{
  "best_metric": 0.3649367392063141,
  "best_model_checkpoint": "mikhail-panzo/zlm_b64_le5_s12000/checkpoint-9500",
  "epoch": 7.956448911222781,
  "eval_steps": 500,
  "global_step": 9500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04187604690117253,
      "grad_norm": 4.788218975067139,
      "learning_rate": 2.4500000000000004e-07,
      "loss": 1.1819,
      "step": 50
    },
    {
      "epoch": 0.08375209380234507,
      "grad_norm": 15.618935585021973,
      "learning_rate": 4.95e-07,
      "loss": 1.0042,
      "step": 100
    },
    {
      "epoch": 0.12562814070351758,
      "grad_norm": 2.7821555137634277,
      "learning_rate": 7.4e-07,
      "loss": 0.9676,
      "step": 150
    },
    {
      "epoch": 0.16750418760469013,
      "grad_norm": 3.759263038635254,
      "learning_rate": 9.85e-07,
      "loss": 0.9362,
      "step": 200
    },
    {
      "epoch": 0.20938023450586266,
      "grad_norm": 3.906439781188965,
      "learning_rate": 1.235e-06,
      "loss": 0.8667,
      "step": 250
    },
    {
      "epoch": 0.25125628140703515,
      "grad_norm": 4.397643566131592,
      "learning_rate": 1.485e-06,
      "loss": 0.8455,
      "step": 300
    },
    {
      "epoch": 0.2931323283082077,
      "grad_norm": 2.94077467918396,
      "learning_rate": 1.7350000000000001e-06,
      "loss": 0.8376,
      "step": 350
    },
    {
      "epoch": 0.33500837520938026,
      "grad_norm": 4.161074161529541,
      "learning_rate": 1.985e-06,
      "loss": 0.7521,
      "step": 400
    },
    {
      "epoch": 0.3768844221105528,
      "grad_norm": 2.8088552951812744,
      "learning_rate": 2.235e-06,
      "loss": 0.7444,
      "step": 450
    },
    {
      "epoch": 0.4187604690117253,
      "grad_norm": 2.62283992767334,
      "learning_rate": 2.4850000000000003e-06,
      "loss": 0.7129,
      "step": 500
    },
    {
      "epoch": 0.4187604690117253,
      "eval_loss": 0.6066474914550781,
      "eval_runtime": 213.4911,
      "eval_samples_per_second": 39.763,
      "eval_steps_per_second": 4.974,
      "step": 500
    },
    {
      "epoch": 0.46063651591289784,
      "grad_norm": 3.654665946960449,
      "learning_rate": 2.7350000000000005e-06,
      "loss": 0.6772,
      "step": 550
    },
    {
      "epoch": 0.5025125628140703,
      "grad_norm": 5.1851887702941895,
      "learning_rate": 2.9850000000000002e-06,
      "loss": 0.6552,
      "step": 600
    },
    {
      "epoch": 0.5443886097152428,
      "grad_norm": 4.017187595367432,
      "learning_rate": 3.2350000000000004e-06,
      "loss": 0.6506,
      "step": 650
    },
    {
      "epoch": 0.5862646566164154,
      "grad_norm": 2.1161563396453857,
      "learning_rate": 3.485e-06,
      "loss": 0.6301,
      "step": 700
    },
    {
      "epoch": 0.628140703517588,
      "grad_norm": 1.857313871383667,
      "learning_rate": 3.7350000000000002e-06,
      "loss": 0.6373,
      "step": 750
    },
    {
      "epoch": 0.6700167504187605,
      "grad_norm": 3.6841330528259277,
      "learning_rate": 3.985000000000001e-06,
      "loss": 0.6305,
      "step": 800
    },
    {
      "epoch": 0.711892797319933,
      "grad_norm": 2.6450912952423096,
      "learning_rate": 4.235e-06,
      "loss": 0.6169,
      "step": 850
    },
    {
      "epoch": 0.7537688442211056,
      "grad_norm": 1.7241184711456299,
      "learning_rate": 4.485e-06,
      "loss": 0.5933,
      "step": 900
    },
    {
      "epoch": 0.7956448911222781,
      "grad_norm": 2.602735996246338,
      "learning_rate": 4.735e-06,
      "loss": 0.583,
      "step": 950
    },
    {
      "epoch": 0.8375209380234506,
      "grad_norm": 12.419028282165527,
      "learning_rate": 4.9850000000000006e-06,
      "loss": 0.5922,
      "step": 1000
    },
    {
      "epoch": 0.8375209380234506,
      "eval_loss": 0.4979843497276306,
      "eval_runtime": 214.0502,
      "eval_samples_per_second": 39.659,
      "eval_steps_per_second": 4.961,
      "step": 1000
    },
    {
      "epoch": 0.8793969849246231,
      "grad_norm": 3.200178861618042,
      "learning_rate": 5.235e-06,
      "loss": 0.5678,
      "step": 1050
    },
    {
      "epoch": 0.9212730318257957,
      "grad_norm": 2.458869218826294,
      "learning_rate": 5.485e-06,
      "loss": 0.5558,
      "step": 1100
    },
    {
      "epoch": 0.9631490787269682,
      "grad_norm": 2.341752290725708,
      "learning_rate": 5.735e-06,
      "loss": 0.5666,
      "step": 1150
    },
    {
      "epoch": 1.0050251256281406,
      "grad_norm": 2.9191603660583496,
