{
  "best_metric": 0.41223829984664917,
  "best_model_checkpoint": "mikhail-panzo/zlm-fil-ceb_b64_le5_s8000/checkpoint-1500",
  "epoch": 173.91304347826087,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 0.7357823252677917,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.4055,
      "step": 50
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 0.7944665551185608,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.3988,
      "step": 100
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 0.7226569652557373,
      "learning_rate": 7.5e-07,
      "loss": 0.4072,
      "step": 150
    },
    {
      "epoch": 8.695652173913043,
      "grad_norm": 0.7845238447189331,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.4028,
      "step": 200
    },
    {
      "epoch": 10.869565217391305,
      "grad_norm": 0.9769566655158997,
      "learning_rate": 1.25e-06,
      "loss": 0.4066,
      "step": 250
    },
    {
      "epoch": 13.043478260869565,
      "grad_norm": 0.8634074330329895,
      "learning_rate": 1.5e-06,
      "loss": 0.405,
      "step": 300
    },
    {
      "epoch": 15.217391304347826,
      "grad_norm": 0.9029327034950256,
      "learning_rate": 1.75e-06,
      "loss": 0.4074,
      "step": 350
    },
    {
      "epoch": 17.391304347826086,
      "grad_norm": 0.7624074220657349,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.406,
      "step": 400
    },
    {
      "epoch": 19.565217391304348,
      "grad_norm": 1.005196213722229,
      "learning_rate": 2.25e-06,
      "loss": 0.4052,
      "step": 450
    },
    {
      "epoch": 21.73913043478261,
      "grad_norm": 1.0276380777359009,
      "learning_rate": 2.5e-06,
      "loss": 0.4019,
      "step": 500
    },
    {
      "epoch": 21.73913043478261,
      "eval_loss": 0.4143177568912506,
      "eval_runtime": 6.4074,
      "eval_samples_per_second": 24.815,
      "eval_steps_per_second": 3.121,
      "step": 500
    },
    {
      "epoch": 23.91304347826087,
      "grad_norm": 0.9067140817642212,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.4021,
      "step": 550
    },
    {
      "epoch": 26.08695652173913,
      "grad_norm": 0.8799042701721191,
      "learning_rate": 3e-06,
      "loss": 0.4071,
      "step": 600
    },
    {
      "epoch": 28.26086956521739,
      "grad_norm": 1.0451533794403076,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.4146,
      "step": 650
    },
    {
      "epoch": 30.434782608695652,
      "grad_norm": 0.7998800277709961,
      "learning_rate": 3.5e-06,
      "loss": 0.4023,
      "step": 700
    },
    {
      "epoch": 32.608695652173914,
      "grad_norm": 1.0301792621612549,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.4045,
      "step": 750
    },
    {
      "epoch": 34.78260869565217,
      "grad_norm": 0.6388185620307922,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.4045,
      "step": 800
    },
    {
      "epoch": 36.95652173913044,
      "grad_norm": 0.9676840901374817,
      "learning_rate": 4.25e-06,
      "loss": 0.4116,
      "step": 850
    },
    {
      "epoch": 39.130434782608695,
      "grad_norm": 0.8355514407157898,
      "learning_rate": 4.5e-06,
      "loss": 0.4013,
      "step": 900
    },
    {
      "epoch": 41.30434782608695,
      "grad_norm": 0.7534067630767822,
      "learning_rate": 4.75e-06,
      "loss": 0.4019,
      "step": 950
    },
    {
      "epoch": 43.47826086956522,
      "grad_norm": 0.9366117119789124,
      "learning_rate": 5e-06,
      "loss": 0.4064,
      "step": 1000
    },
    {
      "epoch": 43.47826086956522,
      "eval_loss": 0.41316837072372437,
      "eval_runtime": 6.5616,
      "eval_samples_per_second": 24.232,
      "eval_steps_per_second": 3.048,
      "step": 1000
    },
    {
      "epoch": 45.65217391304348,
