{
  "best_metric": 0.3861308693885803,
  "best_model_checkpoint": "mikhail-panzo/zlm_b64_le5_s8000/checkpoint-5000",
  "epoch": 4.187604690117253,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04187604690117253,
      "grad_norm": 5.415271759033203,
      "learning_rate": 2.4000000000000003e-07,
      "loss": 1.1819,
      "step": 50
    },
    {
      "epoch": 0.08375209380234507,
      "grad_norm": 16.256921768188477,
      "learning_rate": 4.900000000000001e-07,
      "loss": 1.0035,
      "step": 100
    },
    {
      "epoch": 0.12562814070351758,
      "grad_norm": 3.2439260482788086,
      "learning_rate": 7.350000000000001e-07,
      "loss": 0.9742,
      "step": 150
    },
    {
      "epoch": 0.16750418760469013,
      "grad_norm": 4.24526834487915,
      "learning_rate": 9.85e-07,
      "loss": 0.9361,
      "step": 200
    },
    {
      "epoch": 0.20938023450586266,
      "grad_norm": 3.625537872314453,
      "learning_rate": 1.235e-06,
      "loss": 0.8733,
      "step": 250
    },
    {
      "epoch": 0.25125628140703515,
      "grad_norm": 4.1455464363098145,
      "learning_rate": 1.485e-06,
      "loss": 0.8471,
      "step": 300
    },
    {
      "epoch": 0.2931323283082077,
      "grad_norm": 3.164158582687378,
      "learning_rate": 1.7350000000000001e-06,
      "loss": 0.838,
      "step": 350
    },
    {
      "epoch": 0.33500837520938026,
      "grad_norm": 4.24931001663208,
      "learning_rate": 1.985e-06,
      "loss": 0.7467,
      "step": 400
    },
    {
      "epoch": 0.3768844221105528,
      "grad_norm": 2.332688331604004,
      "learning_rate": 2.235e-06,
      "loss": 0.7389,
      "step": 450
    },
    {
      "epoch": 0.4187604690117253,
      "grad_norm": 2.370961904525757,
      "learning_rate": 2.4850000000000003e-06,
      "loss": 0.7074,
      "step": 500
    },
    {
      "epoch": 0.4187604690117253,
      "eval_loss": 0.6029486060142517,
      "eval_runtime": 220.2253,
      "eval_samples_per_second": 38.547,
      "eval_steps_per_second": 4.822,
      "step": 500
    },
    {
      "epoch": 0.46063651591289784,
      "grad_norm": 14.466177940368652,
      "learning_rate": 2.7350000000000005e-06,
      "loss": 0.6738,
      "step": 550
    },
    {
      "epoch": 0.5025125628140703,
      "grad_norm": 3.282449722290039,
      "learning_rate": 2.9850000000000002e-06,
      "loss": 0.6517,
      "step": 600
    },
    {
      "epoch": 0.5443886097152428,
      "grad_norm": 2.621645212173462,
      "learning_rate": 3.2350000000000004e-06,
      "loss": 0.6479,
      "step": 650
    },
    {
      "epoch": 0.5862646566164154,
      "grad_norm": 2.4493019580841064,
      "learning_rate": 3.485e-06,
      "loss": 0.6321,
      "step": 700
    },
    {
      "epoch": 0.628140703517588,
      "grad_norm": 1.919048547744751,
      "learning_rate": 3.7350000000000002e-06,
      "loss": 0.6407,
      "step": 750
    },
    {
      "epoch": 0.6700167504187605,
      "grad_norm": 2.8260602951049805,
      "learning_rate": 3.985000000000001e-06,
      "loss": 0.6279,
      "step": 800
    },
    {
      "epoch": 0.711892797319933,
      "grad_norm": 3.333841323852539,
      "learning_rate": 4.235e-06,
      "loss": 0.6143,
      "step": 850
    },
    {