      "learning_rate": 5.985000000000001e-06,
      "loss": 0.5345,
      "step": 1200
    },
    {
      "epoch": 1.0469011725293131,
      "grad_norm": 2.68806529045105,
      "learning_rate": 6.235000000000001e-06,
      "loss": 0.5653,
      "step": 1250
    },
    {
      "epoch": 1.0887772194304857,
      "grad_norm": 3.278820276260376,
      "learning_rate": 6.485000000000001e-06,
      "loss": 0.5565,
      "step": 1300
    },
    {
      "epoch": 1.1306532663316582,
      "grad_norm": 3.066511869430542,
      "learning_rate": 6.735000000000001e-06,
      "loss": 0.5644,
      "step": 1350
    },
    {
      "epoch": 1.1725293132328307,
      "grad_norm": 4.269862651824951,
      "learning_rate": 6.985000000000001e-06,
      "loss": 0.5356,
      "step": 1400
    },
    {
      "epoch": 1.2144053601340032,
      "grad_norm": 3.6681156158447266,
      "learning_rate": 7.235000000000001e-06,
      "loss": 0.5246,
      "step": 1450
    },
    {
      "epoch": 1.2562814070351758,
      "grad_norm": 3.1227505207061768,
      "learning_rate": 7.485000000000001e-06,
      "loss": 0.5168,
      "step": 1500
    },
    {
      "epoch": 1.2562814070351758,
      "eval_loss": 0.45785218477249146,
      "eval_runtime": 211.6749,
      "eval_samples_per_second": 40.104,
      "eval_steps_per_second": 5.017,
      "step": 1500
    },
    {
      "epoch": 1.2981574539363483,
      "grad_norm": 2.6239662170410156,
      "learning_rate": 7.735e-06,
      "loss": 0.5153,
      "step": 1550
    },
    {
      "epoch": 1.3400335008375208,
      "grad_norm": 2.881901979446411,
      "learning_rate": 7.985e-06,
      "loss": 0.518,
      "step": 1600
    },
    {
      "epoch": 1.3819095477386933,
      "grad_norm": 2.8052310943603516,
      "learning_rate": 8.235e-06,
      "loss": 0.5272,
      "step": 1650
    },
    {
      "epoch": 1.4237855946398659,
      "grad_norm": 2.6334052085876465,
      "learning_rate": 8.485000000000001e-06,
      "loss": 0.5188,
      "step": 1700
    },
    {
      "epoch": 1.4656616415410384,
      "grad_norm": 3.038274049758911,
      "learning_rate": 8.735000000000002e-06,
      "loss": 0.5139,
      "step": 1750
    },
    {
      "epoch": 1.507537688442211,
      "grad_norm": 3.2438852787017822,
      "learning_rate": 8.985000000000001e-06,
      "loss": 0.5092,
      "step": 1800
    },
    {
      "epoch": 1.5494137353433834,
      "grad_norm": 4.168883800506592,
      "learning_rate": 9.235e-06,
      "loss": 0.5103,
      "step": 1850
    },
    {
      "epoch": 1.591289782244556,
      "grad_norm": 2.6399967670440674,
      "learning_rate": 9.485000000000002e-06,
      "loss": 0.5025,
      "step": 1900
    },
    {
      "epoch": 1.6331658291457285,
      "grad_norm": 1.6828950643539429,
      "learning_rate": 9.735e-06,
      "loss": 0.4963,
      "step": 1950
    },
    {
      "epoch": 1.675041876046901,
      "grad_norm": 1.8955535888671875,
      "learning_rate": 9.985000000000002e-06,
      "loss": 0.4954,
      "step": 2000
    },
    {
      "epoch": 1.675041876046901,
      "eval_loss": 0.43827974796295166,
      "eval_runtime": 212.1455,
      "eval_samples_per_second": 40.015,
      "eval_steps_per_second": 5.006,
      "step": 2000
    },
    {
      "epoch": 1.7169179229480735,
      "grad_norm": 3.9634511470794678,
      "learning_rate": 9.953000000000001e-06,
      "loss": 0.4891,
      "step": 2050
    },
    {
      "epoch": 1.758793969849246,
      "grad_norm": 3.5446126461029053,
      "learning_rate": 9.903e-06,
      "loss": 0.5141,
      "step": 2100
    },
    {
      "epoch": 1.8006700167504186,
      "grad_norm": 3.1621015071868896,
      "learning_rate": 9.853e-06,
      "loss": 0.4941,
      "step": 2150
    },
    {
      "epoch": 1.8425460636515911,
      "grad_norm": 2.6622440814971924,
      "learning_rate": 9.803e-06,
      "loss": 0.4979,
      "step": 2200
    },
    {
      "epoch": 1.8844221105527639,
      "grad_norm": 1.7774969339370728,
      "learning_rate": 9.753e-06,
      "loss": 0.4768,
      "step": 2250
    },
    {
      "epoch": 1.9262981574539364,
      "grad_norm": 2.1141610145568848,
      "learning_rate": 9.703000000000002e-06,
      "loss": 0.4834,
      "step": 2300
    },
    {
      "epoch": 1.968174204355109,
      "grad_norm": 3.4576170444488525,
      "learning_rate": 9.653e-06,
      "loss": 0.493,
      "step": 2350
    },
    {
      "epoch": 2.0100502512562812,
      "grad_norm": 2.4238383769989014,
      "learning_rate": 9.603000000000001e-06,
