": 2.995e-06,
      "loss": 0.5373,
      "step": 600
    },
    {
      "epoch": 28.26086956521739,
      "grad_norm": 0.7158030867576599,
      "learning_rate": 3.2450000000000003e-06,
      "loss": 0.5271,
      "step": 650
    },
    {
      "epoch": 30.434782608695652,
      "grad_norm": 1.2020411491394043,
      "learning_rate": 3.495e-06,
      "loss": 0.5201,
      "step": 700
    },
    {
      "epoch": 32.608695652173914,
      "grad_norm": 0.7339745163917542,
      "learning_rate": 3.745e-06,
      "loss": 0.5126,
      "step": 750
    },
    {
      "epoch": 34.78260869565217,
      "grad_norm": 1.1129218339920044,
      "learning_rate": 3.995000000000001e-06,
      "loss": 0.5136,
      "step": 800
    },
    {
      "epoch": 36.95652173913044,
      "grad_norm": 0.7733961939811707,
      "learning_rate": 4.245e-06,
      "loss": 0.5099,
      "step": 850
    },
    {
      "epoch": 39.130434782608695,
      "grad_norm": 1.1559356451034546,
      "learning_rate": 4.495e-06,
      "loss": 0.5047,
      "step": 900
    },
    {
      "epoch": 41.30434782608695,
      "grad_norm": 0.873518705368042,
      "learning_rate": 4.745e-06,
      "loss": 0.4955,
      "step": 950
    },
    {
      "epoch": 43.47826086956522,
      "grad_norm": 0.703760027885437,
      "learning_rate": 4.9950000000000005e-06,
      "loss": 0.4931,
      "step": 1000
    },
    {
      "epoch": 43.47826086956522,
      "eval_loss": 0.4528730511665344,
      "eval_runtime": 12.5404,
      "eval_samples_per_second": 12.838,
      "eval_steps_per_second": 1.675,
      "step": 1000
    },
    {
      "epoch": 45.65217391304348,
      "grad_norm": 0.9317522644996643,
      "learning_rate": 5.245e-06,
      "loss": 0.4922,
      "step": 1050
    },
    {
      "epoch": 47.82608695652174,
      "grad_norm": 0.9118973016738892,
      "learning_rate": 5.495000000000001e-06,
      "loss": 0.4855,
      "step": 1100
    },
    {
      "epoch": 50.0,
      "grad_norm": 2.0197670459747314,
      "learning_rate": 5.745000000000001e-06,
      "loss": 0.485,
      "step": 1150
    },
    {
      "epoch": 52.17391304347826,
      "grad_norm": 1.1067094802856445,
      "learning_rate": 5.995000000000001e-06,
      "loss": 0.4872,
      "step": 1200
    },
    {
      "epoch": 54.34782608695652,
      "grad_norm": 1.20389723777771,
      "learning_rate": 6.245000000000001e-06,
      "loss": 0.4836,
      "step": 1250
    },
    {
      "epoch": 56.52173913043478,
      "grad_norm": 0.9784926176071167,
      "learning_rate": 6.4950000000000005e-06,
      "loss": 0.4793,
      "step": 1300
    },
    {
      "epoch": 58.69565217391305,
      "grad_norm": 1.4276039600372314,
      "learning_rate": 6.745000000000001e-06,
      "loss": 0.4774,
      "step": 1350
    },
    {
      "epoch": 60.869565217391305,
      "grad_norm": 1.117233157157898,
      "learning_rate": 6.995000000000001e-06,
      "loss": 0.4776,
      "step": 1400
    },
    {
      "epoch": 63.04347826086956,
      "grad_norm": 0.9267759919166565,
      "learning_rate": 7.245000000000001e-06,
      "loss": 0.4682,
      "step": 1450
    },
    {
      "epoch": 65.21739130434783,
      "grad_norm": 0.8514929413795471,
      "learning_rate": 7.495000000000001e-06,
      "loss": 0.4695,
      "step": 1500
    },
    {
      "epoch": 65.21739130434783,
      "eval_loss": 0.43296581506729126,
      "eval_runtime": 12.858,
      "eval_samples_per_second": 12.521,
      "eval_steps_per_second": 1.633,
      "step": 1500
    },
    {
      "epoch": 67.3913043478261,
      "grad_norm": 1.6638849973678589,
      "learning_rate": 7.745e-06,
      "loss": 0.4674,
      "step": 1550
    },
    {
      "epoch": 69.56521739130434,
      "grad_norm": 0.8030112981796265,