      "epoch": 0.7537688442211056,
      "grad_norm": 1.7883094549179077,
      "learning_rate": 4.485e-06,
      "loss": 0.5949,
      "step": 900
    },
    {
      "epoch": 0.7956448911222781,
      "grad_norm": 2.48634934425354,
      "learning_rate": 4.735e-06,
      "loss": 0.5875,
      "step": 950
    },
    {
      "epoch": 0.8375209380234506,
      "grad_norm": 5.4277873039245605,
      "learning_rate": 4.9850000000000006e-06,
      "loss": 0.5916,
      "step": 1000
    },
    {
      "epoch": 0.8375209380234506,
      "eval_loss": 0.49684029817581177,
      "eval_runtime": 219.2621,
      "eval_samples_per_second": 38.716,
      "eval_steps_per_second": 4.844,
      "step": 1000
    },
    {
      "epoch": 0.8793969849246231,
      "grad_norm": 2.675995349884033,
      "learning_rate": 5.235e-06,
      "loss": 0.5676,
      "step": 1050
    },
    {
      "epoch": 0.9212730318257957,
      "grad_norm": 2.507474422454834,
      "learning_rate": 5.485e-06,
      "loss": 0.5535,
      "step": 1100
    },
    {
      "epoch": 0.9631490787269682,
      "grad_norm": 5.360454082489014,
      "learning_rate": 5.735e-06,
      "loss": 0.569,
      "step": 1150
    },
    {
      "epoch": 1.0050251256281406,
      "grad_norm": 2.9445533752441406,
      "learning_rate": 5.985000000000001e-06,
      "loss": 0.5385,
      "step": 1200
    },
    {
      "epoch": 1.0469011725293131,
      "grad_norm": 2.3764073848724365,
      "learning_rate": 6.235000000000001e-06,
      "loss": 0.5647,
      "step": 1250
    },
    {
      "epoch": 1.0887772194304857,
      "grad_norm": 4.364608287811279,
      "learning_rate": 6.485000000000001e-06,
      "loss": 0.5554,
      "step": 1300
    },
    {
      "epoch": 1.1306532663316582,
      "grad_norm": 4.688488006591797,
      "learning_rate": 6.735000000000001e-06,
      "loss": 0.5676,
      "step": 1350
    },
    {
      "epoch": 1.1725293132328307,
      "grad_norm": 3.0410969257354736,
      "learning_rate": 6.985000000000001e-06,
      "loss": 0.5374,
      "step": 1400
    },
    {
      "epoch": 1.2144053601340032,
      "grad_norm": 6.28264856338501,
      "learning_rate": 7.235000000000001e-06,
      "loss": 0.5275,
      "step": 1450
    },
    {
      "epoch": 1.2562814070351758,
      "grad_norm": 3.7520318031311035,
      "learning_rate": 7.485000000000001e-06,
      "loss": 0.5206,
      "step": 1500
    },
    {
      "epoch": 1.2562814070351758,
      "eval_loss": 0.4592318534851074,
      "eval_runtime": 220.2748,
      "eval_samples_per_second": 38.538,
      "eval_steps_per_second": 4.821,
      "step": 1500
    },
    {
      "epoch": 1.2981574539363483,
      "grad_norm": 2.9252817630767822,
      "learning_rate": 7.735e-06,
      "loss": 0.5175,
      "step": 1550
    },
    {
      "epoch": 1.3400335008375208,
      "grad_norm": 2.6018381118774414,
      "learning_rate": 7.985e-06,
      "loss": 0.5191,
      "step": 1600
    },
    {
      "epoch": 1.3819095477386933,
      "grad_norm": 2.215420722961426,
      "learning_rate": 8.235e-06,
      "loss": 0.5274,
      "step": 1650
    },
    {
      "epoch": 1.4237855946398659,
      "grad_norm": 2.787269353866577,
      "learning_rate": 8.485000000000001e-06,
      "loss": 0.5201,
