{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 1536,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0390625,
      "grad_norm": 1.6853455305099487,
      "learning_rate": 8e-05,
      "loss": 2.2667,
      "step": 5
    },
    {
      "epoch": 0.078125,
      "grad_norm": 1.127389907836914,
      "learning_rate": 0.00018,
      "loss": 1.2747,
      "step": 10
    },
    {
      "epoch": 0.1171875,
      "grad_norm": 0.6251115798950195,
      "learning_rate": 0.00019947575360419398,
      "loss": 1.2874,
      "step": 15
    },
    {
      "epoch": 0.15625,
      "grad_norm": 0.8021469116210938,
      "learning_rate": 0.00019882044560943645,
      "loss": 1.1816,
      "step": 20
    },
    {
      "epoch": 0.1953125,
      "grad_norm": 0.7008321285247803,
      "learning_rate": 0.0001981651376146789,
      "loss": 0.9091,
      "step": 25
    },
    {
      "epoch": 0.234375,
      "grad_norm": 0.8806556463241577,
      "learning_rate": 0.00019750982961992138,
      "loss": 1.18,
      "step": 30
    },
    {
      "epoch": 0.2734375,
      "grad_norm": 0.4898707866668701,
      "learning_rate": 0.00019685452162516385,
      "loss": 0.8613,
      "step": 35
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.818252444267273,
      "learning_rate": 0.0001961992136304063,
      "loss": 0.9613,
      "step": 40
    },
    {
      "epoch": 0.3515625,
      "grad_norm": 0.7560004591941833,
      "learning_rate": 0.00019554390563564878,
      "loss": 0.9433,
      "step": 45
    },
    {
      "epoch": 0.390625,
      "grad_norm": 0.5985464453697205,
      "learning_rate": 0.00019488859764089122,
      "loss": 0.8993,
      "step": 50
    },
    {
      "epoch": 0.4296875,
      "grad_norm": 0.7984416484832764,
      "learning_rate": 0.0001942332896461337,
      "loss": 0.9395,
      "step": 55
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.5905727744102478,
      "learning_rate": 0.00019357798165137616,
      "loss": 0.9598,
      "step": 60
    },
    {
      "epoch": 0.5078125,
      "grad_norm": 0.47333383560180664,
      "learning_rate": 0.00019292267365661863,
      "loss": 0.9031,
      "step": 65
    },
    {
      "epoch": 0.546875,
      "grad_norm": 0.5469959378242493,
      "learning_rate": 0.00019226736566186107,
      "loss": 1.1451,
      "step": 70
    },
    {
      "epoch": 0.5859375,
      "grad_norm": 0.5932920575141907,
      "learning_rate": 0.00019161205766710356,
      "loss": 0.8857,
      "step": 75
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.5339898467063904,
      "learning_rate": 0.000190956749672346,
      "loss": 0.7919,
      "step": 80
    },
    {
      "epoch": 0.6640625,
      "grad_norm": 0.48470577597618103,
      "learning_rate": 0.00019030144167758847,
      "loss": 0.9792,
      "step": 85
    },
    {
      "epoch": 0.703125,
      "grad_norm": 0.4082311689853668,
      "learning_rate": 0.00018964613368283094,
      "loss": 0.9229,
      "step": 90
    },
    {
      "epoch": 0.7421875,
      "grad_norm": 0.6408493518829346,
      "learning_rate": 0.0001889908256880734,
      "loss": 0.7129,
      "step": 95
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.45103052258491516,
      "learning_rate": 0.00018833551769331587,
      "loss": 0.9419,
      "step": 100
    },
    {
      "epoch": 0.8203125,
      "grad_norm": 0.6506906747817993,
      "learning_rate": 0.00018768020969855834,
      "loss": 0.836,
      "step": 105
    },
    {
      "epoch": 0.859375,
      "grad_norm": 0.6588282585144043,
      "learning_rate": 0.0001870249017038008,
      "loss": 0.885,
      "step": 110
    },
    {
      "epoch": 0.8984375,
      "grad_norm": 0.5844029188156128,
      "learning_rate": 0.00018636959370904325,
      "loss": 0.8624,
      "step": 115
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.538287878036499,
      "learning_rate": 0.00018571428571428572,
      "loss": 0.9129,