      "learning_rate": 7.995e-06,
      "loss": 0.4652,
      "step": 1600
    },
    {
      "epoch": 71.73913043478261,
      "grad_norm": 1.2407337427139282,
      "learning_rate": 8.245000000000002e-06,
      "loss": 0.4638,
      "step": 1650
    },
    {
      "epoch": 73.91304347826087,
      "grad_norm": 1.0331224203109741,
      "learning_rate": 8.495e-06,
      "loss": 0.4685,
      "step": 1700
    },
    {
      "epoch": 76.08695652173913,
      "grad_norm": 2.4681615829467773,
      "learning_rate": 8.745000000000002e-06,
      "loss": 0.4627,
      "step": 1750
    },
    {
      "epoch": 78.26086956521739,
      "grad_norm": 0.9344178438186646,
      "learning_rate": 8.995000000000001e-06,
      "loss": 0.4575,
      "step": 1800
    },
    {
      "epoch": 80.43478260869566,
      "grad_norm": 1.0561341047286987,
      "learning_rate": 9.245e-06,
      "loss": 0.4621,
      "step": 1850
    },
    {
      "epoch": 82.6086956521739,
      "grad_norm": 0.9816989302635193,
      "learning_rate": 9.495000000000001e-06,
      "loss": 0.4544,
      "step": 1900
    },
    {
      "epoch": 84.78260869565217,
      "grad_norm": 1.4779256582260132,
      "learning_rate": 9.745e-06,
      "loss": 0.4579,
      "step": 1950
    },
    {
      "epoch": 86.95652173913044,
      "grad_norm": 1.1634950637817383,
      "learning_rate": 9.995000000000002e-06,
      "loss": 0.4518,
      "step": 2000
    },
    {
      "epoch": 86.95652173913044,
      "eval_loss": 0.42301931977272034,
      "eval_runtime": 12.725,
      "eval_samples_per_second": 12.652,
      "eval_steps_per_second": 1.65,
      "step": 2000
    },
    {
      "epoch": 89.1304347826087,
      "grad_norm": 1.0197867155075073,
      "learning_rate": 9.918333333333335e-06,
      "loss": 0.4501,
      "step": 2050
    },
    {
      "epoch": 91.30434782608695,
      "grad_norm": 1.2543045282363892,
      "learning_rate": 9.835000000000002e-06,
      "loss": 0.4499,
      "step": 2100
    },
    {
      "epoch": 93.47826086956522,
      "grad_norm": 0.8574827909469604,
      "learning_rate": 9.751666666666667e-06,
      "loss": 0.4485,
      "step": 2150
    },
    {
      "epoch": 95.65217391304348,
      "grad_norm": 1.3784998655319214,
      "learning_rate": 9.668333333333334e-06,
      "loss": 0.4507,
      "step": 2200
    },
    {
      "epoch": 97.82608695652173,
      "grad_norm": 0.9518347978591919,
      "learning_rate": 9.585e-06,
      "loss": 0.4486,
      "step": 2250
    },
    {
      "epoch": 100.0,
      "grad_norm": 2.422320604324341,
      "learning_rate": 9.501666666666667e-06,
      "loss": 0.4455,
      "step": 2300
    },
    {
      "epoch": 102.17391304347827,
      "grad_norm": 1.3441377878189087,
      "learning_rate": 9.418333333333334e-06,
      "loss": 0.45,
      "step": 2350
    },
    {
      "epoch": 104.34782608695652,
      "grad_norm": 0.969160795211792,
      "learning_rate": 9.335000000000001e-06,
      "loss": 0.4454,
      "step": 2400
    },
    {
      "epoch": 106.52173913043478,
      "grad_norm": 2.1275179386138916,
      "learning_rate": 9.251666666666668e-06,
      "loss": 0.4459,
      "step": 2450
    },
    {
      "epoch": 108.69565217391305,
      "grad_norm": 1.3673362731933594,
      "learning_rate": 9.168333333333333e-06,
      "loss": 0.4442,
      "step": 2500
    },
    {
      "epoch": 108.69565217391305,
      "eval_loss": 0.4178585410118103,
      "eval_runtime": 13.0534,
      "eval_samples_per_second": 12.334,
      "eval_steps_per_second": 1.609,
      "step": 2500
    },
    {
      "epoch": 110.8695652173913,
      "grad_norm": 1.3769398927688599,
      "learning_rate": 9.085e-06,
      "loss": 0.4406,
      "step": 2550
    },
    {
      "epoch": 113.04347826086956,
