{
  "best_global_step": 1008,
  "best_metric": 0.38490504026412964,
  "best_model_checkpoint": "saves/prefix-tuning/llama-3-8b-instruct/train_wsc_1754652157/checkpoint-1008",
  "epoch": 10.0,
  "eval_steps": 63,
  "global_step": 1250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "grad_norm": 2.4841861724853516,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 15.2081,
      "num_input_tokens_seen": 2144,
      "step": 5
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.424849510192871,
      "learning_rate": 3.6e-06,
      "loss": 15.3148,
      "num_input_tokens_seen": 4128,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.6069109439849854,
      "learning_rate": 5.600000000000001e-06,
      "loss": 15.3875,
      "num_input_tokens_seen": 6240,
      "step": 15
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.666384696960449,
      "learning_rate": 7.6e-06,
      "loss": 15.1604,
      "num_input_tokens_seen": 8096,
      "step": 20
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.383948564529419,
      "learning_rate": 9.600000000000001e-06,
      "loss": 15.358,
      "num_input_tokens_seen": 10112,
      "step": 25
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.555549383163452,
      "learning_rate": 1.16e-05,
      "loss": 14.8947,
      "num_input_tokens_seen": 12032,
      "step": 30
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.3365485668182373,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 14.9376,
      "num_input_tokens_seen": 13824,
      "step": 35
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.37636137008667,
      "learning_rate": 1.56e-05,
      "loss": 15.0763,
      "num_input_tokens_seen": 15840,
      "step": 40
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.863443374633789,
      "learning_rate": 1.76e-05,
      "loss": 14.7281,
      "num_input_tokens_seen": 17920,
      "step": 45
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.697338819503784,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 14.723,
      "num_input_tokens_seen": 19712,
      "step": 50
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.416774272918701,
      "learning_rate": 2.16e-05,
      "loss": 14.5953,
      "num_input_tokens_seen": 21952,
      "step": 55
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.4722342491149902,
      "learning_rate": 2.36e-05,
      "loss": 13.9838,
      "num_input_tokens_seen": 24160,
      "step": 60
    },
    {
      "epoch": 0.504,
      "eval_loss": 13.871087074279785,
      "eval_runtime": 0.8458,
      "eval_samples_per_second": 66.213,
      "eval_steps_per_second": 16.553,
      "num_input_tokens_seen": 25504,
      "step": 63
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.531996250152588,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 13.8098,
      "num_input_tokens_seen": 26112,
      "step": 65
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.4052157402038574,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 13.7342,
      "num_input_tokens_seen": 28064,
      "step": 70
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.380267381668091,
      "learning_rate": 2.96e-05,
      "loss": 13.545,
      "num_input_tokens_seen": 29824,
      "step": 75
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.2633209228515625,
      "learning_rate": 3.16e-05,
      "loss": 13.168,
      "num_input_tokens_seen": 31904,
      "step": 80
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.5283195972442627,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 13.0662,
      "num_input_tokens_seen": 33984,
      "step": 85
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.547339916229248,
      "learning_rate": 3.56e-05,
      "loss": 12.3617,
      "num_input_tokens_seen": 35776,
      "step": 90
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.478236675262451,
      "learning_rate": 3.76e-05,
      "loss": 12.2163,
      "num_input_tokens_seen": 37472,
      "step": 95
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.3510525226593018,
      "learning_rate": 3.960000000000001e-05,
      "loss": 12.0549,
      "num_input_tokens_seen": 39328,
      "step": 100
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.5149877071380615,
      "learning_rate": 4.16e-05,
      "loss": 11.9157,
      "num_input_tokens_seen": 41280,
      "step": 105
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.451765537261963,
      "learning_rate": 4.36e-05,
      "loss": 11.5939,
      "num_input_tokens_seen": 43552,
      "step": 110
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.4347667694091797,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 10.6461,
      "num_input_tokens_seen": 45216,
      "step": 115
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.257194995880127,
      "learning_rate": 4.76e-05,
      "loss": 10.384,
      "num_input_tokens_seen": 47360,
      "step": 120
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.9155919551849365,
      "learning_rate": 4.96e-05,
      "loss": 9.9251,
      "num_input_tokens_seen": 49376,
      "step": 125
    },
    {
      "epoch": 1.008,
      "eval_loss": 9.737223625183105,
      "eval_runtime": 0.8536,
      "eval_samples_per_second": 65.608,
      "eval_steps_per_second": 16.402,
      "num_input_tokens_seen": 49696,
      "step": 126
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.394456386566162,
      "learning_rate": 4.9998440375027166e-05,
      "loss": 9.5781,
      "num_input_tokens_seen": 51200,
      "step": 130
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.3213348388671875,
      "learning_rate": 4.99921047320825e-05,
      "loss": 8.9157,
      "num_input_tokens_seen": 53216,
      "step": 135
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.2454257011413574,
      "learning_rate": 4.998089682880117e-05,
      "loss": 8.0389,
      "num_input_tokens_seen": 55168,
      "step": 140
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.323469400405884,
      "learning_rate": 4.9964818850186135e-05,
      "loss": 8.6052,
      "num_input_tokens_seen": 56960,
      "step": 145
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.213332414627075,
      "learning_rate": 4.994387393067117e-05,
      "loss": 7.8179,
      "num_input_tokens_seen": 58880,
      "step": 150
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.9772940874099731,
      "learning_rate": 4.9918066153509834e-05,
      "loss": 7.3245,
      "num_input_tokens_seen": 60672,
      "step": 155
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.1031250953674316,
      "learning_rate": 4.988740054997943e-05,
      "loss": 7.6592,
      "num_input_tokens_seen": 62848,
      "step": 160
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5588639974594116,
