{
  "best_metric": 0.31398773193359375,
  "best_model_checkpoint": "mikhail_panzo/zlm_b128_le4_s8000/checkpoint-8000",
  "epoch": 13.403141361256544,
  "eval_steps": 500,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 2.9895308017730713,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 1.0423,
      "step": 50
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 3.051593542098999,
      "learning_rate": 4.950000000000001e-06,
      "loss": 0.8473,
      "step": 100
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 2.0044381618499756,
      "learning_rate": 7.45e-06,
      "loss": 0.733,
      "step": 150
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 3.4974701404571533,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.6511,
      "step": 200
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 1.854073405265808,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.6143,
      "step": 250
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 1.737787127494812,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.5909,
      "step": 300
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 2.0971367359161377,
      "learning_rate": 1.745e-05,
      "loss": 0.5684,
      "step": 350
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 1.8380221128463745,
      "learning_rate": 1.995e-05,
      "loss": 0.5472,
      "step": 400
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 3.9271857738494873,
      "learning_rate": 2.245e-05,
      "loss": 0.5287,
      "step": 450
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 7.809891700744629,
      "learning_rate": 2.495e-05,
      "loss": 0.5174,
      "step": 500
    },
    {
      "epoch": 0.837696335078534,
      "eval_loss": 0.4793977439403534,
      "eval_runtime": 265.0789,
      "eval_samples_per_second": 32.024,
      "eval_steps_per_second": 4.006,
      "step": 500
    },
    {
      "epoch": 0.9214659685863874,
      "grad_norm": 2.2309463024139404,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.5084,
      "step": 550
    },
    {
      "epoch": 1.0052356020942408,
      "grad_norm": 1.8079086542129517,
      "learning_rate": 2.995e-05,
      "loss": 0.4954,
      "step": 600
    },
    {
      "epoch": 1.0890052356020943,
      "grad_norm": 6.239879608154297,
      "learning_rate": 3.245e-05,
      "loss": 0.4954,
      "step": 650
    },
    {
      "epoch": 1.1727748691099475,
      "grad_norm": 12.593622207641602,
      "learning_rate": 3.495e-05,
      "loss": 0.4946,
      "step": 700
    },
    {
      "epoch": 1.256544502617801,
      "grad_norm": 3.1568186283111572,
      "learning_rate": 3.745e-05,
      "loss": 0.4768,
      "step": 750
    },
    {
      "epoch": 1.3403141361256545,
      "grad_norm": 3.9486255645751953,
      "learning_rate": 3.995e-05,
      "loss": 0.4737,
      "step": 800
    },
    {
      "epoch": 1.4240837696335078,
      "grad_norm": 2.641502618789673,
      "learning_rate": 4.245e-05,
      "loss": 0.4706,
      "step": 850
    },
    {
      "epoch": 1.5078534031413613,
      "grad_norm": 1.9798855781555176,
      "learning_rate": 4.495e-05,
      "loss": 0.4605,
      "step": 900
    },
    {
      "epoch": 1.5916230366492146,
      "grad_norm": 2.691363573074341,
      "learning_rate": 4.745e-05,
      "loss": 0.4554,
      "step": 950
    },
    {
      "epoch": 1.675392670157068,
      "grad_norm": 3.704902410507202,
      "learning_rate": 4.995e-05,
      "loss": 0.4561,
      "step": 1000
    },
    {
      "epoch": 1.675392670157068,
      "eval_loss": 0.402164101600647,
      "eval_runtime": 265.618,
      "eval_samples_per_second": 31.959,
      "eval_steps_per_second": 3.998,
      "step": 1000
    },
    {
      "epoch": 1.7591623036649215,
      "grad_norm": 4.97900390625,
      "learning_rate": 5.245e-05,
      "loss": 0.4553,