      "step": 1700
    },
    {
      "epoch": 1.4656616415410384,
      "grad_norm": 3.644702434539795,
      "learning_rate": 8.735000000000002e-06,
      "loss": 0.5157,
      "step": 1750
    },
    {
      "epoch": 1.507537688442211,
      "grad_norm": 2.4051408767700195,
      "learning_rate": 8.985000000000001e-06,
      "loss": 0.5091,
      "step": 1800
    },
    {
      "epoch": 1.5494137353433834,
      "grad_norm": 3.729388952255249,
      "learning_rate": 9.235e-06,
      "loss": 0.5134,
      "step": 1850
    },
    {
      "epoch": 1.591289782244556,
      "grad_norm": 3.360603094100952,
      "learning_rate": 9.485000000000002e-06,
      "loss": 0.5041,
      "step": 1900
    },
    {
      "epoch": 1.6331658291457285,
      "grad_norm": 3.6210439205169678,
      "learning_rate": 9.735e-06,
      "loss": 0.4993,
      "step": 1950
    },
    {
      "epoch": 1.675041876046901,
      "grad_norm": 2.6245505809783936,
      "learning_rate": 9.985000000000002e-06,
      "loss": 0.4979,
      "step": 2000
    },
    {
      "epoch": 1.675041876046901,
      "eval_loss": 0.4388435482978821,
      "eval_runtime": 220.5956,
      "eval_samples_per_second": 38.482,
      "eval_steps_per_second": 4.814,
      "step": 2000
    },
    {
      "epoch": 1.7169179229480735,
      "grad_norm": 3.5152196884155273,
      "learning_rate": 9.921666666666667e-06,
      "loss": 0.4917,
      "step": 2050
    },
    {
      "epoch": 1.758793969849246,
      "grad_norm": 2.4508583545684814,
      "learning_rate": 9.838333333333334e-06,
      "loss": 0.5161,
      "step": 2100
    },
    {
      "epoch": 1.8006700167504186,
      "grad_norm": 3.44022536277771,
      "learning_rate": 9.755e-06,
      "loss": 0.4933,
      "step": 2150
    },
    {
      "epoch": 1.8425460636515911,
      "grad_norm": 2.480679512023926,
      "learning_rate": 9.671666666666668e-06,
      "loss": 0.4983,
      "step": 2200
    },
    {
      "epoch": 1.8844221105527639,
      "grad_norm": 2.595881462097168,
      "learning_rate": 9.588333333333334e-06,
      "loss": 0.4797,
      "step": 2250
    },
    {
      "epoch": 1.9262981574539364,
      "grad_norm": 2.0328593254089355,
      "learning_rate": 9.505000000000001e-06,
      "loss": 0.483,
      "step": 2300
    },
    {
      "epoch": 1.968174204355109,
      "grad_norm": 3.7075388431549072,
      "learning_rate": 9.421666666666668e-06,
      "loss": 0.4922,
      "step": 2350
    },
    {
      "epoch": 2.0100502512562812,
      "grad_norm": 2.3989317417144775,
      "learning_rate": 9.338333333333333e-06,
      "loss": 0.4846,
      "step": 2400
    },
    {
      "epoch": 2.0519262981574538,
      "grad_norm": 2.803882360458374,
      "learning_rate": 9.255e-06,
      "loss": 0.4832,
      "step": 2450
    },
    {
      "epoch": 2.0938023450586263,
      "grad_norm": 3.1467437744140625,
      "learning_rate": 9.171666666666667e-06,
      "loss": 0.4852,
      "step": 2500
    },
    {
      "epoch": 2.0938023450586263,
      "eval_loss": 0.4210895597934723,
      "eval_runtime": 222.2617,
      "eval_samples_per_second": 38.194,
      "eval_steps_per_second": 4.778,
      "step": 2500
    },
    {
      "epoch": 2.135678391959799,
      "grad_norm": 3.0684900283813477,