      "learning_rate": 4.985188309840012e-05,
      "loss": 6.5144,
      "num_input_tokens_seen": 64448,
      "step": 165
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 1.6640610694885254,
      "learning_rate": 4.9811520722969465e-05,
      "loss": 6.9579,
      "num_input_tokens_seen": 66368,
      "step": 170
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.4669989347457886,
      "learning_rate": 4.976632129241252e-05,
      "loss": 5.7934,
      "num_input_tokens_seen": 68128,
      "step": 175
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.5640122890472412,
      "learning_rate": 4.971629361844785e-05,
      "loss": 6.0098,
      "num_input_tokens_seen": 70112,
      "step": 180
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.9767147302627563,
      "learning_rate": 4.966144745406961e-05,
      "loss": 6.0466,
      "num_input_tokens_seen": 72384,
      "step": 185
    },
    {
      "epoch": 1.512,
      "eval_loss": 5.308993816375732,
      "eval_runtime": 0.8536,
      "eval_samples_per_second": 65.601,
      "eval_steps_per_second": 16.4,
      "num_input_tokens_seen": 74112,
      "step": 189
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4375022649765015,
      "learning_rate": 4.960179349164621e-05,
      "loss": 5.7175,
      "num_input_tokens_seen": 74752,
      "step": 190
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.8162568807601929,
      "learning_rate": 4.953734336083583e-05,
      "loss": 5.0706,
      "num_input_tokens_seen": 76640,
      "step": 195
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.0760273933410645,
      "learning_rate": 4.946810962631916e-05,
      "loss": 5.3219,
      "num_input_tokens_seen": 78784,
      "step": 200
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 1.7913469076156616,
      "learning_rate": 4.9394105785349944e-05,
      "loss": 4.6854,
      "num_input_tokens_seen": 80768,
      "step": 205
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 1.471130609512329,
      "learning_rate": 4.9315346265123594e-05,
      "loss": 4.0843,
      "num_input_tokens_seen": 82848,
      "step": 210
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.8904441595077515,
      "learning_rate": 4.923184641996463e-05,
      "loss": 4.0373,
      "num_input_tokens_seen": 84768,
      "step": 215
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.13264536857605,
      "learning_rate": 4.914362252833332e-05,
      "loss": 3.6419,
      "num_input_tokens_seen": 86848,
      "step": 220
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.7601622343063354,
      "learning_rate": 4.905069178965215e-05,
      "loss": 3.3145,
      "num_input_tokens_seen": 88736,
      "step": 225
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 1.8561395406723022,
      "learning_rate": 4.8953072320952745e-05,
      "loss": 3.4327,
      "num_input_tokens_seen": 90848,
      "step": 230
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.9273947477340698,
      "learning_rate": 4.885078315334395e-05,
      "loss": 3.2526,
      "num_input_tokens_seen": 92992,
      "step": 235
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.0955724716186523,
      "learning_rate": 4.874384422830167e-05,
      "loss": 2.791,
      "num_input_tokens_seen": 94880,
      "step": 240
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.2840852737426758,
      "learning_rate": 4.863227639378124e-05,
      "loss": 2.2217,
      "num_input_tokens_seen": 96704,
      "step": 245
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7399126291275024,
      "learning_rate": 4.851610140015304e-05,
      "loss": 1.6788,
      "num_input_tokens_seen": 98240,
      "step": 250
    },
    {
      "epoch": 2.016,
      "eval_loss": 1.980229139328003,
      "eval_runtime": 0.8532,
      "eval_samples_per_second": 65.632,
      "eval_steps_per_second": 16.408,
      "num_input_tokens_seen": 99136,
      "step": 252
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.2386776208877563,
      "learning_rate": 4.839534189596228e-05,
      "loss": 2.0365,
      "num_input_tokens_seen": 100224,
      "step": 255
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.1297599077224731,
      "learning_rate": 4.8270021423513554e-05,
      "loss": 1.4581,
      "num_input_tokens_seen": 101920,
      "step": 260
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.090529203414917,
      "learning_rate": 4.8140164414281306e-05,
      "loss": 1.2941,
      "num_input_tokens_seen": 103808,
      "step": 265
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.0772465467453003,
      "learning_rate": 4.800579618414676e-05,
      "loss": 1.4016,
      "num_input_tokens_seen": 105920,
      "step": 270
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.179438591003418,
      "learning_rate": 4.7866942928462625e-05,
      "loss": 1.5631,
      "num_input_tokens_seen": 108160,
      "step": 275
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.1663882732391357,
      "learning_rate": 4.772363171694622e-05,
      "loss": 0.856,
      "num_input_tokens_seen": 109920,
      "step": 280
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 1.3647843599319458,
      "learning_rate": 4.7575890488402185e-05,
      "loss": 1.1216,
      "num_input_tokens_seen": 111904,
      "step": 285
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.4680769443511963,
      "learning_rate": 4.742374804527575e-05,
      "loss": 0.776,
      "num_input_tokens_seen": 113632,
      "step": 290
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.9575800895690918,
      "learning_rate": 4.7267234048037664e-05,
      "loss": 1.0421,
      "num_input_tokens_seen": 115616,
      "step": 295
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.7263527512550354,
      "learning_rate": 4.710637900940181e-05,
      "loss": 0.7994,
      "num_input_tokens_seen": 117472,
      "step": 300
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.1039828062057495,
      "learning_rate": 4.694121428837668e-05,
      "loss": 1.0609,
      "num_input_tokens_seen": 119616,
      "step": 305
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.9919302463531494,
      "learning_rate": 4.6771772084151885e-05,
      "loss": 0.8039,
      "num_input_tokens_seen": 121568,
      "step": 310
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.072923183441162,
      "learning_rate": 4.659808542982088e-05,
      "loss": 1.0818,
      "num_input_tokens_seen": 123904,
      "step": 315
    },
    {
      "epoch": 2.52,
      "eval_loss": 0.7967647314071655,
      "eval_runtime": 0.8522,
      "eval_samples_per_second": 65.715,
      "eval_steps_per_second": 16.429,
      "num_input_tokens_seen": 123904,