      "step": 1050
    },
    {
      "epoch": 1.8429319371727748,
      "grad_norm": 1.9889676570892334,
      "learning_rate": 5.495e-05,
      "loss": 0.449,
      "step": 1100
    },
    {
      "epoch": 1.9267015706806283,
      "grad_norm": 1.5135546922683716,
      "learning_rate": 5.745e-05,
      "loss": 0.4353,
      "step": 1150
    },
    {
      "epoch": 2.0104712041884816,
      "grad_norm": 7.610673904418945,
      "learning_rate": 5.995000000000001e-05,
      "loss": 0.4311,
      "step": 1200
    },
    {
      "epoch": 2.094240837696335,
      "grad_norm": 2.049562454223633,
      "learning_rate": 6.245000000000001e-05,
      "loss": 0.4312,
      "step": 1250
    },
    {
      "epoch": 2.1780104712041886,
      "grad_norm": 1.4102027416229248,
      "learning_rate": 6.494999999999999e-05,
      "loss": 0.4282,
      "step": 1300
    },
    {
      "epoch": 2.261780104712042,
      "grad_norm": 1.701119065284729,
      "learning_rate": 6.745e-05,
      "loss": 0.4272,
      "step": 1350
    },
    {
      "epoch": 2.345549738219895,
      "grad_norm": 2.0149667263031006,
      "learning_rate": 6.995e-05,
      "loss": 0.4277,
      "step": 1400
    },
    {
      "epoch": 2.4293193717277486,
      "grad_norm": 2.1658883094787598,
      "learning_rate": 7.245000000000001e-05,
      "loss": 0.4247,
      "step": 1450
    },
    {
      "epoch": 2.513089005235602,
      "grad_norm": 2.6821463108062744,
      "learning_rate": 7.495e-05,
      "loss": 0.4169,
      "step": 1500
    },
    {
      "epoch": 2.513089005235602,
      "eval_loss": 0.3853827118873596,
      "eval_runtime": 269.2607,
      "eval_samples_per_second": 31.527,
      "eval_steps_per_second": 3.944,
      "step": 1500
    },
    {
      "epoch": 2.5968586387434556,
      "grad_norm": 1.4221985340118408,
      "learning_rate": 7.745e-05,
      "loss": 0.4145,
      "step": 1550
    },
    {
      "epoch": 2.680628272251309,
      "grad_norm": 3.651655435562134,
      "learning_rate": 7.995e-05,
      "loss": 0.4129,
      "step": 1600
    },
    {
      "epoch": 2.7643979057591626,
      "grad_norm": 1.2700576782226562,
      "learning_rate": 8.245e-05,
      "loss": 0.4106,
      "step": 1650
    },
    {
      "epoch": 2.8481675392670156,
      "grad_norm": 1.8690059185028076,
      "learning_rate": 8.495e-05,
      "loss": 0.4134,
      "step": 1700
    },
    {
      "epoch": 2.931937172774869,
      "grad_norm": 2.483203649520874,
      "learning_rate": 8.745000000000001e-05,
      "loss": 0.413,
      "step": 1750
    },
    {
      "epoch": 3.0157068062827226,
      "grad_norm": 1.738501787185669,
      "learning_rate": 8.995e-05,
      "loss": 0.4063,
      "step": 1800
    },
    {
      "epoch": 3.099476439790576,
      "grad_norm": 1.837342619895935,
      "learning_rate": 9.245e-05,
      "loss": 0.4049,
      "step": 1850
    },
    {
      "epoch": 3.183246073298429,
      "grad_norm": 1.284974455833435,
      "learning_rate": 9.495e-05,
      "loss": 0.4065,
      "step": 1900
    },
    {
      "epoch": 3.2670157068062826,
      "grad_norm": 2.637281656265259,
      "learning_rate": 9.745000000000001e-05,
      "loss": 0.3986,
      "step": 1950
    },
    {
      "epoch": 3.350785340314136,
      "grad_norm": 1.4775941371917725,
      "learning_rate": 9.995e-05,
      "loss": 0.4054,
      "step": 2000
    },
    {
      "epoch": 3.350785340314136,
      "eval_loss": 0.36490994691848755,
      "eval_runtime": 265.9165,
      "eval_samples_per_second": 31.924,
      "eval_steps_per_second": 3.994,
      "step": 2000
    },
    {
      "epoch": 3.4345549738219896,
      "grad_norm": 2.121384859085083,
      "learning_rate": 9.918333333333334e-05,
      "loss": 0.4058,
      "step": 2050
    },
    {
      "epoch": 3.518324607329843,
      "grad_norm": 1.645984411239624,