      "learning_rate": 9.088333333333334e-06,
      "loss": 0.4787,
      "step": 2550
    },
    {
      "epoch": 2.1775544388609713,
      "grad_norm": 2.283045768737793,
      "learning_rate": 9.005000000000001e-06,
      "loss": 0.4781,
      "step": 2600
    },
    {
      "epoch": 2.219430485762144,
      "grad_norm": 2.2983710765838623,
      "learning_rate": 8.921666666666668e-06,
      "loss": 0.4745,
      "step": 2650
    },
    {
      "epoch": 2.2613065326633164,
      "grad_norm": 2.7612133026123047,
      "learning_rate": 8.838333333333335e-06,
      "loss": 0.4659,
      "step": 2700
    },
    {
      "epoch": 2.3031825795644894,
      "grad_norm": 1.8307946920394897,
      "learning_rate": 8.755e-06,
      "loss": 0.4741,
      "step": 2750
    },
    {
      "epoch": 2.3450586264656614,
      "grad_norm": 2.60187029838562,
      "learning_rate": 8.671666666666667e-06,
      "loss": 0.4603,
      "step": 2800
    },
    {
      "epoch": 2.3869346733668344,
      "grad_norm": 1.789478063583374,
      "learning_rate": 8.588333333333334e-06,
      "loss": 0.4587,
      "step": 2850
    },
    {
      "epoch": 2.4288107202680065,
      "grad_norm": 2.2422680854797363,
      "learning_rate": 8.505e-06,
      "loss": 0.4614,
      "step": 2900
    },
    {
      "epoch": 2.4706867671691795,
      "grad_norm": 2.4102437496185303,
      "learning_rate": 8.421666666666668e-06,
      "loss": 0.4618,
      "step": 2950
    },
    {
      "epoch": 2.5125628140703515,
      "grad_norm": 1.9738932847976685,
      "learning_rate": 8.338333333333335e-06,
      "loss": 0.4615,
      "step": 3000
    },
    {
      "epoch": 2.5125628140703515,
      "eval_loss": 0.4087553322315216,
      "eval_runtime": 221.1861,
      "eval_samples_per_second": 38.379,
      "eval_steps_per_second": 4.801,
      "step": 3000
    },
    {
      "epoch": 2.5544388609715245,
      "grad_norm": 2.5020742416381836,
      "learning_rate": 8.255000000000001e-06,
      "loss": 0.4505,
      "step": 3050
    },
    {
      "epoch": 2.5963149078726966,
      "grad_norm": 1.6484034061431885,
      "learning_rate": 8.171666666666668e-06,
      "loss": 0.4548,
      "step": 3100
    },
    {
      "epoch": 2.6381909547738696,
      "grad_norm": 1.7585675716400146,
      "learning_rate": 8.088333333333334e-06,
      "loss": 0.4668,
      "step": 3150
    },
    {
      "epoch": 2.6800670016750416,
      "grad_norm": 2.0797314643859863,
      "learning_rate": 8.006666666666667e-06,
      "loss": 0.4583,
      "step": 3200
    },
    {
      "epoch": 2.7219430485762146,
      "grad_norm": 5.132874011993408,
      "learning_rate": 7.923333333333334e-06,
      "loss": 0.4573,
      "step": 3250
    },
    {
      "epoch": 2.7638190954773867,
      "grad_norm": 2.005979537963867,
      "learning_rate": 7.840000000000001e-06,
      "loss": 0.449,
      "step": 3300
    },
    {
      "epoch": 2.8056951423785597,
      "grad_norm": 2.802713394165039,
      "learning_rate": 7.756666666666666e-06,
      "loss": 0.448,
      "step": 3350
    },
    {
      "epoch": 2.8475711892797317,
      "grad_norm": 1.8899844884872437,
      "learning_rate": 7.673333333333333e-06,
      "loss": 0.4558,
      "step": 3400
    },
    {