      "loss": 0.4836,
      "step": 2400
    },
    {
      "epoch": 2.0519262981574538,
      "grad_norm": 2.2526774406433105,
      "learning_rate": 9.553000000000002e-06,
      "loss": 0.4832,
      "step": 2450
    },
    {
      "epoch": 2.0938023450586263,
      "grad_norm": 2.8719518184661865,
      "learning_rate": 9.503e-06,
      "loss": 0.4836,
      "step": 2500
    },
    {
      "epoch": 2.0938023450586263,
      "eval_loss": 0.4208170473575592,
      "eval_runtime": 214.437,
      "eval_samples_per_second": 39.587,
      "eval_steps_per_second": 4.953,
      "step": 2500
    },
    {
      "epoch": 2.135678391959799,
      "grad_norm": 3.2671754360198975,
      "learning_rate": 9.453e-06,
      "loss": 0.4775,
      "step": 2550
    },
    {
      "epoch": 2.1775544388609713,
      "grad_norm": 2.454883337020874,
      "learning_rate": 9.403000000000001e-06,
      "loss": 0.4777,
      "step": 2600
    },
    {
      "epoch": 2.219430485762144,
      "grad_norm": 2.901097297668457,
      "learning_rate": 9.353000000000002e-06,
      "loss": 0.4757,
      "step": 2650
    },
    {
      "epoch": 2.2613065326633164,
      "grad_norm": 2.980057954788208,
      "learning_rate": 9.303e-06,
      "loss": 0.4644,
      "step": 2700
    },
    {
      "epoch": 2.3031825795644894,
      "grad_norm": 1.7162193059921265,
      "learning_rate": 9.253000000000001e-06,
      "loss": 0.4714,
      "step": 2750
    },
    {
      "epoch": 2.3450586264656614,
      "grad_norm": 2.912874937057495,
      "learning_rate": 9.203000000000002e-06,
      "loss": 0.4596,
      "step": 2800
    },
    {
      "epoch": 2.3869346733668344,
      "grad_norm": 1.7781060934066772,
      "learning_rate": 9.153e-06,
      "loss": 0.4591,
      "step": 2850
    },
    {
      "epoch": 2.4288107202680065,
      "grad_norm": 5.2557902336120605,
      "learning_rate": 9.103e-06,
      "loss": 0.4652,
      "step": 2900
    },
    {
      "epoch": 2.4706867671691795,
      "grad_norm": 1.7230316400527954,
      "learning_rate": 9.053000000000001e-06,
      "loss": 0.46,
      "step": 2950
    },
    {
      "epoch": 2.5125628140703515,
      "grad_norm": 4.071449279785156,
      "learning_rate": 9.003e-06,
      "loss": 0.4623,
      "step": 3000
    },
    {
      "epoch": 2.5125628140703515,
      "eval_loss": 0.4098353981971741,
      "eval_runtime": 214.5603,
      "eval_samples_per_second": 39.565,
      "eval_steps_per_second": 4.95,
      "step": 3000
    },
    {
      "epoch": 2.5544388609715245,
      "grad_norm": 3.5458014011383057,
      "learning_rate": 8.953e-06,
      "loss": 0.4488,
      "step": 3050
    },
    {
      "epoch": 2.5963149078726966,
      "grad_norm": 1.726624846458435,
      "learning_rate": 8.903000000000001e-06,
      "loss": 0.4536,
      "step": 3100
    },
    {
      "epoch": 2.6381909547738696,
      "grad_norm": 3.204106092453003,
      "learning_rate": 8.853e-06,
      "loss": 0.4654,
      "step": 3150
    },
    {
      "epoch": 2.6800670016750416,
      "grad_norm": 2.245302677154541,
      "learning_rate": 8.803e-06,
      "loss": 0.462,
      "step": 3200
    },
    {
      "epoch": 2.7219430485762146,
      "grad_norm": 2.0797626972198486,
      "learning_rate": 8.753e-06,
      "loss": 0.4599,
      "step": 3250
    },
    {
      "epoch": 2.7638190954773867,
      "grad_norm": 1.9437323808670044,
      "learning_rate": 8.703e-06,
      "loss": 0.4473,
      "step": 3300
    },
    {
      "epoch": 2.8056951423785597,
      "grad_norm": 2.5299806594848633,
      "learning_rate": 8.653e-06,
      "loss": 0.4442,
      "step": 3350
    },
    {
      "epoch": 2.8475711892797317,
      "grad_norm": 3.208460807800293,
      "learning_rate": 8.603e-06,
      "loss": 0.4553,
      "step": 3400
    },
    {
      "epoch": 2.8894472361809047,
      "grad_norm": 2.703047037124634,
      "learning_rate": 8.553000000000001e-06,
      "loss": 0.4532,
      "step": 3450
    },
    {
      "epoch": 2.931323283082077,
      "grad_norm": 1.9532470703125,
      "learning_rate": 8.503e-06,
      "loss": 0.4499,
      "step": 3500
    },
    {
      "epoch": 2.931323283082077,
      "eval_loss": 0.4003259837627411,
      "eval_runtime": 213.8518,
      "eval_samples_per_second": 39.696,
      "eval_steps_per_second": 4.966,
      "step": 3500
    },
    {
      "epoch": 2.9731993299832498,
      "grad_norm": 1.9028701782226562,
      "learning_rate": 8.453000000000002e-06,
      "loss": 0.4391,
      "step": 3550
    },
    {
      "epoch": 3.0150753768844223,
