{
  "best_metric": 0.3398433029651642,
  "best_model_checkpoint": "mikhail-panzo/zlm_b128_le4_s4000/checkpoint-3500",
  "epoch": 5.863874345549738,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 2.9717624187469482,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 1.0424,
      "step": 50
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 2.9720630645751953,
      "learning_rate": 4.950000000000001e-06,
      "loss": 0.8474,
      "step": 100
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 2.445929765701294,
      "learning_rate": 7.45e-06,
      "loss": 0.7336,
      "step": 150
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 5.502955913543701,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.6492,
      "step": 200
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 2.3356130123138428,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.6133,
      "step": 250
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 1.937270164489746,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.5889,
      "step": 300
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 2.392244338989258,
      "learning_rate": 1.745e-05,
      "loss": 0.5694,
      "step": 350
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 7.3209919929504395,
      "learning_rate": 1.995e-05,
      "loss": 0.5477,
      "step": 400
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 3.415917158126831,
      "learning_rate": 2.245e-05,
      "loss": 0.5329,
      "step": 450
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 3.0256705284118652,
      "learning_rate": 2.495e-05,
      "loss": 0.5173,
      "step": 500
    },
    {
      "epoch": 0.837696335078534,
      "eval_loss": 0.4566049873828888,
      "eval_runtime": 261.3511,
      "eval_samples_per_second": 32.481,
      "eval_steps_per_second": 4.063,
      "step": 500
    },
    {
      "epoch": 0.9214659685863874,
      "grad_norm": 1.9436837434768677,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.5079,
      "step": 550
    },
    {
      "epoch": 1.0052356020942408,
      "grad_norm": 1.819956660270691,
      "learning_rate": 2.995e-05,
      "loss": 0.4969,
      "step": 600
    },
    {
      "epoch": 1.0890052356020943,
      "grad_norm": 5.457251071929932,
      "learning_rate": 3.245e-05,
      "loss": 0.4977,
      "step": 650
    },
    {
      "epoch": 1.1727748691099475,
      "grad_norm": 3.183980703353882,
      "learning_rate": 3.495e-05,
      "loss": 0.4923,
      "step": 700
    },
    {
      "epoch": 1.256544502617801,
      "grad_norm": 7.1660051345825195,
      "learning_rate": 3.745e-05,
      "loss": 0.4802,
      "step": 750
    },
    {
      "epoch": 1.3403141361256545,
      "grad_norm": 5.499026775360107,
      "learning_rate": 3.995e-05,
      "loss": 0.4754,
      "step": 800
    },
    {
      "epoch": 1.4240837696335078,
      "grad_norm": 2.8053908348083496,
      "learning_rate": 4.245e-05,
      "loss": 0.4669,
      "step": 850
    },
    {
      "epoch": 1.5078534031413613,
      "grad_norm": 3.017005443572998,
      "learning_rate": 4.495e-05,
      "loss": 0.4604,
      "step": 900
    },
    {
      "epoch": 1.5916230366492146,
      "grad_norm": 2.7971177101135254,
      "learning_rate": 4.745e-05,
      "loss": 0.4565,
      "step": 950
    },
    {
      "epoch": 1.675392670157068,
      "grad_norm": 3.1588356494903564,
      "learning_rate": 4.995e-05,
      "loss": 0.455,
      "step": 1000
    },
    {
      "epoch": 1.675392670157068,
      "eval_loss": 0.40312233567237854,
      "eval_runtime": 256.4334,
      "eval_samples_per_second": 33.104,
      "eval_steps_per_second": 4.141,
      "step": 1000
    },
    {
      "epoch": 1.7591623036649215,
      "grad_norm": 2.2053232192993164,
      "learning_rate": 5.245e-05,
      "loss": 0.4543,
      "step": 1050
    },
    {
      "epoch": 1.8429319371727748,
      "grad_norm": 2.0562164783477783,
      "learning_rate": 5.495e-05,
      "loss": 0.4456,
      "step": 1100
    },
    {
      "epoch": 1.9267015706806283,
      "grad_norm": 2.730119466781616,
      "learning_rate": 5.745e-05,
      "loss": 0.4355,
      "step": 1150
    },
    {
      "epoch": 2.0104712041884816,
      "grad_norm": 1.7484283447265625,
      "learning_rate": 5.995000000000001e-05,
      "loss": 0.4299,
      "step": 1200
    },
    {
      "epoch": 2.094240837696335,