      "step": 120
    },
    {
      "epoch": 0.9765625,
      "grad_norm": 0.3959498107433319,
      "learning_rate": 0.00018505897771952819,
      "loss": 1.0968,
      "step": 125
    },
    {
      "epoch": 1.015625,
      "grad_norm": 0.46326589584350586,
      "learning_rate": 0.00018440366972477065,
      "loss": 0.7563,
      "step": 130
    },
    {
      "epoch": 1.0546875,
      "grad_norm": 0.6401046514511108,
      "learning_rate": 0.00018374836173001312,
      "loss": 0.7854,
      "step": 135
    },
    {
      "epoch": 1.09375,
      "grad_norm": 0.6093747615814209,
      "learning_rate": 0.0001830930537352556,
      "loss": 0.7031,
      "step": 140
    },
    {
      "epoch": 1.1328125,
      "grad_norm": 0.48366278409957886,
      "learning_rate": 0.00018243774574049803,
      "loss": 0.4555,
      "step": 145
    },
    {
      "epoch": 1.171875,
      "grad_norm": 0.5257757902145386,
      "learning_rate": 0.0001817824377457405,
      "loss": 0.6499,
      "step": 150
    },
    {
      "epoch": 1.2109375,
      "grad_norm": 0.7223408818244934,
      "learning_rate": 0.00018112712975098296,
      "loss": 0.5893,
      "step": 155
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.4492509663105011,
      "learning_rate": 0.00018047182175622543,
      "loss": 0.8623,
      "step": 160
    },
    {
      "epoch": 1.2890625,
      "grad_norm": 0.6466461420059204,
      "learning_rate": 0.0001798165137614679,
      "loss": 0.7997,
      "step": 165
    },
    {
      "epoch": 1.328125,
      "grad_norm": 0.6021189093589783,
      "learning_rate": 0.00017916120576671037,
      "loss": 0.7151,
      "step": 170
    },
    {
      "epoch": 1.3671875,
      "grad_norm": 0.43464839458465576,
      "learning_rate": 0.00017850589777195283,
      "loss": 0.626,
      "step": 175
    },
    {
      "epoch": 1.40625,
      "grad_norm": 0.49049654603004456,
      "learning_rate": 0.00017785058977719527,
      "loss": 0.7601,
      "step": 180
    },
    {
      "epoch": 1.4453125,
      "grad_norm": 0.6579009294509888,
      "learning_rate": 0.00017719528178243777,
      "loss": 0.6411,
      "step": 185
    },
    {
      "epoch": 1.484375,
      "grad_norm": 0.7494032382965088,
      "learning_rate": 0.0001765399737876802,
      "loss": 0.589,
      "step": 190
    },
    {
      "epoch": 1.5234375,
      "grad_norm": 0.5080376267433167,
      "learning_rate": 0.00017588466579292268,
      "loss": 0.6848,
      "step": 195
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.49630534648895264,
      "learning_rate": 0.00017522935779816515,
      "loss": 0.6273,
      "step": 200
    },
    {
      "epoch": 1.6015625,
      "grad_norm": 0.6087814569473267,
      "learning_rate": 0.0001745740498034076,
      "loss": 0.6083,
      "step": 205
    },
    {
      "epoch": 1.640625,
      "grad_norm": 0.607954740524292,
      "learning_rate": 0.00017391874180865005,
      "loss": 0.8664,
      "step": 210
    },
    {
      "epoch": 1.6796875,
      "grad_norm": 0.44959601759910583,
      "learning_rate": 0.00017326343381389255,
      "loss": 0.5538,
      "step": 215
    },
    {
      "epoch": 1.71875,
      "grad_norm": 0.5550365447998047,
      "learning_rate": 0.000172608125819135,
      "loss": 0.4869,
      "step": 220
    },
    {
      "epoch": 1.7578125,
      "grad_norm": 0.6531190872192383,
      "learning_rate": 0.00017195281782437746,
      "loss": 0.7142,
      "step": 225
    },
    {
      "epoch": 1.796875,
      "grad_norm": 0.6506574153900146,
      "learning_rate": 0.00017129750982961995,
      "loss": 0.6573,
      "step": 230
    },
    {
      "epoch": 1.8359375,
      "grad_norm": 0.5597310662269592,
      "learning_rate": 0.0001706422018348624,
      "loss": 0.6261,
      "step": 235
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.5404195189476013,
      "learning_rate": 0.00016998689384010486,