      "grad_norm": 0.7025179266929626,
      "learning_rate": 5.2500000000000006e-06,
      "loss": 0.4061,
      "step": 1050
    },
    {
      "epoch": 47.82608695652174,
      "grad_norm": 0.9240646958351135,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.4061,
      "step": 1100
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.9685391187667847,
      "learning_rate": 5.75e-06,
      "loss": 0.4063,
      "step": 1150
    },
    {
      "epoch": 52.17391304347826,
      "grad_norm": 0.8591094017028809,
      "learning_rate": 6e-06,
      "loss": 0.4013,
      "step": 1200
    },
    {
      "epoch": 54.34782608695652,
      "grad_norm": 0.9348228573799133,
      "learning_rate": 6.25e-06,
      "loss": 0.402,
      "step": 1250
    },
    {
      "epoch": 56.52173913043478,
      "grad_norm": 0.7908409237861633,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.4046,
      "step": 1300
    },
    {
      "epoch": 58.69565217391305,
      "grad_norm": 1.109017014503479,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.3995,
      "step": 1350
    },
    {
      "epoch": 60.869565217391305,
      "grad_norm": 0.8529478311538696,
      "learning_rate": 7e-06,
      "loss": 0.4018,
      "step": 1400
    },
    {
      "epoch": 63.04347826086956,
      "grad_norm": 0.9947476387023926,
      "learning_rate": 7.25e-06,
      "loss": 0.401,
      "step": 1450
    },
    {
      "epoch": 65.21739130434783,
      "grad_norm": 1.2476813793182373,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4034,
      "step": 1500
    },
    {
      "epoch": 65.21739130434783,
      "eval_loss": 0.41223829984664917,
      "eval_runtime": 6.5846,
      "eval_samples_per_second": 24.147,
      "eval_steps_per_second": 3.037,
      "step": 1500
    },
    {
      "epoch": 67.3913043478261,
      "grad_norm": 0.8658013343811035,
      "learning_rate": 7.75e-06,
      "loss": 0.4007,
      "step": 1550
    },
    {
      "epoch": 69.56521739130434,
      "grad_norm": 0.7932788133621216,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.4054,
      "step": 1600
    },
    {
      "epoch": 71.73913043478261,
      "grad_norm": 2.33347225189209,
      "learning_rate": 8.25e-06,
      "loss": 0.4075,
      "step": 1650
    },
    {
      "epoch": 73.91304347826087,
      "grad_norm": 1.0095164775848389,
      "learning_rate": 8.5e-06,
      "loss": 0.3982,
      "step": 1700
    },
    {
      "epoch": 76.08695652173913,
      "grad_norm": 0.8895902633666992,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.4018,
      "step": 1750
    },
    {
      "epoch": 78.26086956521739,
      "grad_norm": 1.2252330780029297,
      "learning_rate": 9e-06,
      "loss": 0.4024,
      "step": 1800
    },
    {
      "epoch": 80.43478260869566,
      "grad_norm": 0.7731898427009583,
      "learning_rate": 9.250000000000001e-06,
      "loss": 0.3994,
      "step": 1850
    },
    {
      "epoch": 82.6086956521739,
      "grad_norm": 1.047607660293579,
      "learning_rate": 9.5e-06,
      "loss": 0.4013,
      "step": 1900
    },
    {
      "epoch": 84.78260869565217,
      "grad_norm": 0.9284490346908569,
      "learning_rate": 9.75e-06,
      "loss": 0.3986,
      "step": 1950
    },
    {
      "epoch": 86.95652173913044,
      "grad_norm": 4.725491046905518,
      "learning_rate": 1e-05,
      "loss": 0.3985,
      "step": 2000
    },
    {
      "epoch": 86.95652173913044,
      "eval_loss": 0.4147759974002838,
      "eval_runtime": 6.5475,
      "eval_samples_per_second": 24.284,
      "eval_steps_per_second": 3.055,
      "step": 2000
    },
    {
      "epoch": 89.1304347826087,
      "grad_norm": 1.1678683757781982,
      "learning_rate": 9.916666666666668e-06,
      "loss": 0.4066,
      "step": 2050
    },
    {
      "epoch": 91.30434782608695,