      "grad_norm": 10.080265045166016,
      "learning_rate": 8.403e-06,
      "loss": 0.4462,
      "step": 3600
    },
    {
      "epoch": 3.056951423785595,
      "grad_norm": 2.996328115463257,
      "learning_rate": 8.353000000000001e-06,
      "loss": 0.45,
      "step": 3650
    },
    {
      "epoch": 3.0988274706867673,
      "grad_norm": 2.221226930618286,
      "learning_rate": 8.303000000000002e-06,
      "loss": 0.4469,
      "step": 3700
    },
    {
      "epoch": 3.14070351758794,
      "grad_norm": 2.961524248123169,
      "learning_rate": 8.253e-06,
      "loss": 0.4535,
      "step": 3750
    },
    {
      "epoch": 3.1825795644891124,
      "grad_norm": 2.487060785293579,
      "learning_rate": 8.203000000000001e-06,
      "loss": 0.4444,
      "step": 3800
    },
    {
      "epoch": 3.224455611390285,
      "grad_norm": 1.5449366569519043,
      "learning_rate": 8.153000000000001e-06,
      "loss": 0.449,
      "step": 3850
    },
    {
      "epoch": 3.2663316582914574,
      "grad_norm": 1.6193325519561768,
      "learning_rate": 8.103e-06,
      "loss": 0.4423,
      "step": 3900
    },
    {
      "epoch": 3.30820770519263,
      "grad_norm": 2.442594528198242,
      "learning_rate": 8.053e-06,
      "loss": 0.4456,
      "step": 3950
    },
    {
      "epoch": 3.3500837520938025,
      "grad_norm": 1.759829044342041,
      "learning_rate": 8.003000000000001e-06,
      "loss": 0.4421,
      "step": 4000
    },
    {
      "epoch": 3.3500837520938025,
      "eval_loss": 0.39415255188941956,
      "eval_runtime": 212.9896,
      "eval_samples_per_second": 39.856,
      "eval_steps_per_second": 4.986,
      "step": 4000
    },
    {
      "epoch": 3.391959798994975,
      "grad_norm": 1.6665236949920654,
      "learning_rate": 7.953e-06,
      "loss": 0.4626,
      "step": 4050
    },
    {
      "epoch": 3.4338358458961475,
      "grad_norm": 2.2765064239501953,
      "learning_rate": 7.903e-06,
      "loss": 0.4467,
      "step": 4100
    },
    {
      "epoch": 3.47571189279732,
      "grad_norm": 2.401715040206909,
      "learning_rate": 7.853000000000001e-06,
      "loss": 0.4523,
      "step": 4150
    },
    {
      "epoch": 3.5175879396984926,
      "grad_norm": 4.074472427368164,
      "learning_rate": 7.804e-06,
      "loss": 0.4422,
      "step": 4200
    },
    {
      "epoch": 3.559463986599665,
      "grad_norm": 2.0492303371429443,
      "learning_rate": 7.754e-06,
      "loss": 0.4417,
      "step": 4250
    },
    {
      "epoch": 3.6013400335008376,
      "grad_norm": 2.8641483783721924,
      "learning_rate": 7.704000000000001e-06,
      "loss": 0.4364,
      "step": 4300
    },
    {
      "epoch": 3.64321608040201,
      "grad_norm": 2.1105141639709473,
      "learning_rate": 7.654e-06,
      "loss": 0.4456,
      "step": 4350
    },
    {
      "epoch": 3.6850921273031827,
      "grad_norm": 3.2357890605926514,
      "learning_rate": 7.604e-06,
      "loss": 0.4314,
      "step": 4400
    },
    {
      "epoch": 3.726968174204355,
      "grad_norm": 1.3265612125396729,
      "learning_rate": 7.554e-06,
      "loss": 0.4309,
      "step": 4450
    },
    {
      "epoch": 3.7688442211055277,
      "grad_norm": 1.558438777923584,
      "learning_rate": 7.5040000000000005e-06,
      "loss": 0.4345,
      "step": 4500
    },
    {
      "epoch": 3.7688442211055277,
      "eval_loss": 0.3893836736679077,
      "eval_runtime": 214.7943,
      "eval_samples_per_second": 39.522,
      "eval_steps_per_second": 4.944,
      "step": 4500
    },
    {
      "epoch": 3.8107202680067003,
      "grad_norm": 1.6830787658691406,
      "learning_rate": 7.454e-06,
      "loss": 0.4279,
      "step": 4550
    },
    {
      "epoch": 3.852596314907873,
      "grad_norm": 2.7424182891845703,
      "learning_rate": 7.404e-06,
      "loss": 0.4312,
      "step": 4600
    },
    {
      "epoch": 3.8944723618090453,
      "grad_norm": 1.8119078874588013,
      "learning_rate": 7.354000000000001e-06,
      "loss": 0.4262,
      "step": 4650
    },
    {
      "epoch": 3.936348408710218,
      "grad_norm": 1.8333683013916016,
      "learning_rate": 7.304000000000001e-06,
      "loss": 0.4225,
      "step": 4700
    },
    {
      "epoch": 3.9782244556113904,
      "grad_norm": 1.8040356636047363,
      "learning_rate": 7.254000000000001e-06,
      "loss": 0.4239,
      "step": 4750
    },
    {
      "epoch": 4.0201005025125625,
      "grad_norm": 1.7291460037231445,
      "learning_rate": 7.204000000000001e-06,