      "loss": 0.5054,
      "step": 240
    },
    {
      "epoch": 1.9140625,
      "grad_norm": 0.611003041267395,
      "learning_rate": 0.00016933158584534733,
      "loss": 0.6949,
      "step": 245
    },
    {
      "epoch": 1.953125,
      "grad_norm": 0.4925813674926758,
      "learning_rate": 0.0001686762778505898,
      "loss": 0.6684,
      "step": 250
    },
    {
      "epoch": 1.9921875,
      "grad_norm": 0.5423117876052856,
      "learning_rate": 0.00016802096985583224,
      "loss": 0.7782,
      "step": 255
    },
    {
      "epoch": 2.03125,
      "grad_norm": 0.4928165078163147,
      "learning_rate": 0.00016736566186107473,
      "loss": 0.4515,
      "step": 260
    },
    {
      "epoch": 2.0703125,
      "grad_norm": 0.6966648101806641,
      "learning_rate": 0.00016671035386631717,
      "loss": 0.4123,
      "step": 265
    },
    {
      "epoch": 2.109375,
      "grad_norm": 0.7156907916069031,
      "learning_rate": 0.00016605504587155964,
      "loss": 0.591,
      "step": 270
    },
    {
      "epoch": 2.1484375,
      "grad_norm": 0.5283113718032837,
      "learning_rate": 0.0001653997378768021,
      "loss": 0.4631,
      "step": 275
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.7045680284500122,
      "learning_rate": 0.00016474442988204457,
      "loss": 0.4343,
      "step": 280
    },
    {
      "epoch": 2.2265625,
      "grad_norm": 0.7731931805610657,
      "learning_rate": 0.00016408912188728701,
      "loss": 0.4591,
      "step": 285
    },
    {
      "epoch": 2.265625,
      "grad_norm": 0.7124219536781311,
      "learning_rate": 0.0001634338138925295,
      "loss": 0.4534,
      "step": 290
    },
    {
      "epoch": 2.3046875,
      "grad_norm": 0.66915363073349,
      "learning_rate": 0.00016277850589777198,
      "loss": 0.5908,
      "step": 295
    },
    {
      "epoch": 2.34375,
      "grad_norm": 0.6559345722198486,
      "learning_rate": 0.00016212319790301442,
      "loss": 0.5065,
      "step": 300
    },
    {
      "epoch": 2.3828125,
      "grad_norm": 0.776062548160553,
      "learning_rate": 0.00016146788990825688,
      "loss": 0.4548,
      "step": 305
    },
    {
      "epoch": 2.421875,
      "grad_norm": 0.5407435297966003,
      "learning_rate": 0.00016081258191349935,
      "loss": 0.4586,
      "step": 310
    },
    {
      "epoch": 2.4609375,
      "grad_norm": 0.7619644403457642,
      "learning_rate": 0.00016015727391874182,
      "loss": 0.6065,
      "step": 315
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6659480333328247,
      "learning_rate": 0.0001595019659239843,
      "loss": 0.4892,
      "step": 320
    },
    {
      "epoch": 2.5390625,
      "grad_norm": 0.586632490158081,
      "learning_rate": 0.00015884665792922676,
      "loss": 0.5094,
      "step": 325
    },
    {
      "epoch": 2.578125,
      "grad_norm": 0.6501973867416382,
      "learning_rate": 0.0001581913499344692,
      "loss": 0.5017,
      "step": 330
    },
    {
      "epoch": 2.6171875,
      "grad_norm": 0.5939526557922363,
      "learning_rate": 0.00015753604193971166,
      "loss": 0.357,
      "step": 335
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.6541431546211243,
      "learning_rate": 0.00015688073394495413,
      "loss": 0.5687,
      "step": 340
    },
    {
      "epoch": 2.6953125,
      "grad_norm": 0.7392444014549255,
      "learning_rate": 0.0001562254259501966,
      "loss": 0.5922,
      "step": 345
    },
    {
      "epoch": 2.734375,
      "grad_norm": 0.7246791124343872,
      "learning_rate": 0.00015557011795543907,
      "loss": 0.388,
      "step": 350
    },
    {
      "epoch": 2.7734375,
      "grad_norm": 1.0469605922698975,
      "learning_rate": 0.00015491480996068153,
      "loss": 0.4028,
      "step": 355
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.7362831830978394,
      "learning_rate": 0.000154259501965924,
      "loss": 0.6255,