      "step": 315
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6072083711624146,
      "learning_rate": 4.642018818594107e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 125696,
      "step": 320
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.9770075678825378,
      "learning_rate": 4.6238115033932636e-05,
      "loss": 0.6088,
      "num_input_tokens_seen": 127488,
      "step": 325
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.5782830715179443,
      "learning_rate": 4.605190146931731e-05,
      "loss": 0.7891,
      "num_input_tokens_seen": 129632,
      "step": 330
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.686546802520752,
      "learning_rate": 4.586158379479848e-05,
      "loss": 0.7432,
      "num_input_tokens_seen": 131680,
      "step": 335
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.8748849630355835,
      "learning_rate": 4.566719911318389e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 133472,
      "step": 340
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.690437912940979,
      "learning_rate": 4.5468785320152365e-05,
      "loss": 0.5338,
      "num_input_tokens_seen": 135200,
      "step": 345
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.1224329471588135,
      "learning_rate": 4.5266381096866e-05,
      "loss": 0.9588,
      "num_input_tokens_seen": 137536,
      "step": 350
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.998359203338623,
      "learning_rate": 4.5060025902429174e-05,
      "loss": 0.7235,
      "num_input_tokens_seen": 139744,
      "step": 355
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.1624242067337036,
      "learning_rate": 4.484975996619589e-05,
      "loss": 0.703,
      "num_input_tokens_seen": 141760,
      "step": 360
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.1226049661636353,
      "learning_rate": 4.4635624279927044e-05,
      "loss": 0.6324,
      "num_input_tokens_seen": 143872,
      "step": 365
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6500603556632996,
      "learning_rate": 4.441766058979898e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 145856,
      "step": 370
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9589425325393677,
      "learning_rate": 4.4195911388264946e-05,
      "loss": 0.8394,
      "num_input_tokens_seen": 147648,
      "step": 375
    },
    {
      "epoch": 3.024,
      "eval_loss": 0.5601035952568054,
      "eval_runtime": 0.8544,
      "eval_samples_per_second": 65.545,
      "eval_steps_per_second": 16.386,
      "num_input_tokens_seen": 148736,
      "step": 378
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.4394139051437378,
      "learning_rate": 4.3970419905771145e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 149472,
      "step": 380
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.8933354020118713,
      "learning_rate": 4.374123010232888e-05,
      "loss": 0.6238,
      "num_input_tokens_seen": 151552,
      "step": 385
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.8065023422241211,
      "learning_rate": 4.350838665894446e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 153568,
      "step": 390
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.6302071213722229,
      "learning_rate": 4.3271934968908514e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 155616,
      "step": 395
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.5202668309211731,
      "learning_rate": 4.303192112894652e-05,
      "loss": 0.6315,
      "num_input_tokens_seen": 157728,
      "step": 400
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.9103642702102661,
      "learning_rate": 4.278839193023214e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 159488,
      "step": 405
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 0.6605204939842224,
      "learning_rate": 4.254139484926519e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 161600,
      "step": 410
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.8729458451271057,
      "learning_rate": 4.2290978038616e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 163712,
      "step": 415
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.7586544752120972,
      "learning_rate": 4.2037190317538e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 165536,
      "step": 420
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.765929102897644,
      "learning_rate": 4.178008116245024e-05,
      "loss": 0.5842,
      "num_input_tokens_seen": 167872,
      "step": 425
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.9758070111274719,
      "learning_rate": 4.1519700697291944e-05,
      "loss": 0.6883,
      "num_input_tokens_seen": 170112,
      "step": 430
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.6534411907196045,
      "learning_rate": 4.125609968375072e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 172000,
      "step": 435
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.6654028296470642,
      "learning_rate": 4.098932951136645e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 174016,
      "step": 440
    },
    {
      "epoch": 3.528,
      "eval_loss": 0.4781542420387268,
      "eval_runtime": 0.8555,
      "eval_samples_per_second": 65.46,
      "eval_steps_per_second": 16.365,
      "num_input_tokens_seen": 174432,
      "step": 441
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.7983071208000183,
      "learning_rate": 4.071944218751282e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 175776,
      "step": 445
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9768652319908142,
      "learning_rate": 4.044649032725836e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 177952,
      "step": 450
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.1632089614868164,
      "learning_rate": 4.017052714310906e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 179968,
      "step": 455
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.7327042818069458,
      "learning_rate": 3.989160643463445e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 181952,
      "step": 460
    },
    {
      "epoch": 3.7199999999999998,
      "grad_norm": 1.0546882152557373,
      "learning_rate": 3.960978257797931e-05,
      "loss": 0.3364,
      "num_input_tokens_seen": 183680,
      "step": 465
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.5163419246673584,
      "learning_rate": 3.932511051526289e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 185632,
      "step": 470
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.192272663116455,
      "learning_rate": 3.903764574386786e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 187552,
      "step": 475