      "learning_rate": 9.835e-05,
      "loss": 0.4021,
      "step": 2100
    },
    {
      "epoch": 3.6020942408376966,
      "grad_norm": 1.246239185333252,
      "learning_rate": 9.751666666666666e-05,
      "loss": 0.3991,
      "step": 2150
    },
    {
      "epoch": 3.6858638743455496,
      "grad_norm": 1.9096795320510864,
      "learning_rate": 9.668333333333334e-05,
      "loss": 0.3961,
      "step": 2200
    },
    {
      "epoch": 3.769633507853403,
      "grad_norm": 1.8867601156234741,
      "learning_rate": 9.585000000000001e-05,
      "loss": 0.3904,
      "step": 2250
    },
    {
      "epoch": 3.8534031413612566,
      "grad_norm": 1.7438101768493652,
      "learning_rate": 9.501666666666668e-05,
      "loss": 0.3895,
      "step": 2300
    },
    {
      "epoch": 3.93717277486911,
      "grad_norm": 1.1799490451812744,
      "learning_rate": 9.418333333333334e-05,
      "loss": 0.4027,
      "step": 2350
    },
    {
      "epoch": 4.020942408376963,
      "grad_norm": 1.1952763795852661,
      "learning_rate": 9.335e-05,
      "loss": 0.3893,
      "step": 2400
    },
    {
      "epoch": 4.104712041884817,
      "grad_norm": 2.008756160736084,
      "learning_rate": 9.251666666666667e-05,
      "loss": 0.3878,
      "step": 2450
    },
    {
      "epoch": 4.18848167539267,
      "grad_norm": 2.2693591117858887,
      "learning_rate": 9.168333333333333e-05,
      "loss": 0.3863,
      "step": 2500
    },
    {
      "epoch": 4.18848167539267,
      "eval_loss": 0.3528364896774292,
      "eval_runtime": 272.7627,
      "eval_samples_per_second": 31.122,
      "eval_steps_per_second": 3.893,
      "step": 2500
    },
    {
      "epoch": 4.272251308900524,
      "grad_norm": 1.0409568548202515,
      "learning_rate": 9.086666666666666e-05,
      "loss": 0.3894,
      "step": 2550
    },
    {
      "epoch": 4.356020942408377,
      "grad_norm": 2.047908306121826,
      "learning_rate": 9.003333333333333e-05,
      "loss": 0.3862,
      "step": 2600
    },
    {
      "epoch": 4.439790575916231,
      "grad_norm": 1.767168641090393,
      "learning_rate": 8.92e-05,
      "loss": 0.3832,
      "step": 2650
    },
    {
      "epoch": 4.523560209424084,
      "grad_norm": 1.1918233633041382,
      "learning_rate": 8.836666666666667e-05,
      "loss": 0.3839,
      "step": 2700
    },
    {
      "epoch": 4.607329842931938,
      "grad_norm": 0.9553185105323792,
      "learning_rate": 8.753333333333334e-05,
      "loss": 0.3807,
      "step": 2750
    },
    {
      "epoch": 4.69109947643979,
      "grad_norm": 1.1633609533309937,
      "learning_rate": 8.67e-05,
      "loss": 0.3836,
      "step": 2800
    },
    {
      "epoch": 4.774869109947644,
      "grad_norm": 1.1633822917938232,
      "learning_rate": 8.586666666666668e-05,
      "loss": 0.3815,
      "step": 2850
    },
    {
      "epoch": 4.858638743455497,
      "grad_norm": 1.7152239084243774,
      "learning_rate": 8.503333333333334e-05,
      "loss": 0.3809,
      "step": 2900
    },
    {
      "epoch": 4.942408376963351,
      "grad_norm": 1.2379584312438965,
      "learning_rate": 8.42e-05,
      "loss": 0.3801,
      "step": 2950
    },
    {
      "epoch": 5.026178010471204,
      "grad_norm": 1.1686451435089111,
      "learning_rate": 8.336666666666667e-05,
      "loss": 0.3807,
      "step": 3000
    },
    {
      "epoch": 5.026178010471204,
      "eval_loss": 0.3480445444583893,
      "eval_runtime": 272.7064,
      "eval_samples_per_second": 31.129,
      "eval_steps_per_second": 3.894,
      "step": 3000
    },
    {
      "epoch": 5.109947643979058,
      "grad_norm": 1.5680384635925293,
      "learning_rate": 8.253333333333334e-05,
      "loss": 0.3739,
      "step": 3050
    },
    {
      "epoch": 5.193717277486911,
      "grad_norm": 1.5168578624725342,