      "step": 360
    },
    {
      "epoch": 2.8515625,
      "grad_norm": 0.6058784127235413,
      "learning_rate": 0.00015360419397116644,
      "loss": 0.5189,
      "step": 365
    },
    {
      "epoch": 2.890625,
      "grad_norm": 0.6939958333969116,
      "learning_rate": 0.00015294888597640894,
      "loss": 0.501,
      "step": 370
    },
    {
      "epoch": 2.9296875,
      "grad_norm": 0.8468016982078552,
      "learning_rate": 0.00015229357798165138,
      "loss": 0.5747,
      "step": 375
    },
    {
      "epoch": 2.96875,
      "grad_norm": 0.6065675616264343,
      "learning_rate": 0.00015163826998689384,
      "loss": 0.3813,
      "step": 380
    },
    {
      "epoch": 3.0078125,
      "grad_norm": 0.5093637108802795,
      "learning_rate": 0.0001509829619921363,
      "loss": 0.5125,
      "step": 385
    },
    {
      "epoch": 3.046875,
      "grad_norm": 0.7048936486244202,
      "learning_rate": 0.00015032765399737878,
      "loss": 0.3766,
      "step": 390
    },
    {
      "epoch": 3.0859375,
      "grad_norm": 1.191715955734253,
      "learning_rate": 0.00014967234600262122,
      "loss": 0.4224,
      "step": 395
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.6624323129653931,
      "learning_rate": 0.00014901703800786372,
      "loss": 0.4212,
      "step": 400
    },
    {
      "epoch": 3.1640625,
      "grad_norm": 1.3422083854675293,
      "learning_rate": 0.00014836173001310616,
      "loss": 0.3319,
      "step": 405
    },
    {
      "epoch": 3.203125,
      "grad_norm": 0.5813243985176086,
      "learning_rate": 0.00014770642201834862,
      "loss": 0.4468,
      "step": 410
    },
    {
      "epoch": 3.2421875,
      "grad_norm": 0.7296664118766785,
      "learning_rate": 0.0001470511140235911,
      "loss": 0.3234,
      "step": 415
    },
    {
      "epoch": 3.28125,
      "grad_norm": 0.7492959499359131,
      "learning_rate": 0.00014639580602883356,
      "loss": 0.3102,
      "step": 420
    },
    {
      "epoch": 3.3203125,
      "grad_norm": 0.994613528251648,
      "learning_rate": 0.000145740498034076,
      "loss": 0.4286,
      "step": 425
    },
    {
      "epoch": 3.359375,
      "grad_norm": 0.9514994025230408,
      "learning_rate": 0.0001450851900393185,
      "loss": 0.3267,
      "step": 430
    },
    {
      "epoch": 3.3984375,
      "grad_norm": 0.7083520293235779,
      "learning_rate": 0.00014442988204456096,
      "loss": 0.4743,
      "step": 435
    },
    {
      "epoch": 3.4375,
      "grad_norm": 0.6460224390029907,
      "learning_rate": 0.0001437745740498034,
      "loss": 0.3128,
      "step": 440
    },
    {
      "epoch": 3.4765625,
      "grad_norm": 0.7406665086746216,
      "learning_rate": 0.0001431192660550459,
      "loss": 0.3762,
      "step": 445
    },
    {
      "epoch": 3.515625,
      "grad_norm": 0.7346643805503845,
      "learning_rate": 0.00014246395806028834,
      "loss": 0.3573,
      "step": 450
    },
    {
      "epoch": 3.5546875,
      "grad_norm": 0.3775249123573303,
      "learning_rate": 0.0001418086500655308,
      "loss": 0.3488,
      "step": 455
    },
    {
      "epoch": 3.59375,
      "grad_norm": 0.9807206988334656,
      "learning_rate": 0.00014115334207077327,
      "loss": 0.2672,
      "step": 460
    },
    {
      "epoch": 3.6328125,
      "grad_norm": 0.5825705528259277,
      "learning_rate": 0.00014049803407601574,
      "loss": 0.2388,
      "step": 465
    },
    {
      "epoch": 3.671875,
      "grad_norm": 1.1724300384521484,
      "learning_rate": 0.00013984272608125818,
      "loss": 0.2998,
      "step": 470
    },
    {
      "epoch": 3.7109375,
      "grad_norm": 0.6543852090835571,
      "learning_rate": 0.00013918741808650068,
      "loss": 0.3704,
      "step": 475
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.6687126755714417,
      "learning_rate": 0.00013853211009174312,
      "loss": 0.3478,