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.180910348892212,
      "learning_rate": 3.8747444305621e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 189408,
      "step": 480
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7053818702697754,
      "learning_rate": 3.8454562775867684e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 191488,
      "step": 485
    },
    {
      "epoch": 3.92,
      "grad_norm": 1.434988260269165,
      "learning_rate": 3.8159058252442446e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 193312,
      "step": 490
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.8368187546730042,
      "learning_rate": 3.786098834453766e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 195424,
      "step": 495
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.403336763381958,
      "learning_rate": 3.7560411161472456e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 197024,
      "step": 500
    },
    {
      "epoch": 4.032,
      "eval_loss": 0.46134254336357117,
      "eval_runtime": 0.8491,
      "eval_samples_per_second": 65.955,
      "eval_steps_per_second": 16.489,
      "num_input_tokens_seen": 198656,
      "step": 504
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.6058657169342041,
      "learning_rate": 3.725738530136422e-05,
      "loss": 0.5412,
      "num_input_tokens_seen": 199040,
      "step": 505
    },
    {
      "epoch": 4.08,
      "grad_norm": 1.2126246690750122,
      "learning_rate": 3.695196983970481e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 200960,
      "step": 510
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.5591365098953247,
      "learning_rate": 3.664422431784361e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 203008,
      "step": 515
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.040564775466919,
      "learning_rate": 3.633420873137988e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 204672,
      "step": 520
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.4747677743434906,
      "learning_rate": 3.602198351846647e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 206784,
      "step": 525
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.8027376532554626,
      "learning_rate": 3.570760954802726e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 208672,
      "step": 530
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.7639246582984924,
      "learning_rate": 3.53911481078907e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 210752,
      "step": 535
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7363210320472717,
      "learning_rate": 3.507266089284157e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 213472,
      "step": 540
    },
    {
      "epoch": 4.36,
      "grad_norm": 1.636357307434082,
      "learning_rate": 3.475220999259349e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 215616,
      "step": 545
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.7487366795539856,
      "learning_rate": 3.442985787968442e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 217664,
      "step": 550
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.5726836323738098,
      "learning_rate": 3.410566739729746e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 219584,
      "step": 555
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.7676336169242859,
      "learning_rate": 3.3779701747009504e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 221504,
      "step": 560
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.8204706907272339,
      "learning_rate": 3.3452024476469934e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 223424,
      "step": 565
    },
    {
      "epoch": 4.536,
      "eval_loss": 0.43884095549583435,
      "eval_runtime": 0.8563,
      "eval_samples_per_second": 65.4,
      "eval_steps_per_second": 16.35,
      "num_input_tokens_seen": 224032,
      "step": 567
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 1.2742042541503906,
      "learning_rate": 3.312269946701191e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 225216,
      "step": 570
    },
    {
      "epoch": 4.6,
      "grad_norm": 1.1421749591827393,
      "learning_rate": 3.279179092119855e-05,
      "loss": 0.3681,
      "num_input_tokens_seen": 227008,
      "step": 575
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.359496831893921,
      "learning_rate": 3.245936335030651e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 228736,
      "step": 580
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7598556280136108,
      "learning_rate": 3.21254815617494e-05,
      "loss": 0.3985,
      "num_input_tokens_seen": 230240,
      "step": 585
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.4738464057445526,
      "learning_rate": 3.179021064644347e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 232192,
      "step": 590
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.8662070035934448,
      "learning_rate": 3.145361596611795e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 234368,
      "step": 595
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7484014630317688,
      "learning_rate": 3.111576314057268e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 236032,
      "step": 600
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.9487155079841614,
      "learning_rate": 3.0776718034885454e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 237920,
      "step": 605
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.6988774538040161,
      "learning_rate": 3.0436546746571372e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 239680,
      "step": 610
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.8400317430496216,
      "learning_rate": 3.0095315592697126e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 241504,
      "step": 615
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.9846633672714233,
      "learning_rate": 2.9753091096952255e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 243584,
      "step": 620
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6307002902030945,
      "learning_rate": 2.9409939976680313e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 245472,
      "step": 625
    },
    {
      "epoch": 5.04,
      "grad_norm": 1.2824044227600098,
      "learning_rate": 2.9065929129872094e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 247424,
      "step": 630
    },
    {
      "epoch": 5.04,
      "eval_loss": 0.42154452204704285,
      "eval_runtime": 0.8526,
      "eval_samples_per_second": 65.682,
      "eval_steps_per_second": 16.42,
      "num_input_tokens_seen": 247424,
      "step": 630
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.6836050152778625,