      "learning_rate": 8.17e-05,
      "loss": 0.3756,
      "step": 3100
    },
    {
      "epoch": 5.277486910994765,
      "grad_norm": 1.0925545692443848,
      "learning_rate": 8.086666666666666e-05,
      "loss": 0.3739,
      "step": 3150
    },
    {
      "epoch": 5.361256544502618,
      "grad_norm": 0.8022084832191467,
      "learning_rate": 8.003333333333333e-05,
      "loss": 0.3736,
      "step": 3200
    },
    {
      "epoch": 5.445026178010472,
      "grad_norm": 1.6964820623397827,
      "learning_rate": 7.920000000000001e-05,
      "loss": 0.379,
      "step": 3250
    },
    {
      "epoch": 5.528795811518324,
      "grad_norm": 1.0717741250991821,
      "learning_rate": 7.836666666666667e-05,
      "loss": 0.3762,
      "step": 3300
    },
    {
      "epoch": 5.612565445026178,
      "grad_norm": 1.2660809755325317,
      "learning_rate": 7.753333333333334e-05,
      "loss": 0.3708,
      "step": 3350
    },
    {
      "epoch": 5.696335078534031,
      "grad_norm": 1.6065802574157715,
      "learning_rate": 7.670000000000001e-05,
      "loss": 0.3717,
      "step": 3400
    },
    {
      "epoch": 5.780104712041885,
      "grad_norm": 1.2089744806289673,
      "learning_rate": 7.586666666666668e-05,
      "loss": 0.3715,
      "step": 3450
    },
    {
      "epoch": 5.863874345549738,
      "grad_norm": 3.1524300575256348,
      "learning_rate": 7.503333333333333e-05,
      "loss": 0.3708,
      "step": 3500
    },
    {
      "epoch": 5.863874345549738,
      "eval_loss": 0.3431868255138397,
      "eval_runtime": 272.9384,
      "eval_samples_per_second": 31.102,
      "eval_steps_per_second": 3.891,
      "step": 3500
    },
    {
      "epoch": 5.947643979057592,
      "grad_norm": 1.914294719696045,
      "learning_rate": 7.42e-05,
      "loss": 0.3673,
      "step": 3550
    },
    {
      "epoch": 6.031413612565445,
      "grad_norm": 1.3186005353927612,
      "learning_rate": 7.336666666666667e-05,
      "loss": 0.3665,
      "step": 3600
    },
    {
      "epoch": 6.115183246073299,
      "grad_norm": 2.019273519515991,
      "learning_rate": 7.253333333333334e-05,
      "loss": 0.3697,
      "step": 3650
    },
    {
      "epoch": 6.198952879581152,
      "grad_norm": 1.0517597198486328,
      "learning_rate": 7.17e-05,
      "loss": 0.3674,
      "step": 3700
    },
    {
      "epoch": 6.282722513089006,
      "grad_norm": 1.0202686786651611,
      "learning_rate": 7.086666666666666e-05,
      "loss": 0.3706,
      "step": 3750
    },
    {
      "epoch": 6.366492146596858,
      "grad_norm": 1.4179818630218506,
      "learning_rate": 7.003333333333335e-05,
      "loss": 0.3681,
      "step": 3800
    },
    {
      "epoch": 6.450261780104712,
      "grad_norm": 1.3820505142211914,
      "learning_rate": 6.92e-05,
      "loss": 0.3671,
      "step": 3850
    },
    {
      "epoch": 6.534031413612565,
      "grad_norm": 1.3857202529907227,
      "learning_rate": 6.836666666666667e-05,
      "loss": 0.3641,
      "step": 3900
    },
    {
      "epoch": 6.617801047120419,
      "grad_norm": 1.0996108055114746,
      "learning_rate": 6.753333333333334e-05,
      "loss": 0.3685,
      "step": 3950
    },
    {
      "epoch": 6.701570680628272,
      "grad_norm": 0.9405946731567383,
      "learning_rate": 6.670000000000001e-05,
      "loss": 0.366,
      "step": 4000
    },
    {
      "epoch": 6.701570680628272,
      "eval_loss": 0.33282962441444397,
      "eval_runtime": 271.6112,
      "eval_samples_per_second": 31.254,
      "eval_steps_per_second": 3.91,
      "step": 4000
    },
    {
      "epoch": 6.785340314136126,
      "grad_norm": 0.9167115092277527,
      "learning_rate": 6.586666666666666e-05,
      "loss": 0.3618,
      "step": 4050
    },
    {
      "epoch": 6.869109947643979,
      "grad_norm": 1.0496320724487305,