      "step": 480
    },
    {
      "epoch": 3.7890625,
      "grad_norm": 0.8228131532669067,
      "learning_rate": 0.00013787680209698558,
      "loss": 0.3139,
      "step": 485
    },
    {
      "epoch": 3.828125,
      "grad_norm": 0.65690678358078,
      "learning_rate": 0.00013722149410222805,
      "loss": 0.4469,
      "step": 490
    },
    {
      "epoch": 3.8671875,
      "grad_norm": 0.7769365906715393,
      "learning_rate": 0.00013656618610747052,
      "loss": 0.4247,
      "step": 495
    },
    {
      "epoch": 3.90625,
      "grad_norm": 0.7008833289146423,
      "learning_rate": 0.000135910878112713,
      "loss": 0.4327,
      "step": 500
    },
    {
      "epoch": 3.9453125,
      "grad_norm": 0.6874434947967529,
      "learning_rate": 0.00013525557011795545,
      "loss": 0.5225,
      "step": 505
    },
    {
      "epoch": 3.984375,
      "grad_norm": 0.4368758499622345,
      "learning_rate": 0.00013460026212319792,
      "loss": 0.3697,
      "step": 510
    },
    {
      "epoch": 4.0234375,
      "grad_norm": 1.0020313262939453,
      "learning_rate": 0.00013394495412844036,
      "loss": 0.2611,
      "step": 515
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.8647730946540833,
      "learning_rate": 0.00013328964613368286,
      "loss": 0.2297,
      "step": 520
    },
    {
      "epoch": 4.1015625,
      "grad_norm": 1.0684905052185059,
      "learning_rate": 0.0001326343381389253,
      "loss": 0.2594,
      "step": 525
    },
    {
      "epoch": 4.140625,
      "grad_norm": 0.6783558130264282,
      "learning_rate": 0.00013197903014416777,
      "loss": 0.2302,
      "step": 530
    },
    {
      "epoch": 4.1796875,
      "grad_norm": 0.7600467205047607,
      "learning_rate": 0.00013132372214941023,
      "loss": 0.1469,
      "step": 535
    },
    {
      "epoch": 4.21875,
      "grad_norm": 0.9370886087417603,
      "learning_rate": 0.0001306684141546527,
      "loss": 0.3683,
      "step": 540
    },
    {
      "epoch": 4.2578125,
      "grad_norm": 0.6307783722877502,
      "learning_rate": 0.00013001310615989514,
      "loss": 0.1998,
      "step": 545
    },
    {
      "epoch": 4.296875,
      "grad_norm": 0.9554206728935242,
      "learning_rate": 0.0001293577981651376,
      "loss": 0.3081,
      "step": 550
    },
    {
      "epoch": 4.3359375,
      "grad_norm": 0.8178610801696777,
      "learning_rate": 0.00012870249017038008,
      "loss": 0.2767,
      "step": 555
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.6448714137077332,
      "learning_rate": 0.00012804718217562254,
      "loss": 0.2248,
      "step": 560
    },
    {
      "epoch": 4.4140625,
      "grad_norm": 0.9795539379119873,
      "learning_rate": 0.000127391874180865,
      "loss": 0.3152,
      "step": 565
    },
    {
      "epoch": 4.453125,
      "grad_norm": 0.7778314352035522,
      "learning_rate": 0.00012673656618610748,
      "loss": 0.3028,
      "step": 570
    },
    {
      "epoch": 4.4921875,
      "grad_norm": 0.9457613825798035,
      "learning_rate": 0.00012608125819134995,
      "loss": 0.309,
      "step": 575
    },
    {
      "epoch": 4.53125,
      "grad_norm": 0.7530558705329895,
      "learning_rate": 0.0001254259501965924,
      "loss": 0.3125,
      "step": 580
    },
    {
      "epoch": 4.5703125,
      "grad_norm": 0.7017265558242798,
      "learning_rate": 0.00012477064220183488,
      "loss": 0.273,
      "step": 585
    },
    {
      "epoch": 4.609375,
      "grad_norm": 0.8178383708000183,
      "learning_rate": 0.00012411533420707732,
      "loss": 0.3205,
      "step": 590
    },
    {
      "epoch": 4.6484375,
      "grad_norm": 1.5198026895523071,
      "learning_rate": 0.0001234600262123198,
      "loss": 0.3607,
      "step": 595
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.8270261883735657,
      "learning_rate": 0.00012280471821756226,
      "loss": 0.3297,