      "grad_norm": 1.988890528678894,
      "learning_rate": 9.835000000000002e-06,
      "loss": 0.3997,
      "step": 2100
    },
    {
      "epoch": 93.47826086956522,
      "grad_norm": 1.0278427600860596,
      "learning_rate": 9.751666666666667e-06,
      "loss": 0.3948,
      "step": 2150
    },
    {
      "epoch": 95.65217391304348,
      "grad_norm": 0.9606735110282898,
      "learning_rate": 9.67e-06,
      "loss": 0.42,
      "step": 2200
    },
    {
      "epoch": 97.82608695652173,
      "grad_norm": 1.0861903429031372,
      "learning_rate": 9.586666666666667e-06,
      "loss": 0.398,
      "step": 2250
    },
    {
      "epoch": 100.0,
      "grad_norm": 1.4872643947601318,
      "learning_rate": 9.503333333333334e-06,
      "loss": 0.4059,
      "step": 2300
    },
    {
      "epoch": 102.17391304347827,
      "grad_norm": 1.0460981130599976,
      "learning_rate": 9.42e-06,
      "loss": 0.4014,
      "step": 2350
    },
    {
      "epoch": 104.34782608695652,
      "grad_norm": 1.906476378440857,
      "learning_rate": 9.336666666666666e-06,
      "loss": 0.4015,
      "step": 2400
    },
    {
      "epoch": 106.52173913043478,
      "grad_norm": 1.3965051174163818,
      "learning_rate": 9.253333333333333e-06,
      "loss": 0.4063,
      "step": 2450
    },
    {
      "epoch": 108.69565217391305,
      "grad_norm": 1.7410107851028442,
      "learning_rate": 9.17e-06,
      "loss": 0.396,
      "step": 2500
    },
    {
      "epoch": 108.69565217391305,
      "eval_loss": 0.41233959794044495,
      "eval_runtime": 6.6031,
      "eval_samples_per_second": 24.08,
      "eval_steps_per_second": 3.029,
      "step": 2500
    },
    {
      "epoch": 110.8695652173913,
      "grad_norm": 1.4562053680419922,
      "learning_rate": 9.088333333333334e-06,
      "loss": 0.4042,
      "step": 2550
    },
    {
      "epoch": 113.04347826086956,
      "grad_norm": 0.8741047978401184,
      "learning_rate": 9.005000000000001e-06,
      "loss": 0.3985,
      "step": 2600
    },
    {
      "epoch": 115.21739130434783,
      "grad_norm": 1.1817028522491455,
      "learning_rate": 8.921666666666668e-06,
      "loss": 0.4036,
      "step": 2650
    },
    {
      "epoch": 117.3913043478261,
      "grad_norm": 0.8515534996986389,
      "learning_rate": 8.838333333333335e-06,
      "loss": 0.3986,
      "step": 2700
    },
    {
      "epoch": 119.56521739130434,
      "grad_norm": 0.9545428156852722,
      "learning_rate": 8.755e-06,
      "loss": 0.3946,
      "step": 2750
    },
    {
      "epoch": 121.73913043478261,
      "grad_norm": 0.6653321385383606,
      "learning_rate": 8.671666666666667e-06,
      "loss": 0.3957,
      "step": 2800
    },
    {
      "epoch": 123.91304347826087,
      "grad_norm": 0.8100114464759827,
      "learning_rate": 8.588333333333334e-06,
      "loss": 0.3953,
      "step": 2850
    },
    {
      "epoch": 126.08695652173913,
      "grad_norm": 1.4174237251281738,
      "learning_rate": 8.505e-06,
      "loss": 0.3933,
      "step": 2900
    },
    {
      "epoch": 128.2608695652174,
      "grad_norm": 1.6101123094558716,
      "learning_rate": 8.421666666666668e-06,
      "loss": 0.4027,
      "step": 2950
    },
    {
      "epoch": 130.43478260869566,
      "grad_norm": 0.7453967332839966,
      "learning_rate": 8.338333333333335e-06,
      "loss": 0.3958,
      "step": 3000
    },
    {
      "epoch": 130.43478260869566,
      "eval_loss": 0.4138609766960144,
      "eval_runtime": 6.6093,
      "eval_samples_per_second": 24.057,
      "eval_steps_per_second": 3.026,
      "step": 3000
    },
    {
      "epoch": 132.6086956521739,
      "grad_norm": 1.061684012413025,
      "learning_rate": 8.255000000000001e-06,
      "loss": 0.3966,
      "step": 3050
    },
    {