      "learning_rate": 6.503333333333333e-05,
      "loss": 0.3624,
      "step": 4100
    },
    {
      "epoch": 6.952879581151833,
      "grad_norm": 1.7927188873291016,
      "learning_rate": 6.42e-05,
      "loss": 0.3621,
      "step": 4150
    },
    {
      "epoch": 7.036649214659686,
      "grad_norm": 1.3362383842468262,
      "learning_rate": 6.336666666666667e-05,
      "loss": 0.3631,
      "step": 4200
    },
    {
      "epoch": 7.12041884816754,
      "grad_norm": 1.5686593055725098,
      "learning_rate": 6.253333333333333e-05,
      "loss": 0.3672,
      "step": 4250
    },
    {
      "epoch": 7.204188481675392,
      "grad_norm": 2.042232036590576,
      "learning_rate": 6.170000000000001e-05,
      "loss": 0.3598,
      "step": 4300
    },
    {
      "epoch": 7.287958115183246,
      "grad_norm": 1.2040599584579468,
      "learning_rate": 6.086666666666667e-05,
      "loss": 0.3628,
      "step": 4350
    },
    {
      "epoch": 7.371727748691099,
      "grad_norm": 1.3659878969192505,
      "learning_rate": 6.003333333333334e-05,
      "loss": 0.3626,
      "step": 4400
    },
    {
      "epoch": 7.455497382198953,
      "grad_norm": 1.9217562675476074,
      "learning_rate": 5.92e-05,
      "loss": 0.3636,
      "step": 4450
    },
    {
      "epoch": 7.539267015706806,
      "grad_norm": 1.5920127630233765,
      "learning_rate": 5.836666666666667e-05,
      "loss": 0.3575,
      "step": 4500
    },
    {
      "epoch": 7.539267015706806,
      "eval_loss": 0.33003953099250793,
      "eval_runtime": 274.282,
      "eval_samples_per_second": 30.95,
      "eval_steps_per_second": 3.872,
      "step": 4500
    },
    {
      "epoch": 7.62303664921466,
      "grad_norm": 1.3059043884277344,
      "learning_rate": 5.753333333333334e-05,
      "loss": 0.3603,
      "step": 4550
    },
    {
      "epoch": 7.706806282722513,
      "grad_norm": 1.6313170194625854,
      "learning_rate": 5.6699999999999996e-05,
      "loss": 0.3622,
      "step": 4600
    },
    {
      "epoch": 7.790575916230367,
      "grad_norm": 1.7790875434875488,
      "learning_rate": 5.5866666666666665e-05,
      "loss": 0.3635,
      "step": 4650
    },
    {
      "epoch": 7.87434554973822,
      "grad_norm": 1.1874761581420898,
      "learning_rate": 5.5033333333333334e-05,
      "loss": 0.3635,
      "step": 4700
    },
    {
      "epoch": 7.958115183246074,
      "grad_norm": 0.9745219349861145,
      "learning_rate": 5.420000000000001e-05,
      "loss": 0.357,
      "step": 4750
    },
    {
      "epoch": 8.041884816753926,
      "grad_norm": 1.3074321746826172,
      "learning_rate": 5.3366666666666665e-05,
      "loss": 0.3543,
      "step": 4800
    },
    {
      "epoch": 8.12565445026178,
      "grad_norm": 1.3384989500045776,
      "learning_rate": 5.2533333333333334e-05,
      "loss": 0.3556,
      "step": 4850
    },
    {
      "epoch": 8.209424083769633,
      "grad_norm": 1.0133076906204224,
      "learning_rate": 5.17e-05,
      "loss": 0.3654,
      "step": 4900
    },
    {
      "epoch": 8.293193717277488,
      "grad_norm": 1.194360375404358,
      "learning_rate": 5.086666666666667e-05,
      "loss": 0.3547,
      "step": 4950
    },
    {
      "epoch": 8.37696335078534,
      "grad_norm": 1.049672245979309,
      "learning_rate": 5.0033333333333334e-05,
      "loss": 0.3607,
      "step": 5000
    },
    {
      "epoch": 8.37696335078534,
      "eval_loss": 0.3284249007701874,
      "eval_runtime": 274.7134,
      "eval_samples_per_second": 30.901,
      "eval_steps_per_second": 3.866,
      "step": 5000
    },
    {
      "epoch": 8.460732984293193,
      "grad_norm": 1.1301287412643433,
      "learning_rate": 4.92e-05,
      "loss": 0.3542,
      "step": 5050
    },
    {
      "epoch": 8.544502617801047,
      "grad_norm": 1.0920559167861938,