      "learning_rate": 2.8721125622123806e-05,
      "loss": 0.3778,
      "num_input_tokens_seen": 249472,
      "step": 635
    },
    {
      "epoch": 5.12,
      "grad_norm": 1.2711609601974487,
      "learning_rate": 2.8375596673562482e-05,
      "loss": 0.3189,
      "num_input_tokens_seen": 251296,
      "step": 640
    },
    {
      "epoch": 5.16,
      "grad_norm": 1.1848688125610352,
      "learning_rate": 2.8029409645741267e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 253344,
      "step": 645
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.43571653962135315,
      "learning_rate": 2.7682632028507167e-05,
      "loss": 0.3687,
      "num_input_tokens_seen": 255104,
      "step": 650
    },
    {
      "epoch": 5.24,
      "grad_norm": 2.1147029399871826,
      "learning_rate": 2.733533142684377e-05,
      "loss": 0.3907,
      "num_input_tokens_seen": 256832,
      "step": 655
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.5585914850234985,
      "learning_rate": 2.6987575547691497e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 258720,
      "step": 660
    },
    {
      "epoch": 5.32,
      "grad_norm": 1.013206124305725,
      "learning_rate": 2.6639432186748043e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 260576,
      "step": 665
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.6060461401939392,
      "learning_rate": 2.6290969215251416e-05,
      "loss": 0.3817,
      "num_input_tokens_seen": 262368,
      "step": 670
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.9579009413719177,
      "learning_rate": 2.594225456674837e-05,
      "loss": 0.3826,
      "num_input_tokens_seen": 264320,
      "step": 675
    },
    {
      "epoch": 5.44,
      "grad_norm": 1.1483041048049927,
      "learning_rate": 2.559335622385055e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 266304,
      "step": 680
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.8167299628257751,
      "learning_rate": 2.524434220498123e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 268384,
      "step": 685
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.5977395176887512,
      "learning_rate": 2.4895280551114907e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 270208,
      "step": 690
    },
    {
      "epoch": 5.5440000000000005,
      "eval_loss": 0.4073176383972168,
      "eval_runtime": 0.8504,
      "eval_samples_per_second": 65.848,
      "eval_steps_per_second": 16.462,
      "num_input_tokens_seen": 271232,
      "step": 693
    },
    {
      "epoch": 5.5600000000000005,
      "grad_norm": 1.8195998668670654,
      "learning_rate": 2.4546239312512635e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 271840,
      "step": 695
    },
    {
      "epoch": 5.6,
      "grad_norm": 1.045985460281372,
      "learning_rate": 2.4197286535455464e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 273888,
      "step": 700
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.0764570236206055,
      "learning_rate": 2.384849024897869e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 275904,
      "step": 705
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.2045237421989441,
      "learning_rate": 2.349991845160949e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 277888,
      "step": 710
    },
    {
      "epoch": 5.72,
      "grad_norm": 1.0971755981445312,
      "learning_rate": 2.3151639098110377e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 279872,
      "step": 715
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.39166244864463806,
      "learning_rate": 2.280372008623142e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 281664,
      "step": 720
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.7744928002357483,
      "learning_rate": 2.2456229243473345e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 283776,
      "step": 725
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.4687577486038208,
      "learning_rate": 2.2109234313864465e-05,
      "loss": 0.3788,
      "num_input_tokens_seen": 285568,
      "step": 730
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.439230352640152,
      "learning_rate": 2.176280294475383e-05,
      "loss": 0.3795,
      "num_input_tokens_seen": 287360,
      "step": 735
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.6617324948310852,
      "learning_rate": 2.1417002673623264e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 289632,
      "step": 740
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.32485124468803406,
      "learning_rate": 2.1071900914920816e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 291552,
      "step": 745
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.3294658660888672,
      "learning_rate": 2.0727564946918087e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 293616,
      "step": 750
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.7842015027999878,
      "learning_rate": 2.038406189859433e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 295440,
      "step": 755
    },
    {
      "epoch": 6.048,
      "eval_loss": 0.40050727128982544,
      "eval_runtime": 0.8608,
      "eval_samples_per_second": 65.059,
      "eval_steps_per_second": 16.265,
      "num_input_tokens_seen": 295728,
      "step": 756
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.7505316734313965,
      "learning_rate": 2.004145873654942e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 297360,
      "step": 760
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.787607729434967,
      "learning_rate": 1.969982225194864e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 299312,
      "step": 765
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.8527250289916992,
      "learning_rate": 1.9359219047501565e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 301488,
      "step": 770
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.8097345232963562,
      "learning_rate": 1.9019715524477767e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 303696,
      "step": 775
    },
    {
      "epoch": 6.24,
      "grad_norm": 1.303476095199585,
      "learning_rate": 1.868137786976177e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 305360,
      "step": 780
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.7829755544662476,
      "learning_rate": 1.8344272042949724e-05,
      "loss": 0.3779,
      "num_input_tokens_seen": 307408,
      "step": 785
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.659096360206604,
      "learning_rate": 1.800846376349051e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 309232,
      "step": 790
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.6200412511825562,