      "loss": 0.4327,
      "step": 4800
    },
    {
      "epoch": 4.061976549413735,
      "grad_norm": 1.8311458826065063,
      "learning_rate": 7.154000000000001e-06,
      "loss": 0.4297,
      "step": 4850
    },
    {
      "epoch": 4.1038525963149075,
      "grad_norm": 2.10210919380188,
      "learning_rate": 7.104000000000001e-06,
      "loss": 0.4207,
      "step": 4900
    },
    {
      "epoch": 4.1457286432160805,
      "grad_norm": 2.5385794639587402,
      "learning_rate": 7.0540000000000006e-06,
      "loss": 0.4342,
      "step": 4950
    },
    {
      "epoch": 4.187604690117253,
      "grad_norm": 1.8550605773925781,
      "learning_rate": 7.004000000000001e-06,
      "loss": 0.4233,
      "step": 5000
    },
    {
      "epoch": 4.187604690117253,
      "eval_loss": 0.384044349193573,
      "eval_runtime": 215.1398,
      "eval_samples_per_second": 39.458,
      "eval_steps_per_second": 4.936,
      "step": 5000
    },
    {
      "epoch": 4.2294807370184255,
      "grad_norm": 1.6772557497024536,
      "learning_rate": 6.954000000000001e-06,
      "loss": 0.4258,
      "step": 5050
    },
    {
      "epoch": 4.271356783919598,
      "grad_norm": 1.554694652557373,
      "learning_rate": 6.904e-06,
      "loss": 0.4277,
      "step": 5100
    },
    {
      "epoch": 4.313232830820771,
      "grad_norm": 1.6072709560394287,
      "learning_rate": 6.854000000000001e-06,
      "loss": 0.4271,
      "step": 5150
    },
    {
      "epoch": 4.355108877721943,
      "grad_norm": 2.7782182693481445,
      "learning_rate": 6.804e-06,
      "loss": 0.4206,
      "step": 5200
    },
    {
      "epoch": 4.396984924623116,
      "grad_norm": 2.203001022338867,
      "learning_rate": 6.754000000000001e-06,
      "loss": 0.4237,
      "step": 5250
    },
    {
      "epoch": 4.438860971524288,
      "grad_norm": 1.6368170976638794,
      "learning_rate": 6.7040000000000005e-06,
      "loss": 0.4221,
      "step": 5300
    },
    {
      "epoch": 4.480737018425461,
      "grad_norm": 1.8552517890930176,
      "learning_rate": 6.654e-06,
      "loss": 0.4219,
      "step": 5350
    },
    {
      "epoch": 4.522613065326633,
      "grad_norm": 2.663379192352295,
      "learning_rate": 6.604000000000001e-06,
      "loss": 0.4335,
      "step": 5400
    },
    {
      "epoch": 4.564489112227806,
      "grad_norm": 1.8602125644683838,
      "learning_rate": 6.554e-06,
      "loss": 0.4194,
      "step": 5450
    },
    {
      "epoch": 4.606365159128979,
      "grad_norm": 1.888647437095642,
      "learning_rate": 6.504e-06,
      "loss": 0.4288,
      "step": 5500
    },
    {
      "epoch": 4.606365159128979,
      "eval_loss": 0.3807646930217743,
      "eval_runtime": 214.2921,
      "eval_samples_per_second": 39.614,
      "eval_steps_per_second": 4.956,
      "step": 5500
    },
    {
      "epoch": 4.648241206030151,
      "grad_norm": 1.3516755104064941,
      "learning_rate": 6.454e-06,
      "loss": 0.4183,
      "step": 5550
    },
    {
      "epoch": 4.690117252931323,
      "grad_norm": 1.143816590309143,
      "learning_rate": 6.404e-06,
      "loss": 0.4227,
      "step": 5600
    },
    {
      "epoch": 4.731993299832496,
      "grad_norm": 2.002258062362671,
      "learning_rate": 6.3540000000000005e-06,
      "loss": 0.4211,
      "step": 5650
    },
    {
      "epoch": 4.773869346733669,
      "grad_norm": 2.188854217529297,
      "learning_rate": 6.304e-06,
      "loss": 0.4186,
      "step": 5700
    },
    {
      "epoch": 4.815745393634841,
      "grad_norm": 1.709869384765625,
      "learning_rate": 6.254e-06,
      "loss": 0.4244,
      "step": 5750
    },
    {
      "epoch": 4.857621440536013,
      "grad_norm": 1.9853215217590332,
      "learning_rate": 6.204e-06,
      "loss": 0.4171,
      "step": 5800
    },
    {
      "epoch": 4.899497487437186,
      "grad_norm": 1.9427616596221924,
      "learning_rate": 6.154e-06,
      "loss": 0.4178,
      "step": 5850
    },
    {
      "epoch": 4.941373534338359,
      "grad_norm": 1.9131799936294556,
      "learning_rate": 6.104000000000001e-06,
      "loss": 0.4205,
      "step": 5900
    },
    {
      "epoch": 4.983249581239531,
      "grad_norm": 2.243724822998047,
      "learning_rate": 6.054000000000001e-06,
      "loss": 0.4144,
      "step": 5950
    },
    {
      "epoch": 5.025125628140704,
      "grad_norm": 1.4981979131698608,
      "learning_rate": 6.004000000000001e-06,
      "loss": 0.4218,
      "step": 6000
    },
    {
      "epoch": 5.025125628140704,