      "learning_rate": 4.836666666666667e-05,
      "loss": 0.3534,
      "step": 5100
    },
    {
      "epoch": 8.6282722513089,
      "grad_norm": 1.1312081813812256,
      "learning_rate": 4.7533333333333334e-05,
      "loss": 0.3598,
      "step": 5150
    },
    {
      "epoch": 8.712041884816754,
      "grad_norm": 1.5819182395935059,
      "learning_rate": 4.6700000000000003e-05,
      "loss": 0.3537,
      "step": 5200
    },
    {
      "epoch": 8.795811518324607,
      "grad_norm": 1.0059967041015625,
      "learning_rate": 4.5866666666666666e-05,
      "loss": 0.3553,
      "step": 5250
    },
    {
      "epoch": 8.879581151832461,
      "grad_norm": 1.1407588720321655,
      "learning_rate": 4.5033333333333335e-05,
      "loss": 0.3557,
      "step": 5300
    },
    {
      "epoch": 8.963350785340314,
      "grad_norm": 1.296221137046814,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.3545,
      "step": 5350
    },
    {
      "epoch": 9.047120418848168,
      "grad_norm": 1.4306052923202515,
      "learning_rate": 4.3366666666666666e-05,
      "loss": 0.3583,
      "step": 5400
    },
    {
      "epoch": 9.13089005235602,
      "grad_norm": 1.1458420753479004,
      "learning_rate": 4.2533333333333335e-05,
      "loss": 0.3503,
      "step": 5450
    },
    {
      "epoch": 9.214659685863875,
      "grad_norm": 0.9508205056190491,
      "learning_rate": 4.17e-05,
      "loss": 0.352,
      "step": 5500
    },
    {
      "epoch": 9.214659685863875,
      "eval_loss": 0.3231986165046692,
      "eval_runtime": 273.3343,
      "eval_samples_per_second": 31.057,
      "eval_steps_per_second": 3.885,
      "step": 5500
    },
    {
      "epoch": 9.298429319371728,
      "grad_norm": 2.127049446105957,
      "learning_rate": 4.086666666666667e-05,
      "loss": 0.3547,
      "step": 5550
    },
    {
      "epoch": 9.38219895287958,
      "grad_norm": 1.5531483888626099,
      "learning_rate": 4.0033333333333335e-05,
      "loss": 0.3576,
      "step": 5600
    },
    {
      "epoch": 9.465968586387435,
      "grad_norm": 1.5712753534317017,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.3492,
      "step": 5650
    },
    {
      "epoch": 9.549738219895287,
      "grad_norm": 1.1668370962142944,
      "learning_rate": 3.8366666666666666e-05,
      "loss": 0.3529,
      "step": 5700
    },
    {
      "epoch": 9.633507853403142,
      "grad_norm": 1.1639858484268188,
      "learning_rate": 3.7533333333333335e-05,
      "loss": 0.3526,
      "step": 5750
    },
    {
      "epoch": 9.717277486910994,
      "grad_norm": 1.4188268184661865,
      "learning_rate": 3.6700000000000004e-05,
      "loss": 0.3527,
      "step": 5800
    },
    {
      "epoch": 9.801047120418849,
      "grad_norm": 1.8606890439987183,
      "learning_rate": 3.586666666666667e-05,
      "loss": 0.3517,
      "step": 5850
    },
    {
      "epoch": 9.884816753926701,
      "grad_norm": 1.2933155298233032,
      "learning_rate": 3.5033333333333336e-05,
      "loss": 0.3463,
      "step": 5900
    },
    {
      "epoch": 9.968586387434556,
      "grad_norm": 1.4728389978408813,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.3505,
      "step": 5950
    },
    {
      "epoch": 10.052356020942408,
      "grad_norm": 1.6055335998535156,
      "learning_rate": 3.336666666666667e-05,
      "loss": 0.3471,
      "step": 6000
    },
    {
      "epoch": 10.052356020942408,
      "eval_loss": 0.3189197778701782,
      "eval_runtime": 273.3525,
      "eval_samples_per_second": 31.055,
      "eval_steps_per_second": 3.885,
      "step": 6000
    },
    {
      "epoch": 10.136125654450261,
      "grad_norm": 1.572178840637207,
      "learning_rate": 3.253333333333333e-05,
      "loss": 0.3492,
      "step": 6050
    },
    {
      "epoch": 10.219895287958115,
      "grad_norm": 1.063959002494812,