      "epoch": 134.7826086956522,
      "grad_norm": 1.0820105075836182,
      "learning_rate": 8.171666666666668e-06,
      "loss": 0.3937,
      "step": 3100
    },
    {
      "epoch": 136.95652173913044,
      "grad_norm": 0.8147062063217163,
      "learning_rate": 8.088333333333334e-06,
      "loss": 0.3961,
      "step": 3150
    },
    {
      "epoch": 139.1304347826087,
      "grad_norm": 0.7450351715087891,
      "learning_rate": 8.005e-06,
      "loss": 0.3923,
      "step": 3200
    },
    {
      "epoch": 141.30434782608697,
      "grad_norm": 1.05226731300354,
      "learning_rate": 7.921666666666667e-06,
      "loss": 0.3951,
      "step": 3250
    },
    {
      "epoch": 143.47826086956522,
      "grad_norm": 1.5119163990020752,
      "learning_rate": 7.838333333333334e-06,
      "loss": 0.3957,
      "step": 3300
    },
    {
      "epoch": 145.65217391304347,
      "grad_norm": 1.0422383546829224,
      "learning_rate": 7.755000000000001e-06,
      "loss": 0.3942,
      "step": 3350
    },
    {
      "epoch": 147.82608695652175,
      "grad_norm": 1.2120935916900635,
      "learning_rate": 7.671666666666668e-06,
      "loss": 0.3946,
      "step": 3400
    },
    {
      "epoch": 150.0,
      "grad_norm": 1.3127976655960083,
      "learning_rate": 7.588333333333334e-06,
      "loss": 0.3933,
      "step": 3450
    },
    {
      "epoch": 152.17391304347825,
      "grad_norm": 0.7704339027404785,
      "learning_rate": 7.505e-06,
      "loss": 0.3928,
      "step": 3500
    },
    {
      "epoch": 152.17391304347825,
      "eval_loss": 0.4132215976715088,
      "eval_runtime": 6.6162,
      "eval_samples_per_second": 24.032,
      "eval_steps_per_second": 3.023,
      "step": 3500
    },
    {
      "epoch": 154.34782608695653,
      "grad_norm": 1.4827122688293457,
      "learning_rate": 7.421666666666667e-06,
      "loss": 0.3908,
      "step": 3550
    },
    {
      "epoch": 156.52173913043478,
      "grad_norm": 1.0464049577713013,
      "learning_rate": 7.338333333333334e-06,
      "loss": 0.3993,
      "step": 3600
    },
    {
      "epoch": 158.69565217391303,
      "grad_norm": 0.8279913663864136,
      "learning_rate": 7.255000000000001e-06,
      "loss": 0.3967,
      "step": 3650
    },
    {
      "epoch": 160.8695652173913,
      "grad_norm": 0.9175034165382385,
      "learning_rate": 7.171666666666667e-06,
      "loss": 0.3911,
      "step": 3700
    },
    {
      "epoch": 163.04347826086956,
      "grad_norm": 1.0080233812332153,
      "learning_rate": 7.088333333333334e-06,
      "loss": 0.3898,
      "step": 3750
    },
    {
      "epoch": 165.2173913043478,
      "grad_norm": 0.9944823980331421,
      "learning_rate": 7.005000000000001e-06,
      "loss": 0.3901,
      "step": 3800
    },
    {
      "epoch": 167.3913043478261,
      "grad_norm": 1.8222392797470093,
      "learning_rate": 6.921666666666668e-06,
      "loss": 0.3901,
      "step": 3850
    },
    {
      "epoch": 169.56521739130434,
      "grad_norm": 0.9506675004959106,
      "learning_rate": 6.838333333333334e-06,
      "loss": 0.3945,
      "step": 3900
    },
    {
      "epoch": 171.7391304347826,
      "grad_norm": 4.512270450592041,
      "learning_rate": 6.7550000000000005e-06,
      "loss": 0.3913,
      "step": 3950
    },
    {
      "epoch": 173.91304347826087,
      "grad_norm": 0.74826580286026,
      "learning_rate": 6.6716666666666674e-06,
      "loss": 0.3914,
      "step": 4000
    },
    {
      "epoch": 173.91304347826087,
      "eval_loss": 0.4138501286506653,
      "eval_runtime": 6.5399,
      "eval_samples_per_second": 24.312,
      "eval_steps_per_second": 3.058,
      "step": 4000
    }
  ],
  "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": 5.923192104795082e+16,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}