      "learning_rate": 1.767401849787357e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 311184,
      "step": 795
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.5834780335426331,
      "learning_rate": 1.73410014468661e-05,
      "loss": 0.3727,
      "num_input_tokens_seen": 313072,
      "step": 800
    },
    {
      "epoch": 6.44,
      "grad_norm": 1.6976076364517212,
      "learning_rate": 1.7009477532802054e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 314832,
      "step": 805
    },
    {
      "epoch": 6.48,
      "grad_norm": 1.1492793560028076,
      "learning_rate": 1.6679511386925337e-05,
      "loss": 0.3356,
      "num_input_tokens_seen": 316560,
      "step": 810
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.6304459571838379,
      "learning_rate": 1.635116733678988e-05,
      "loss": 0.427,
      "num_input_tokens_seen": 318960,
      "step": 815
    },
    {
      "epoch": 6.552,
      "eval_loss": 0.40054336190223694,
      "eval_runtime": 0.8527,
      "eval_samples_per_second": 65.674,
      "eval_steps_per_second": 16.418,
      "num_input_tokens_seen": 320464,
      "step": 819
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 0.825368344783783,
      "learning_rate": 1.6024509393718844e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 320880,
      "step": 820
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.5931684970855713,
      "learning_rate": 1.5699601240325474e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 323184,
      "step": 825
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.8659843802452087,
      "learning_rate": 1.5376506218098015e-05,
      "loss": 0.371,
      "num_input_tokens_seen": 325168,
      "step": 830
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.6564781069755554,
      "learning_rate": 1.505528731505126e-05,
      "loss": 0.3795,
      "num_input_tokens_seen": 326992,
      "step": 835
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.9912712574005127,
      "learning_rate": 1.4736007153446801e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 329104,
      "step": 840
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.7603425979614258,
      "learning_rate": 1.4418727977584774e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 331088,
      "step": 845
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.5329681038856506,
      "learning_rate": 1.4103511641669152e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 333008,
      "step": 850
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.5622363686561584,
      "learning_rate": 1.3790419597749199e-05,
      "loss": 0.3725,
      "num_input_tokens_seen": 335024,
      "step": 855
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.9950863718986511,
      "learning_rate": 1.3479512883739232e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 337104,
      "step": 860
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.7129436135292053,
      "learning_rate": 1.3170852111519175e-05,
      "loss": 0.3773,
      "num_input_tokens_seen": 338960,
      "step": 865
    },
    {
      "epoch": 6.96,
      "grad_norm": 1.585489273071289,
      "learning_rate": 1.2864497455118152e-05,
      "loss": 0.3702,
      "num_input_tokens_seen": 340848,
      "step": 870
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.2564765214920044,
      "learning_rate": 1.2560508638983437e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 343040,
      "step": 875
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.6992194056510925,
      "learning_rate": 1.2258944926337057e-05,
      "loss": 0.3347,
      "num_input_tokens_seen": 345056,
      "step": 880
    },
    {
      "epoch": 7.056,
      "eval_loss": 0.41065576672554016,
      "eval_runtime": 0.8558,
      "eval_samples_per_second": 65.435,
      "eval_steps_per_second": 16.359,
      "num_input_tokens_seen": 345856,
      "step": 882
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.7301174402236938,
      "learning_rate": 1.1959865107622307e-05,
      "loss": 0.3597,
      "num_input_tokens_seen": 347232,
      "step": 885
    },
    {
      "epoch": 7.12,
      "grad_norm": 1.776132345199585,
      "learning_rate": 1.1663327489042435e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 349504,
      "step": 890
    },
    {
      "epoch": 7.16,
      "grad_norm": 0.7216264605522156,
      "learning_rate": 1.1369389881193749e-05,
      "loss": 0.4164,
      "num_input_tokens_seen": 351296,
      "step": 895
    },
    {
      "epoch": 7.2,
      "grad_norm": 1.4257539510726929,
      "learning_rate": 1.107810958779531e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 353248,
      "step": 900
    },
    {
      "epoch": 7.24,
      "grad_norm": 1.8053635358810425,
      "learning_rate": 1.0789543394517435e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 355232,
      "step": 905
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.6325796246528625,
      "learning_rate": 1.050374755791127e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 357376,
      "step": 910
    },
    {
      "epoch": 7.32,
      "grad_norm": 1.1207585334777832,
      "learning_rate": 1.022077779444145e-05,
      "loss": 0.3539,
      "num_input_tokens_seen": 359232,
      "step": 915
    },
    {
      "epoch": 7.36,
      "grad_norm": 1.4070286750793457,
      "learning_rate": 9.94068926962404e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 361024,
      "step": 920
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.627964973449707,
      "learning_rate": 9.663536587271902e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 363200,
      "step": 925
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.6566188335418701,
      "learning_rate": 9.389373778849612e-06,
      "loss": 0.3517,
      "num_input_tokens_seen": 365152,
      "step": 930
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.6392218470573425,
      "learning_rate": 9.11825429293989e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 367328,
      "step": 935
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.5262883901596069,
      "learning_rate": 8.850230984823735e-06,
      "loss": 0.358,
      "num_input_tokens_seen": 369248,
      "step": 940
    },
    {
      "epoch": 7.5600000000000005,
      "grad_norm": 0.9610680937767029,
      "learning_rate": 8.585356106176094e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 371040,
      "step": 945
    },
    {
      "epoch": 7.5600000000000005,
      "eval_loss": 0.4088888466358185,
      "eval_runtime": 0.8481,
      "eval_samples_per_second": 66.028,
      "eval_steps_per_second": 16.507,
      "num_input_tokens_seen": 371040,