      "eval_loss": 0.3775295317173004,
      "eval_runtime": 216.3465,
      "eval_samples_per_second": 39.238,
      "eval_steps_per_second": 4.909,
      "step": 6000
    },
    {
      "epoch": 5.067001675041876,
      "grad_norm": 2.0180861949920654,
      "learning_rate": 5.954000000000001e-06,
      "loss": 0.4109,
      "step": 6050
    },
    {
      "epoch": 5.108877721943049,
      "grad_norm": 1.9829692840576172,
      "learning_rate": 5.9040000000000006e-06,
      "loss": 0.4174,
      "step": 6100
    },
    {
      "epoch": 5.150753768844221,
      "grad_norm": 2.0144565105438232,
      "learning_rate": 5.854000000000001e-06,
      "loss": 0.4158,
      "step": 6150
    },
    {
      "epoch": 5.192629815745394,
      "grad_norm": 1.966562032699585,
      "learning_rate": 5.804000000000001e-06,
      "loss": 0.4158,
      "step": 6200
    },
    {
      "epoch": 5.234505862646566,
      "grad_norm": 2.074537992477417,
      "learning_rate": 5.754e-06,
      "loss": 0.414,
      "step": 6250
    },
    {
      "epoch": 5.276381909547739,
      "grad_norm": 2.5158934593200684,
      "learning_rate": 5.704000000000001e-06,
      "loss": 0.4132,
      "step": 6300
    },
    {
      "epoch": 5.318257956448911,
      "grad_norm": 1.9309169054031372,
      "learning_rate": 5.654e-06,
      "loss": 0.4154,
      "step": 6350
    },
    {
      "epoch": 5.360134003350084,
      "grad_norm": 1.3020102977752686,
      "learning_rate": 5.604000000000001e-06,
      "loss": 0.423,
      "step": 6400
    },
    {
      "epoch": 5.402010050251256,
      "grad_norm": 1.4237923622131348,
      "learning_rate": 5.5540000000000005e-06,
      "loss": 0.413,
      "step": 6450
    },
    {
      "epoch": 5.443886097152429,
      "grad_norm": 1.6322195529937744,
      "learning_rate": 5.504e-06,
      "loss": 0.4174,
      "step": 6500
    },
    {
      "epoch": 5.443886097152429,
      "eval_loss": 0.3745593726634979,
      "eval_runtime": 216.3114,
      "eval_samples_per_second": 39.244,
      "eval_steps_per_second": 4.91,
      "step": 6500
    },
    {
      "epoch": 5.485762144053601,
      "grad_norm": 2.4924020767211914,
      "learning_rate": 5.454000000000001e-06,
      "loss": 0.4081,
      "step": 6550
    },
    {
      "epoch": 5.527638190954773,
      "grad_norm": 1.912190318107605,
      "learning_rate": 5.404e-06,
      "loss": 0.4138,
      "step": 6600
    },
    {
      "epoch": 5.569514237855946,
      "grad_norm": 2.1161069869995117,
      "learning_rate": 5.354e-06,
      "loss": 0.4109,
      "step": 6650
    },
    {
      "epoch": 5.611390284757119,
      "grad_norm": 1.890815258026123,
      "learning_rate": 5.304e-06,
      "loss": 0.4103,
      "step": 6700
    },
    {
      "epoch": 5.653266331658291,
      "grad_norm": 1.7124552726745605,
      "learning_rate": 5.254e-06,
      "loss": 0.414,
      "step": 6750
    },
    {
      "epoch": 5.695142378559464,
      "grad_norm": 1.9143486022949219,
      "learning_rate": 5.2040000000000005e-06,
      "loss": 0.4153,
      "step": 6800
    },
    {
      "epoch": 5.7370184254606365,
      "grad_norm": 2.9802162647247314,
      "learning_rate": 5.154e-06,
      "loss": 0.4137,
      "step": 6850
    },
    {
      "epoch": 5.778894472361809,
      "grad_norm": 1.3769222497940063,
      "learning_rate": 5.104e-06,
      "loss": 0.407,
      "step": 6900
    },
    {
      "epoch": 5.8207705192629815,
      "grad_norm": 1.4017168283462524,
      "learning_rate": 5.054e-06,
      "loss": 0.4121,
      "step": 6950
    },
    {
      "epoch": 5.8626465661641545,
      "grad_norm": 1.4576807022094727,
      "learning_rate": 5.004e-06,
      "loss": 0.4075,
      "step": 7000
    },
    {
      "epoch": 5.8626465661641545,
      "eval_loss": 0.37438860535621643,
      "eval_runtime": 215.1568,
      "eval_samples_per_second": 39.455,
      "eval_steps_per_second": 4.936,
      "step": 7000
    },
    {
      "epoch": 5.9045226130653266,
      "grad_norm": 1.978575587272644,
      "learning_rate": 4.954e-06,
      "loss": 0.4211,
      "step": 7050
    },
    {
      "epoch": 5.9463986599664995,
      "grad_norm": 1.596773386001587,
      "learning_rate": 4.904000000000001e-06,
      "loss": 0.4092,
      "step": 7100
    },
    {
      "epoch": 5.988274706867672,
      "grad_norm": 1.6083742380142212,
      "learning_rate": 4.8540000000000005e-06,
      "loss": 0.4158,
      "step": 7150
    },
    {
      "epoch": 6.030150753768845,
      "grad_norm": 1.3935966491699219,