      "learning_rate": 3.1700000000000005e-05,
      "loss": 0.3525,
      "step": 6100
    },
    {
      "epoch": 10.303664921465968,
      "grad_norm": 1.1579703092575073,
      "learning_rate": 3.086666666666667e-05,
      "loss": 0.3486,
      "step": 6150
    },
    {
      "epoch": 10.387434554973822,
      "grad_norm": 1.4260714054107666,
      "learning_rate": 3.0033333333333336e-05,
      "loss": 0.3483,
      "step": 6200
    },
    {
      "epoch": 10.471204188481675,
      "grad_norm": 1.453321099281311,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 0.3481,
      "step": 6250
    },
    {
      "epoch": 10.55497382198953,
      "grad_norm": 1.8545498847961426,
      "learning_rate": 2.836666666666667e-05,
      "loss": 0.3482,
      "step": 6300
    },
    {
      "epoch": 10.638743455497382,
      "grad_norm": 1.073957920074463,
      "learning_rate": 2.7533333333333333e-05,
      "loss": 0.348,
      "step": 6350
    },
    {
      "epoch": 10.722513089005236,
      "grad_norm": 1.0049316883087158,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 0.3487,
      "step": 6400
    },
    {
      "epoch": 10.806282722513089,
      "grad_norm": 1.4970500469207764,
      "learning_rate": 2.5866666666666667e-05,
      "loss": 0.3468,
      "step": 6450
    },
    {
      "epoch": 10.890052356020943,
      "grad_norm": 1.6566526889801025,
      "learning_rate": 2.5033333333333336e-05,
      "loss": 0.3469,
      "step": 6500
    },
    {
      "epoch": 10.890052356020943,
      "eval_loss": 0.31707289814949036,
      "eval_runtime": 275.2143,
      "eval_samples_per_second": 30.845,
      "eval_steps_per_second": 3.859,
      "step": 6500
    },
    {
      "epoch": 10.973821989528796,
      "grad_norm": 1.1960996389389038,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.3478,
      "step": 6550
    },
    {
      "epoch": 11.057591623036648,
      "grad_norm": 0.97001713514328,
      "learning_rate": 2.3366666666666668e-05,
      "loss": 0.3475,
      "step": 6600
    },
    {
      "epoch": 11.141361256544503,
      "grad_norm": 1.1384519338607788,
      "learning_rate": 2.2533333333333333e-05,
      "loss": 0.3484,
      "step": 6650
    },
    {
      "epoch": 11.225130890052355,
      "grad_norm": 0.9649496078491211,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 0.3454,
      "step": 6700
    },
    {
      "epoch": 11.30890052356021,
      "grad_norm": 1.0407809019088745,
      "learning_rate": 2.0866666666666668e-05,
      "loss": 0.3446,
      "step": 6750
    },
    {
      "epoch": 11.392670157068062,
      "grad_norm": 1.087108850479126,
      "learning_rate": 2.0033333333333334e-05,
      "loss": 0.3475,
      "step": 6800
    },
    {
      "epoch": 11.476439790575917,
      "grad_norm": 0.8870049715042114,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.3454,
      "step": 6850
    },
    {
      "epoch": 11.56020942408377,
      "grad_norm": 1.0377373695373535,
      "learning_rate": 1.8366666666666668e-05,
      "loss": 0.3447,
      "step": 6900
    },
    {
      "epoch": 11.643979057591624,
      "grad_norm": 1.138604760169983,
      "learning_rate": 1.7533333333333334e-05,
      "loss": 0.345,
      "step": 6950
    },
    {
      "epoch": 11.727748691099476,
      "grad_norm": 1.6464053392410278,
      "learning_rate": 1.6700000000000003e-05,
      "loss": 0.3465,
      "step": 7000
    },
    {
      "epoch": 11.727748691099476,
      "eval_loss": 0.31611359119415283,
      "eval_runtime": 271.5426,
      "eval_samples_per_second": 31.262,
      "eval_steps_per_second": 3.911,
      "step": 7000
    },
    {
      "epoch": 11.81151832460733,
      "grad_norm": 0.9007428288459778,
      "learning_rate": 1.586666666666667e-05,
      "loss": 0.3461,
      "step": 7050
    },
    {
      "epoch": 11.895287958115183,