      "grad_norm": 1.1786061525344849,
      "learning_rate": 6.245000000000001e-05,
      "loss": 0.4305,
      "step": 1250
    },
    {
      "epoch": 2.1780104712041886,
      "grad_norm": 1.98978590965271,
      "learning_rate": 6.494999999999999e-05,
      "loss": 0.4295,
      "step": 1300
    },
    {
      "epoch": 2.261780104712042,
      "grad_norm": 2.818659782409668,
      "learning_rate": 6.745e-05,
      "loss": 0.4235,
      "step": 1350
    },
    {
      "epoch": 2.345549738219895,
      "grad_norm": 2.3864262104034424,
      "learning_rate": 6.995e-05,
      "loss": 0.4271,
      "step": 1400
    },
    {
      "epoch": 2.4293193717277486,
      "grad_norm": 1.3647903203964233,
      "learning_rate": 7.245000000000001e-05,
      "loss": 0.4208,
      "step": 1450
    },
    {
      "epoch": 2.513089005235602,
      "grad_norm": 2.2144172191619873,
      "learning_rate": 7.495e-05,
      "loss": 0.4175,
      "step": 1500
    },
    {
      "epoch": 2.513089005235602,
      "eval_loss": 0.3777858018875122,
      "eval_runtime": 260.5025,
      "eval_samples_per_second": 32.587,
      "eval_steps_per_second": 4.077,
      "step": 1500
    },
    {
      "epoch": 2.5968586387434556,
      "grad_norm": 1.6483193635940552,
      "learning_rate": 7.745e-05,
      "loss": 0.414,
      "step": 1550
    },
    {
      "epoch": 2.680628272251309,
      "grad_norm": 1.7688554525375366,
      "learning_rate": 7.995e-05,
      "loss": 0.4153,
      "step": 1600
    },
    {
      "epoch": 2.7643979057591626,
      "grad_norm": 1.2314317226409912,
      "learning_rate": 8.245e-05,
      "loss": 0.4089,
      "step": 1650
    },
    {
      "epoch": 2.8481675392670156,
      "grad_norm": 1.6623793840408325,
      "learning_rate": 8.495e-05,
      "loss": 0.4124,
      "step": 1700
    },
    {
      "epoch": 2.931937172774869,
      "grad_norm": 3.812507390975952,
      "learning_rate": 8.745000000000001e-05,
      "loss": 0.4112,
      "step": 1750
    },
    {
      "epoch": 3.0157068062827226,
      "grad_norm": 2.141019821166992,
      "learning_rate": 8.995e-05,
      "loss": 0.4081,
      "step": 1800
    },
    {
      "epoch": 3.099476439790576,
      "grad_norm": 1.8928133249282837,
      "learning_rate": 9.245e-05,
      "loss": 0.4067,
      "step": 1850
    },
    {
      "epoch": 3.183246073298429,
      "grad_norm": 2.322817087173462,
      "learning_rate": 9.495e-05,
      "loss": 0.4088,
      "step": 1900
    },
    {
      "epoch": 3.2670157068062826,
      "grad_norm": 2.1984918117523193,
      "learning_rate": 9.745000000000001e-05,
      "loss": 0.3976,
      "step": 1950
    },
    {
      "epoch": 3.350785340314136,
      "grad_norm": 2.0455121994018555,
      "learning_rate": 9.995e-05,
      "loss": 0.4022,
      "step": 2000
    },
    {
      "epoch": 3.350785340314136,
      "eval_loss": 0.3677983582019806,
      "eval_runtime": 257.3111,
      "eval_samples_per_second": 32.991,
      "eval_steps_per_second": 4.127,
      "step": 2000
    },
    {
      "epoch": 3.4345549738219896,
      "grad_norm": 2.0729174613952637,
      "learning_rate": 9.755000000000001e-05,
      "loss": 0.4017,
      "step": 2050
    },
    {
      "epoch": 3.518324607329843,
      "grad_norm": 1.4496002197265625,
      "learning_rate": 9.505e-05,
      "loss": 0.3999,
      "step": 2100
    },
    {
      "epoch": 3.6020942408376966,
      "grad_norm": 1.313783884048462,
      "learning_rate": 9.255e-05,
      "loss": 0.3971,
      "step": 2150
    },
    {
      "epoch": 3.6858638743455496,
      "grad_norm": 2.4938676357269287,
      "learning_rate": 9.005000000000001e-05,
      "loss": 0.3958,
      "step": 2200
    },
    {
      "epoch": 3.769633507853403,
      "grad_norm": 1.6081351041793823,
      "learning_rate": 8.755e-05,
      "loss": 0.389,
      "step": 2250
    },
    {
      "epoch": 3.8534031413612566,
      "grad_norm": 1.5322096347808838,
      "learning_rate": 8.505000000000001e-05,
      "loss": 0.3887,
      "step": 2300
    },
    {
      "epoch": 3.93717277486911,
      "grad_norm": 2.3951408863067627,
      "learning_rate": 8.26e-05,
      "loss": 0.392,
      "step": 2350
    },
    {
      "epoch": 4.020942408376963,
      "grad_norm": 1.0958179235458374,
      "learning_rate": 8.010000000000001e-05,
      "loss": 0.3846,
      "step": 2400
    },
    {