      "epoch": 2.8894472361809047,
      "grad_norm": 3.7795069217681885,
      "learning_rate": 7.590000000000001e-06,
      "loss": 0.4533,
      "step": 3450
    },
    {
      "epoch": 2.931323283082077,
      "grad_norm": 4.659352779388428,
      "learning_rate": 7.506666666666668e-06,
      "loss": 0.4521,
      "step": 3500
    },
    {
      "epoch": 2.931323283082077,
      "eval_loss": 0.400156170129776,
      "eval_runtime": 222.3025,
      "eval_samples_per_second": 38.187,
      "eval_steps_per_second": 4.777,
      "step": 3500
    },
    {
      "epoch": 2.9731993299832498,
      "grad_norm": 2.298712730407715,
      "learning_rate": 7.423333333333333e-06,
      "loss": 0.4398,
      "step": 3550
    },
    {
      "epoch": 3.0150753768844223,
      "grad_norm": 2.995213508605957,
      "learning_rate": 7.340000000000001e-06,
      "loss": 0.4496,
      "step": 3600
    },
    {
      "epoch": 3.056951423785595,
      "grad_norm": 2.17159366607666,
      "learning_rate": 7.256666666666668e-06,
      "loss": 0.4519,
      "step": 3650
    },
    {
      "epoch": 3.0988274706867673,
      "grad_norm": 2.4102556705474854,
      "learning_rate": 7.173333333333335e-06,
      "loss": 0.4497,
      "step": 3700
    },
    {
      "epoch": 3.14070351758794,
      "grad_norm": 1.7901294231414795,
      "learning_rate": 7.09e-06,
      "loss": 0.4545,
      "step": 3750
    },
    {
      "epoch": 3.1825795644891124,
      "grad_norm": 2.869340181350708,
      "learning_rate": 7.006666666666667e-06,
      "loss": 0.4492,
      "step": 3800
    },
    {
      "epoch": 3.224455611390285,
      "grad_norm": 2.1695075035095215,
      "learning_rate": 6.9233333333333345e-06,
      "loss": 0.4505,
      "step": 3850
    },
    {
      "epoch": 3.2663316582914574,
      "grad_norm": 2.9931163787841797,
      "learning_rate": 6.8400000000000014e-06,
      "loss": 0.4421,
      "step": 3900
    },
    {
      "epoch": 3.30820770519263,
      "grad_norm": 3.8682026863098145,
      "learning_rate": 6.756666666666667e-06,
      "loss": 0.4439,
      "step": 3950
    },
    {
      "epoch": 3.3500837520938025,
      "grad_norm": 1.6634626388549805,
      "learning_rate": 6.6733333333333335e-06,
      "loss": 0.4431,
      "step": 4000
    },
    {
      "epoch": 3.3500837520938025,
      "eval_loss": 0.39482244849205017,
      "eval_runtime": 222.4171,
      "eval_samples_per_second": 38.167,
      "eval_steps_per_second": 4.775,
      "step": 4000
    },
    {
      "epoch": 3.391959798994975,
      "grad_norm": 3.006056308746338,
      "learning_rate": 6.5900000000000004e-06,
      "loss": 0.4615,
      "step": 4050
    },
    {
      "epoch": 3.4338358458961475,
      "grad_norm": 1.9082704782485962,
      "learning_rate": 6.5066666666666665e-06,
      "loss": 0.4486,
      "step": 4100
    },
    {
      "epoch": 3.47571189279732,
      "grad_norm": 3.1924006938934326,
      "learning_rate": 6.423333333333333e-06,
      "loss": 0.4526,
      "step": 4150
    },
    {
      "epoch": 3.5175879396984926,
      "grad_norm": 1.9817140102386475,
      "learning_rate": 6.34e-06,
      "loss": 0.4428,
      "step": 4200
    },
    {
      "epoch": 3.559463986599665,
      "grad_norm": 1.3777679204940796,