      "step": 945
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.7290608882904053,
      "learning_rate": 8.323681294879394e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 372928,
      "step": 950
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.532922089099884,
      "learning_rate": 8.06525756495657e-06,
      "loss": 0.4047,
      "num_input_tokens_seen": 374816,
      "step": 955
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.77403724193573,
      "learning_rate": 7.810135296625818e-06,
      "loss": 0.3748,
      "num_input_tokens_seen": 376704,
      "step": 960
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.5218887329101562,
      "learning_rate": 7.558364226478842e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 378624,
      "step": 965
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.5533618927001953,
      "learning_rate": 7.309993437784624e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 380480,
      "step": 970
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.6465098261833191,
      "learning_rate": 7.065071350920538e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 382144,
      "step": 975
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.6888776421546936,
      "learning_rate": 6.823645713932708e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 384320,
      "step": 980
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.6575106382369995,
      "learning_rate": 6.58576359322742e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 386656,
      "step": 985
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.4115797281265259,
      "learning_rate": 6.3514713643954475e-06,
      "loss": 0.3612,
      "num_input_tokens_seen": 388672,
      "step": 990
    },
    {
      "epoch": 7.96,
      "grad_norm": 1.37602698802948,
      "learning_rate": 6.120814703171024e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 390400,
      "step": 995
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.3773898780345917,
      "learning_rate": 5.893838576527275e-06,
      "loss": 0.3604,
      "num_input_tokens_seen": 392080,
      "step": 1000
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.7060356140136719,
      "learning_rate": 5.6705872339098186e-06,
      "loss": 0.4144,
      "num_input_tokens_seen": 394160,
      "step": 1005
    },
    {
      "epoch": 8.064,
      "eval_loss": 0.38490504026412964,
      "eval_runtime": 0.8524,
      "eval_samples_per_second": 65.693,
      "eval_steps_per_second": 16.423,
      "num_input_tokens_seen": 395216,
      "step": 1008
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.7593961954116821,
      "learning_rate": 5.451104198610249e-06,
      "loss": 0.3568,
      "num_input_tokens_seen": 395888,
      "step": 1010
    },
    {
      "epoch": 8.12,
      "grad_norm": 1.2459512948989868,
      "learning_rate": 5.235432259281175e-06,
      "loss": 0.3882,
      "num_input_tokens_seen": 398032,
      "step": 1015
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.0449175834655762,
      "learning_rate": 5.023613461594512e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 399856,
      "step": 1020
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.7643663287162781,
      "learning_rate": 4.8156891000445406e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 401616,
      "step": 1025
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.7826510071754456,
      "learning_rate": 4.6116997098975465e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 403568,
      "step": 1030
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.6359825134277344,
      "learning_rate": 4.411685059289314e-06,
      "loss": 0.3725,
      "num_input_tokens_seen": 405712,
      "step": 1035
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.4394562244415283,
      "learning_rate": 4.215684141472292e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 407888,
      "step": 1040
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.6596431732177734,
      "learning_rate": 4.023735167213752e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 409712,
      "step": 1045
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.9594627618789673,
      "learning_rate": 3.835875557346552e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 411504,
      "step": 1050
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.5970214009284973,
      "learning_rate": 3.6521419354738738e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 413168,
      "step": 1055
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.4645940959453583,
      "learning_rate": 3.4725701208293435e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 414960,
      "step": 1060
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.6897419691085815,
      "learning_rate": 3.297195121294022e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 416880,
      "step": 1065
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.8192169666290283,
      "learning_rate": 3.126051126571561e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 418768,
      "step": 1070
    },
    {
      "epoch": 8.568,
      "eval_loss": 0.38693496584892273,
      "eval_runtime": 0.8505,
      "eval_samples_per_second": 65.843,
      "eval_steps_per_second": 16.461,
      "num_input_tokens_seen": 419184,
      "step": 1071
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.9851546287536621,
      "learning_rate": 2.9591715015228284e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 421008,
      "step": 1075
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.5261719226837158,
      "learning_rate": 2.7965887796613884e-06,
      "loss": 0.3589,
      "num_input_tokens_seen": 422864,
      "step": 1080
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.7083816528320312,
      "learning_rate": 2.6383346568110062e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 424976,
      "step": 1085
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.8124401569366455,
      "learning_rate": 2.4844399849264928e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 427120,
      "step": 1090
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.7263258695602417,
      "learning_rate": 2.3349347660790582e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 429264,
      "step": 1095
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.6860739588737488,
      "learning_rate": 2.189848146607348e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 431056,
      "step": 1100
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.49945640563964844,
      "learning_rate": 2.0492084114352965e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 433136,
      "step": 1105