      "learning_rate": 4.804e-06,
      "loss": 0.4065,
      "step": 7200
    },
    {
      "epoch": 6.072026800670017,
      "grad_norm": 1.670521855354309,
      "learning_rate": 4.7540000000000006e-06,
      "loss": 0.4117,
      "step": 7250
    },
    {
      "epoch": 6.11390284757119,
      "grad_norm": 1.903156042098999,
      "learning_rate": 4.704e-06,
      "loss": 0.407,
      "step": 7300
    },
    {
      "epoch": 6.155778894472362,
      "grad_norm": 1.7410374879837036,
      "learning_rate": 4.654e-06,
      "loss": 0.4148,
      "step": 7350
    },
    {
      "epoch": 6.197654941373535,
      "grad_norm": 2.5548722743988037,
      "learning_rate": 4.604e-06,
      "loss": 0.4068,
      "step": 7400
    },
    {
      "epoch": 6.239530988274707,
      "grad_norm": 2.150181293487549,
      "learning_rate": 4.554000000000001e-06,
      "loss": 0.4006,
      "step": 7450
    },
    {
      "epoch": 6.28140703517588,
      "grad_norm": 1.514251470565796,
      "learning_rate": 4.504e-06,
      "loss": 0.4042,
      "step": 7500
    },
    {
      "epoch": 6.28140703517588,
      "eval_loss": 0.37153398990631104,
      "eval_runtime": 214.3016,
      "eval_samples_per_second": 39.612,
      "eval_steps_per_second": 4.956,
      "step": 7500
    },
    {
      "epoch": 6.323283082077052,
      "grad_norm": 1.6764007806777954,
      "learning_rate": 4.454000000000001e-06,
      "loss": 0.4132,
      "step": 7550
    },
    {
      "epoch": 6.365159128978225,
      "grad_norm": 1.9812299013137817,
      "learning_rate": 4.4040000000000005e-06,
      "loss": 0.4051,
      "step": 7600
    },
    {
      "epoch": 6.407035175879397,
      "grad_norm": 1.9731210470199585,
      "learning_rate": 4.354e-06,
      "loss": 0.4065,
      "step": 7650
    },
    {
      "epoch": 6.44891122278057,
      "grad_norm": 1.4154566526412964,
      "learning_rate": 4.304000000000001e-06,
      "loss": 0.4035,
      "step": 7700
    },
    {
      "epoch": 6.490787269681742,
      "grad_norm": 1.3490220308303833,
      "learning_rate": 4.254e-06,
      "loss": 0.4124,
      "step": 7750
    },
    {
      "epoch": 6.532663316582915,
      "grad_norm": 1.3685543537139893,
      "learning_rate": 4.204e-06,
      "loss": 0.4073,
      "step": 7800
    },
    {
      "epoch": 6.574539363484087,
      "grad_norm": 2.5781421661376953,
      "learning_rate": 4.154e-06,
      "loss": 0.4051,
      "step": 7850
    },
    {
      "epoch": 6.61641541038526,
      "grad_norm": 1.4581451416015625,
      "learning_rate": 4.104e-06,
      "loss": 0.4063,
      "step": 7900
    },
    {
      "epoch": 6.658291457286432,
      "grad_norm": 1.944769024848938,
      "learning_rate": 4.0540000000000005e-06,
      "loss": 0.4019,
      "step": 7950
    },
    {
      "epoch": 6.700167504187605,
      "grad_norm": 1.7144460678100586,
      "learning_rate": 4.004e-06,
      "loss": 0.4075,
      "step": 8000
    },
    {
      "epoch": 6.700167504187605,
      "eval_loss": 0.3694958984851837,
      "eval_runtime": 216.5797,
      "eval_samples_per_second": 39.196,
      "eval_steps_per_second": 4.904,
      "step": 8000
    },
    {
      "epoch": 6.742043551088777,
      "grad_norm": 1.6752500534057617,
      "learning_rate": 3.954e-06,
      "loss": 0.4038,
      "step": 8050
    },
    {
      "epoch": 6.78391959798995,
      "grad_norm": 1.8653851747512817,
      "learning_rate": 3.904e-06,
      "loss": 0.4138,
      "step": 8100
    },
    {
      "epoch": 6.825795644891122,
      "grad_norm": 1.4554862976074219,
      "learning_rate": 3.854000000000001e-06,
      "loss": 0.3997,
      "step": 8150
    },
    {
      "epoch": 6.867671691792295,
      "grad_norm": 1.586377501487732,
      "learning_rate": 3.8040000000000003e-06,
      "loss": 0.4114,
      "step": 8200
    },
    {
      "epoch": 6.909547738693467,
      "grad_norm": 1.9160149097442627,
      "learning_rate": 3.7540000000000004e-06,
      "loss": 0.4028,
      "step": 8250
    },
    {
      "epoch": 6.95142378559464,
      "grad_norm": 1.3755500316619873,
      "learning_rate": 3.7040000000000005e-06,
      "loss": 0.411,
      "step": 8300
    },
    {
      "epoch": 6.993299832495812,
      "grad_norm": 1.6453169584274292,
      "learning_rate": 3.655e-06,
      "loss": 0.4067,
      "step": 8350
    },
    {
      "epoch": 7.035175879396985,
      "grad_norm": 1.8807151317596436,
      "learning_rate": 3.6050000000000002e-06,
      "loss": 0.4056,
      "step": 8400
    },
    {
      "epoch": 7.077051926298157,