      "grad_norm": 1.1172393560409546,
      "learning_rate": 1.5033333333333336e-05,
      "loss": 0.3489,
      "step": 7100
    },
    {
      "epoch": 11.979057591623036,
      "grad_norm": 1.0454591512680054,
      "learning_rate": 1.42e-05,
      "loss": 0.3429,
      "step": 7150
    },
    {
      "epoch": 12.06282722513089,
      "grad_norm": 1.1494097709655762,
      "learning_rate": 1.3383333333333335e-05,
      "loss": 0.3523,
      "step": 7200
    },
    {
      "epoch": 12.146596858638743,
      "grad_norm": 0.8459360003471375,
      "learning_rate": 1.255e-05,
      "loss": 0.3469,
      "step": 7250
    },
    {
      "epoch": 12.230366492146597,
      "grad_norm": 1.034494161605835,
      "learning_rate": 1.1716666666666667e-05,
      "loss": 0.3451,
      "step": 7300
    },
    {
      "epoch": 12.31413612565445,
      "grad_norm": 1.1340311765670776,
      "learning_rate": 1.0883333333333335e-05,
      "loss": 0.3441,
      "step": 7350
    },
    {
      "epoch": 12.397905759162304,
      "grad_norm": 1.4063786268234253,
      "learning_rate": 1.005e-05,
      "loss": 0.3472,
      "step": 7400
    },
    {
      "epoch": 12.481675392670157,
      "grad_norm": 1.271904468536377,
      "learning_rate": 9.216666666666666e-06,
      "loss": 0.3458,
      "step": 7450
    },
    {
      "epoch": 12.565445026178011,
      "grad_norm": 1.2097103595733643,
      "learning_rate": 8.383333333333333e-06,
      "loss": 0.3443,
      "step": 7500
    },
    {
      "epoch": 12.565445026178011,
      "eval_loss": 0.3158954679965973,
      "eval_runtime": 271.3497,
      "eval_samples_per_second": 31.284,
      "eval_steps_per_second": 3.914,
      "step": 7500
    },
    {
      "epoch": 12.649214659685864,
      "grad_norm": 1.0162887573242188,
      "learning_rate": 7.55e-06,
      "loss": 0.3432,
      "step": 7550
    },
    {
      "epoch": 12.732984293193716,
      "grad_norm": 0.977488100528717,
      "learning_rate": 6.716666666666667e-06,
      "loss": 0.3457,
      "step": 7600
    },
    {
      "epoch": 12.81675392670157,
      "grad_norm": 0.8293308019638062,
      "learning_rate": 5.8833333333333335e-06,
      "loss": 0.3445,
      "step": 7650
    },
    {
      "epoch": 12.900523560209423,
      "grad_norm": 0.9540147185325623,
      "learning_rate": 5.050000000000001e-06,
      "loss": 0.3444,
      "step": 7700
    },
    {
      "epoch": 12.984293193717278,
      "grad_norm": 1.6243274211883545,
      "learning_rate": 4.216666666666666e-06,
      "loss": 0.3481,
      "step": 7750
    },
    {
      "epoch": 13.06806282722513,
      "grad_norm": 0.8080894351005554,
      "learning_rate": 3.3833333333333337e-06,
      "loss": 0.3416,
      "step": 7800
    },
    {
      "epoch": 13.151832460732985,
      "grad_norm": 0.8952440023422241,
      "learning_rate": 2.55e-06,
      "loss": 0.3423,
      "step": 7850
    },
    {
      "epoch": 13.235602094240837,
      "grad_norm": 0.6296280026435852,
      "learning_rate": 1.7166666666666668e-06,
      "loss": 0.3454,
      "step": 7900
    },
    {
      "epoch": 13.319371727748692,
      "grad_norm": 1.5492085218429565,
      "learning_rate": 8.833333333333334e-07,
      "loss": 0.344,
      "step": 7950
    },
    {
      "epoch": 13.403141361256544,
      "grad_norm": 0.6316419839859009,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.3432,
      "step": 8000
    },
    {
      "epoch": 13.403141361256544,
      "eval_loss": 0.31398773193359375,
      "eval_runtime": 271.4112,
      "eval_samples_per_second": 31.277,
      "eval_steps_per_second": 3.913,
      "step": 8000
    }
  ],
  "logging_steps": 50,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 14,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4332007404692806e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}