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.8145077228546143,
      "learning_rate": 1.913042978557944e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 435216,
      "step": 1110
    },
    {
      "epoch": 8.92,
      "grad_norm": 1.1362378597259521,
      "learning_rate": 1.7813783936962258e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 437040,
      "step": 1115
    },
    {
      "epoch": 8.96,
      "grad_norm": 1.012280821800232,
      "learning_rate": 1.654240325121831e-06,
      "loss": 0.3873,
      "num_input_tokens_seen": 439088,
      "step": 1120
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.0121548175811768,
      "learning_rate": 1.5316535586531483e-06,
      "loss": 0.4015,
      "num_input_tokens_seen": 440848,
      "step": 1125
    },
    {
      "epoch": 9.04,
      "grad_norm": 0.9955313205718994,
      "learning_rate": 1.4136419928231892e-06,
      "loss": 0.3714,
      "num_input_tokens_seen": 442864,
      "step": 1130
    },
    {
      "epoch": 9.072,
      "eval_loss": 0.38989585638046265,
      "eval_runtime": 0.8537,
      "eval_samples_per_second": 65.599,
      "eval_steps_per_second": 16.4,
      "num_input_tokens_seen": 444560,
      "step": 1134
    },
    {
      "epoch": 9.08,
      "grad_norm": 0.6308397054672241,
      "learning_rate": 1.3002286342205462e-06,
      "loss": 0.3985,
      "num_input_tokens_seen": 445040,
      "step": 1135
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.438340961933136,
      "learning_rate": 1.1914355930041837e-06,
      "loss": 0.3583,
      "num_input_tokens_seen": 446864,
      "step": 1140
    },
    {
      "epoch": 9.16,
      "grad_norm": 0.80867600440979,
      "learning_rate": 1.087284078593051e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 449008,
      "step": 1145
    },
    {
      "epoch": 9.2,
      "grad_norm": 1.2045706510543823,
      "learning_rate": 9.877943955312552e-07,
      "loss": 0.3471,
      "num_input_tokens_seen": 450832,
      "step": 1150
    },
    {
      "epoch": 9.24,
      "grad_norm": 1.406896948814392,
      "learning_rate": 8.929859395296364e-07,
      "loss": 0.3737,
      "num_input_tokens_seen": 452784,
      "step": 1155
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.849940836429596,
      "learning_rate": 8.028771936845342e-07,
      "loss": 0.3719,
      "num_input_tokens_seen": 454992,
      "step": 1160
    },
    {
      "epoch": 9.32,
      "grad_norm": 0.889966607093811,
      "learning_rate": 7.174857248745004e-07,
      "loss": 0.3571,
      "num_input_tokens_seen": 456528,
      "step": 1165
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.1385633945465088,
      "learning_rate": 6.368281803355691e-07,
      "loss": 0.3921,
      "num_input_tokens_seen": 458768,
      "step": 1170
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.6559231877326965,
      "learning_rate": 5.609202844158723e-07,
      "loss": 0.3935,
      "num_input_tokens_seen": 460816,
      "step": 1175
    },
    {
      "epoch": 9.44,
      "grad_norm": 0.8425350785255432,
      "learning_rate": 4.897768355101084e-07,
      "loss": 0.3657,
      "num_input_tokens_seen": 462384,
      "step": 1180
    },
    {
      "epoch": 9.48,
      "grad_norm": 0.9185330271720886,
      "learning_rate": 4.234117031746143e-07,
      "loss": 0.3923,
      "num_input_tokens_seen": 464304,
      "step": 1185
    },
    {
      "epoch": 9.52,
      "grad_norm": 0.5886545777320862,
      "learning_rate": 3.6183782542343057e-07,
      "loss": 0.3341,
      "num_input_tokens_seen": 466384,
      "step": 1190
    },
    {
      "epoch": 9.56,
      "grad_norm": 0.7718898057937622,
      "learning_rate": 3.050672062060278e-07,
      "loss": 0.3858,
      "num_input_tokens_seen": 468368,
      "step": 1195
    },
    {
      "epoch": 9.576,
      "eval_loss": 0.3861676752567291,
      "eval_runtime": 0.8505,
      "eval_samples_per_second": 65.845,
      "eval_steps_per_second": 16.461,
      "num_input_tokens_seen": 469104,
      "step": 1197
    },
    {
      "epoch": 9.6,
      "grad_norm": 1.3115769624710083,
      "learning_rate": 2.531109130671061e-07,
      "loss": 0.3761,
      "num_input_tokens_seen": 470192,
      "step": 1200
    },
    {
      "epoch": 9.64,
      "grad_norm": 0.625157356262207,
      "learning_rate": 2.0597907498896007e-07,
      "loss": 0.3878,
      "num_input_tokens_seen": 472240,
      "step": 1205
    },
    {
      "epoch": 9.68,
      "grad_norm": 1.4318360090255737,
      "learning_rate": 1.6368088041681108e-07,
      "loss": 0.3748,
      "num_input_tokens_seen": 474160,
      "step": 1210
    },
    {
      "epoch": 9.72,
      "grad_norm": 0.62953782081604,
      "learning_rate": 1.2622457546749567e-07,
      "loss": 0.4067,
      "num_input_tokens_seen": 476784,
      "step": 1215
    },
    {
      "epoch": 9.76,
      "grad_norm": 1.3049143552780151,
      "learning_rate": 9.361746232188495e-08,
      "loss": 0.3679,
      "num_input_tokens_seen": 478864,
      "step": 1220
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.5870345234870911,
      "learning_rate": 6.586589780128716e-08,
      "loss": 0.3603,
      "num_input_tokens_seen": 480816,
      "step": 1225
    },
    {
      "epoch": 9.84,
      "grad_norm": 0.5798351764678955,
      "learning_rate": 4.2975292128200064e-08,
      "loss": 0.384,
      "num_input_tokens_seen": 482832,
      "step": 1230
    },
    {
      "epoch": 9.88,
      "grad_norm": 0.6793395280838013,
      "learning_rate": 2.4950107871549167e-08,
      "loss": 0.3827,
      "num_input_tokens_seen": 484912,
      "step": 1235
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.9757604002952576,
      "learning_rate": 1.179385907672248e-08,
      "loss": 0.3863,
      "num_input_tokens_seen": 486640,
      "step": 1240
    },
    {
      "epoch": 9.96,
      "grad_norm": 0.7342190742492676,
      "learning_rate": 3.5091105804907487e-09,
      "loss": 0.3537,
      "num_input_tokens_seen": 488336,
      "step": 1245
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.7796341180801392,
      "learning_rate": 9.747751098521107e-11,
      "loss": 0.3453,
      "num_input_tokens_seen": 490000,
      "step": 1250
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 490000,
      "step": 1250,
      "total_flos": 2.206447853568e+16,
      "train_loss": 2.282332957649231,
      "train_runtime": 172.1756,
      "train_samples_per_second": 28.924,
      "train_steps_per_second": 7.26
    }
  ],
  "logging_steps": 5,
  "max_steps": 1250,
  "num_input_tokens_seen": 490000,
  "num_train_epochs": 10,
  "save_steps": 63,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.206447853568e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}