      "epoch": 4.104712041884817,
      "grad_norm": 1.2315720319747925,
      "learning_rate": 7.76e-05,
      "loss": 0.3867,
      "step": 2450
    },
    {
      "epoch": 4.18848167539267,
      "grad_norm": 1.3186556100845337,
      "learning_rate": 7.510000000000001e-05,
      "loss": 0.3848,
      "step": 2500
    },
    {
      "epoch": 4.18848167539267,
      "eval_loss": 0.3522670865058899,
      "eval_runtime": 269.3867,
      "eval_samples_per_second": 31.512,
      "eval_steps_per_second": 3.942,
      "step": 2500
    },
    {
      "epoch": 4.272251308900524,
      "grad_norm": 1.6223663091659546,
      "learning_rate": 7.26e-05,
      "loss": 0.3868,
      "step": 2550
    },
    {
      "epoch": 4.356020942408377,
      "grad_norm": 1.2300375699996948,
      "learning_rate": 7.01e-05,
      "loss": 0.3834,
      "step": 2600
    },
    {
      "epoch": 4.439790575916231,
      "grad_norm": 1.930428147315979,
      "learning_rate": 6.76e-05,
      "loss": 0.3789,
      "step": 2650
    },
    {
      "epoch": 4.523560209424084,
      "grad_norm": 1.2785884141921997,
      "learning_rate": 6.510000000000001e-05,
      "loss": 0.3821,
      "step": 2700
    },
    {
      "epoch": 4.607329842931938,
      "grad_norm": 2.007714033126831,
      "learning_rate": 6.26e-05,
      "loss": 0.3777,
      "step": 2750
    },
    {
      "epoch": 4.69109947643979,
      "grad_norm": 1.490657925605774,
      "learning_rate": 6.0100000000000004e-05,
      "loss": 0.3817,
      "step": 2800
    },
    {
      "epoch": 4.774869109947644,
      "grad_norm": 1.7175956964492798,
      "learning_rate": 5.76e-05,
      "loss": 0.3778,
      "step": 2850
    },
    {
      "epoch": 4.858638743455497,
      "grad_norm": 1.9603019952774048,
      "learning_rate": 5.5100000000000004e-05,
      "loss": 0.3787,
      "step": 2900
    },
    {
      "epoch": 4.942408376963351,
      "grad_norm": 0.9075130224227905,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 0.3775,
      "step": 2950
    },
    {
      "epoch": 5.026178010471204,
      "grad_norm": 1.1054623126983643,
      "learning_rate": 5.0100000000000005e-05,
      "loss": 0.3763,
      "step": 3000
    },
    {
      "epoch": 5.026178010471204,
      "eval_loss": 0.34259819984436035,
      "eval_runtime": 256.7226,
      "eval_samples_per_second": 33.067,
      "eval_steps_per_second": 4.137,
      "step": 3000
    },
    {
      "epoch": 5.109947643979058,
      "grad_norm": 0.8834495544433594,
      "learning_rate": 4.76e-05,
      "loss": 0.3718,
      "step": 3050
    },
    {
      "epoch": 5.193717277486911,
      "grad_norm": 1.585207223892212,
      "learning_rate": 4.5100000000000005e-05,
      "loss": 0.3732,
      "step": 3100
    },
    {
      "epoch": 5.277486910994765,
      "grad_norm": 1.8974605798721313,
      "learning_rate": 4.26e-05,
      "loss": 0.3714,
      "step": 3150
    },
    {
      "epoch": 5.361256544502618,
      "grad_norm": 1.280151128768921,
      "learning_rate": 4.0100000000000006e-05,
      "loss": 0.3703,
      "step": 3200
    },
    {
      "epoch": 5.445026178010472,
      "grad_norm": 1.453099012374878,
      "learning_rate": 3.76e-05,
      "loss": 0.3733,
      "step": 3250
    },
    {
      "epoch": 5.528795811518324,
      "grad_norm": 1.4233230352401733,
      "learning_rate": 3.51e-05,
      "loss": 0.3725,
      "step": 3300
    },
    {
      "epoch": 5.612565445026178,
      "grad_norm": 0.9149217009544373,
      "learning_rate": 3.26e-05,
      "loss": 0.3674,
      "step": 3350
    },
    {
      "epoch": 5.696335078534031,
      "grad_norm": 0.9855162501335144,
      "learning_rate": 3.01e-05,
      "loss": 0.3685,
      "step": 3400
    },
    {
      "epoch": 5.780104712041885,
      "grad_norm": 1.0689723491668701,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.3684,
      "step": 3450
    },
    {
      "epoch": 5.863874345549738,
      "grad_norm": 1.527082920074463,
      "learning_rate": 2.51e-05,
      "loss": 0.3665,
      "step": 3500
    },
    {
      "epoch": 5.863874345549738,
      "eval_loss": 0.3398433029651642,
      "eval_runtime": 259.8568,
      "eval_samples_per_second": 32.668,
      "eval_steps_per_second": 4.087,
      "step": 3500
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 6.269975786347891e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}