      "learning_rate": 6.256666666666667e-06,
      "loss": 0.443,
      "step": 4250
    },
    {
      "epoch": 3.6013400335008376,
      "grad_norm": 1.8882936239242554,
      "learning_rate": 6.173333333333333e-06,
      "loss": 0.4404,
      "step": 4300
    },
    {
      "epoch": 3.64321608040201,
      "grad_norm": 1.632260799407959,
      "learning_rate": 6.09e-06,
      "loss": 0.4451,
      "step": 4350
    },
    {
      "epoch": 3.6850921273031827,
      "grad_norm": 7.229835033416748,
      "learning_rate": 6.006666666666667e-06,
      "loss": 0.4352,
      "step": 4400
    },
    {
      "epoch": 3.726968174204355,
      "grad_norm": 1.7340707778930664,
      "learning_rate": 5.923333333333334e-06,
      "loss": 0.433,
      "step": 4450
    },
    {
      "epoch": 3.7688442211055277,
      "grad_norm": 1.81380295753479,
      "learning_rate": 5.84e-06,
      "loss": 0.4393,
      "step": 4500
    },
    {
      "epoch": 3.7688442211055277,
      "eval_loss": 0.3913716673851013,
      "eval_runtime": 221.2245,
      "eval_samples_per_second": 38.373,
      "eval_steps_per_second": 4.801,
      "step": 4500
    },
    {
      "epoch": 3.8107202680067003,
      "grad_norm": 2.9950456619262695,
      "learning_rate": 5.756666666666667e-06,
      "loss": 0.4303,
      "step": 4550
    },
    {
      "epoch": 3.852596314907873,
      "grad_norm": 1.537866473197937,
      "learning_rate": 5.673333333333334e-06,
      "loss": 0.4322,
      "step": 4600
    },
    {
      "epoch": 3.8944723618090453,
      "grad_norm": 2.1190896034240723,
      "learning_rate": 5.590000000000001e-06,
      "loss": 0.4299,
      "step": 4650
    },
    {
      "epoch": 3.936348408710218,
      "grad_norm": 1.5541598796844482,
      "learning_rate": 5.506666666666667e-06,
      "loss": 0.4259,
      "step": 4700
    },
    {
      "epoch": 3.9782244556113904,
      "grad_norm": 2.0810546875,
      "learning_rate": 5.423333333333334e-06,
      "loss": 0.4265,
      "step": 4750
    },
    {
      "epoch": 4.0201005025125625,
      "grad_norm": 1.5855557918548584,
      "learning_rate": 5.3400000000000005e-06,
      "loss": 0.4343,
      "step": 4800
    },
    {
      "epoch": 4.061976549413735,
      "grad_norm": 2.2456538677215576,
      "learning_rate": 5.256666666666667e-06,
      "loss": 0.432,
      "step": 4850
    },
    {
      "epoch": 4.1038525963149075,
      "grad_norm": 2.1106362342834473,
      "learning_rate": 5.1733333333333335e-06,
      "loss": 0.4212,
      "step": 4900
    },
    {
      "epoch": 4.1457286432160805,
      "grad_norm": 2.1016457080841064,
      "learning_rate": 5.09e-06,
      "loss": 0.4343,
      "step": 4950
    },
    {
      "epoch": 4.187604690117253,
      "grad_norm": 2.1289258003234863,
      "learning_rate": 5.006666666666667e-06,
      "loss": 0.4271,
      "step": 5000
    },
    {
      "epoch": 4.187604690117253,
      "eval_loss": 0.3861308693885803,
      "eval_runtime": 221.7311,
      "eval_samples_per_second": 38.285,
      "eval_steps_per_second": 4.79,
      "step": 5000
    }
  ],
  "logging_steps": 50,
  "max_steps": 8000,
  "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": 4.851101140979386e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}