      "grad_norm": 1.33198881149292,
      "learning_rate": 3.5550000000000003e-06,
      "loss": 0.4105,
      "step": 8450
    },
    {
      "epoch": 7.11892797319933,
      "grad_norm": 1.4984444379806519,
      "learning_rate": 3.505e-06,
      "loss": 0.4064,
      "step": 8500
    },
    {
      "epoch": 7.11892797319933,
      "eval_loss": 0.3666473627090454,
      "eval_runtime": 213.5237,
      "eval_samples_per_second": 39.757,
      "eval_steps_per_second": 4.974,
      "step": 8500
    },
    {
      "epoch": 7.160804020100502,
      "grad_norm": 1.289669394493103,
      "learning_rate": 3.455e-06,
      "loss": 0.3988,
      "step": 8550
    },
    {
      "epoch": 7.202680067001675,
      "grad_norm": 1.299459457397461,
      "learning_rate": 3.4050000000000004e-06,
      "loss": 0.409,
      "step": 8600
    },
    {
      "epoch": 7.244556113902847,
      "grad_norm": 1.4774562120437622,
      "learning_rate": 3.3550000000000005e-06,
      "loss": 0.4054,
      "step": 8650
    },
    {
      "epoch": 7.28643216080402,
      "grad_norm": 1.2796870470046997,
      "learning_rate": 3.3050000000000005e-06,
      "loss": 0.4009,
      "step": 8700
    },
    {
      "epoch": 7.328308207705192,
      "grad_norm": 1.6995893716812134,
      "learning_rate": 3.255e-06,
      "loss": 0.4043,
      "step": 8750
    },
    {
      "epoch": 7.370184254606365,
      "grad_norm": 1.35319983959198,
      "learning_rate": 3.2050000000000002e-06,
      "loss": 0.4043,
      "step": 8800
    },
    {
      "epoch": 7.4120603015075375,
      "grad_norm": 2.3558692932128906,
      "learning_rate": 3.1550000000000003e-06,
      "loss": 0.4048,
      "step": 8850
    },
    {
      "epoch": 7.45393634840871,
      "grad_norm": 1.3804599046707153,
      "learning_rate": 3.1050000000000003e-06,
      "loss": 0.397,
      "step": 8900
    },
    {
      "epoch": 7.4958123953098825,
      "grad_norm": 2.1145002841949463,
      "learning_rate": 3.0550000000000004e-06,
      "loss": 0.4068,
      "step": 8950
    },
    {
      "epoch": 7.5376884422110555,
      "grad_norm": 1.3504544496536255,
      "learning_rate": 3.005e-06,
      "loss": 0.4007,
      "step": 9000
    },
    {
      "epoch": 7.5376884422110555,
      "eval_loss": 0.36628037691116333,
      "eval_runtime": 213.9427,
      "eval_samples_per_second": 39.679,
      "eval_steps_per_second": 4.964,
      "step": 9000
    },
    {
      "epoch": 7.579564489112228,
      "grad_norm": 1.9124435186386108,
      "learning_rate": 2.955e-06,
      "loss": 0.399,
      "step": 9050
    },
    {
      "epoch": 7.6214405360134005,
      "grad_norm": 1.7034064531326294,
      "learning_rate": 2.905e-06,
      "loss": 0.3989,
      "step": 9100
    },
    {
      "epoch": 7.663316582914573,
      "grad_norm": 1.259141206741333,
      "learning_rate": 2.855e-06,
      "loss": 0.4029,
      "step": 9150
    },
    {
      "epoch": 7.705192629815746,
      "grad_norm": 2.1062421798706055,
      "learning_rate": 2.8050000000000007e-06,
      "loss": 0.4021,
      "step": 9200
    },
    {
      "epoch": 7.747068676716918,
      "grad_norm": 1.868788242340088,
      "learning_rate": 2.7550000000000003e-06,
      "loss": 0.4108,
      "step": 9250
    },
    {
      "epoch": 7.788944723618091,
      "grad_norm": 1.1658709049224854,
      "learning_rate": 2.7050000000000004e-06,
      "loss": 0.4059,
      "step": 9300
    },
    {
      "epoch": 7.830820770519263,
      "grad_norm": 1.230908989906311,
      "learning_rate": 2.6550000000000004e-06,
      "loss": 0.4047,
      "step": 9350
    },
    {
      "epoch": 7.872696817420436,
      "grad_norm": 1.4613652229309082,
      "learning_rate": 2.6050000000000005e-06,
      "loss": 0.4071,
      "step": 9400
    },
    {
      "epoch": 7.914572864321608,
      "grad_norm": 1.526199221611023,
      "learning_rate": 2.555e-06,
      "loss": 0.3978,
      "step": 9450
    },
    {
      "epoch": 7.956448911222781,
      "grad_norm": 1.6618636846542358,
      "learning_rate": 2.505e-06,
      "loss": 0.399,
      "step": 9500
    },
    {
      "epoch": 7.956448911222781,
      "eval_loss": 0.3649367392063141,
      "eval_runtime": 214.1,
      "eval_samples_per_second": 39.65,
      "eval_steps_per_second": 4.96,
      "step": 9500
    }
  ],
  "logging_steps": 50,
  "max_steps": 12000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 11,
  "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": 9.25136433794137e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}