      "grad_norm": 0.9913681745529175,
      "learning_rate": 9.001666666666667e-06,
      "loss": 0.4389,
      "step": 2600
    },
    {
      "epoch": 115.21739130434783,
      "grad_norm": 1.1747106313705444,
      "learning_rate": 8.918333333333334e-06,
      "loss": 0.4386,
      "step": 2650
    },
    {
      "epoch": 117.3913043478261,
      "grad_norm": 1.0514781475067139,
      "learning_rate": 8.835000000000001e-06,
      "loss": 0.4393,
      "step": 2700
    },
    {
      "epoch": 119.56521739130434,
      "grad_norm": 1.8967915773391724,
      "learning_rate": 8.751666666666668e-06,
      "loss": 0.4421,
      "step": 2750
    },
    {
      "epoch": 121.73913043478261,
      "grad_norm": 0.8795832395553589,
      "learning_rate": 8.668333333333335e-06,
      "loss": 0.436,
      "step": 2800
    },
    {
      "epoch": 123.91304347826087,
      "grad_norm": 0.7928704023361206,
      "learning_rate": 8.585000000000002e-06,
      "loss": 0.4333,
      "step": 2850
    },
    {
      "epoch": 126.08695652173913,
      "grad_norm": 1.2805510759353638,
      "learning_rate": 8.501666666666667e-06,
      "loss": 0.4326,
      "step": 2900
    },
    {
      "epoch": 128.2608695652174,
      "grad_norm": 1.420920968055725,
      "learning_rate": 8.418333333333334e-06,
      "loss": 0.4317,
      "step": 2950
    },
    {
      "epoch": 130.43478260869566,
      "grad_norm": 0.7063888907432556,
      "learning_rate": 8.335e-06,
      "loss": 0.4344,
      "step": 3000
    },
    {
      "epoch": 130.43478260869566,
      "eval_loss": 0.41350311040878296,
      "eval_runtime": 12.8066,
      "eval_samples_per_second": 12.572,
      "eval_steps_per_second": 1.64,
      "step": 3000
    },
    {
      "epoch": 132.6086956521739,
      "grad_norm": 1.8065855503082275,
      "learning_rate": 8.251666666666668e-06,
      "loss": 0.4361,
      "step": 3050
    },
    {
      "epoch": 134.7826086956522,
      "grad_norm": 0.8073704838752747,
      "learning_rate": 8.168333333333334e-06,
      "loss": 0.4339,
      "step": 3100
    },
    {
      "epoch": 136.95652173913044,
      "grad_norm": 1.2890065908432007,
      "learning_rate": 8.085000000000001e-06,
      "loss": 0.4325,
      "step": 3150
    },
    {
      "epoch": 139.1304347826087,
      "grad_norm": 1.336401104927063,
      "learning_rate": 8.001666666666668e-06,
      "loss": 0.4334,
      "step": 3200
    },
    {
      "epoch": 141.30434782608697,
      "grad_norm": 1.2965891361236572,
      "learning_rate": 7.918333333333333e-06,
      "loss": 0.4298,
      "step": 3250
    },
    {
      "epoch": 143.47826086956522,
      "grad_norm": 0.8761409521102905,
      "learning_rate": 7.835e-06,
      "loss": 0.4231,
      "step": 3300
    },
    {
      "epoch": 145.65217391304347,
      "grad_norm": 1.1475930213928223,
      "learning_rate": 7.751666666666667e-06,
      "loss": 0.4316,
      "step": 3350
    },
    {
      "epoch": 147.82608695652175,
      "grad_norm": 0.8305974006652832,
      "learning_rate": 7.668333333333334e-06,
      "loss": 0.4277,
      "step": 3400
    },
    {
      "epoch": 150.0,
      "grad_norm": 1.6335935592651367,
      "learning_rate": 7.585e-06,
      "loss": 0.4248,
      "step": 3450
    },
    {
      "epoch": 152.17391304347825,
      "grad_norm": 1.1171984672546387,
      "learning_rate": 7.501666666666667e-06,
      "loss": 0.4318,
      "step": 3500
    },
    {
      "epoch": 152.17391304347825,
      "eval_loss": 0.4111216962337494,
      "eval_runtime": 13.0991,
      "eval_samples_per_second": 12.291,
      "eval_steps_per_second": 1.603,
      "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": 4.241687452614216e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}