      "step": 600
    },
    {
      "epoch": 4.7265625,
      "grad_norm": 0.8817920088768005,
      "learning_rate": 0.00012214941022280473,
      "loss": 0.1601,
      "step": 605
    },
    {
      "epoch": 4.765625,
      "grad_norm": 0.9366243481636047,
      "learning_rate": 0.00012149410222804718,
      "loss": 0.1668,
      "step": 610
    },
    {
      "epoch": 4.8046875,
      "grad_norm": 0.706917405128479,
      "learning_rate": 0.00012083879423328965,
      "loss": 0.2827,
      "step": 615
    },
    {
      "epoch": 4.84375,
      "grad_norm": 0.8291599154472351,
      "learning_rate": 0.0001201834862385321,
      "loss": 0.2726,
      "step": 620
    },
    {
      "epoch": 4.8828125,
      "grad_norm": 0.6848894357681274,
      "learning_rate": 0.00011952817824377458,
      "loss": 0.1808,
      "step": 625
    },
    {
      "epoch": 4.921875,
      "grad_norm": 0.9057679176330566,
      "learning_rate": 0.00011887287024901705,
      "loss": 0.3169,
      "step": 630
    },
    {
      "epoch": 4.9609375,
      "grad_norm": 0.570704460144043,
      "learning_rate": 0.0001182175622542595,
      "loss": 0.3689,
      "step": 635
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8146092295646667,
      "learning_rate": 0.00011756225425950199,
      "loss": 0.3264,
      "step": 640
    },
    {
      "epoch": 5.0390625,
      "grad_norm": 0.5888718366622925,
      "learning_rate": 0.00011690694626474443,
      "loss": 0.167,
      "step": 645
    },
    {
      "epoch": 5.078125,
      "grad_norm": 0.9465442299842834,
      "learning_rate": 0.00011625163826998691,
      "loss": 0.1873,
      "step": 650
    },
    {
      "epoch": 5.1171875,
      "grad_norm": 0.5365155339241028,
      "learning_rate": 0.00011559633027522936,
      "loss": 0.1689,
      "step": 655
    },
    {
      "epoch": 5.15625,
      "grad_norm": 0.9071202278137207,
      "learning_rate": 0.00011494102228047183,
      "loss": 0.1666,
      "step": 660
    },
    {
      "epoch": 5.1953125,
      "grad_norm": 0.7092397212982178,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.1557,
      "step": 665
    },
    {
      "epoch": 5.234375,
      "grad_norm": 0.7074161767959595,
      "learning_rate": 0.00011363040629095676,
      "loss": 0.2316,
      "step": 670
    },
    {
      "epoch": 5.2734375,
      "grad_norm": 0.9464021325111389,
      "learning_rate": 0.0001129750982961992,
      "loss": 0.2005,
      "step": 675
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.6366726160049438,
      "learning_rate": 0.00011231979030144169,
      "loss": 0.1569,
      "step": 680
    },
    {
      "epoch": 5.3515625,
      "grad_norm": 0.6061714291572571,
      "learning_rate": 0.00011166448230668414,
      "loss": 0.2708,
      "step": 685
    },
    {
      "epoch": 5.390625,
      "grad_norm": 0.8460837602615356,
      "learning_rate": 0.00011100917431192661,
      "loss": 0.1406,
      "step": 690
    },
    {
      "epoch": 5.4296875,
      "grad_norm": 0.7344151139259338,
      "learning_rate": 0.00011035386631716909,
      "loss": 0.1954,
      "step": 695
    },
    {
      "epoch": 5.46875,
      "grad_norm": 1.1099109649658203,
      "learning_rate": 0.00010969855832241154,
      "loss": 0.1321,
      "step": 700
    },
    {
      "epoch": 5.5078125,
      "grad_norm": 0.8708857297897339,
      "learning_rate": 0.00010904325032765401,
      "loss": 0.2918,
      "step": 705
    },
    {
      "epoch": 5.546875,
      "grad_norm": 0.8755677938461304,
      "learning_rate": 0.00010838794233289647,
      "loss": 0.2321,
      "step": 710
    },
    {
      "epoch": 5.5859375,
      "grad_norm": 0.7879914045333862,
      "learning_rate": 0.00010773263433813893,
      "loss": 0.2525,
      "step": 715
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.8013678193092346,
      "learning_rate": 0.00010707732634338139,
      "loss": 0.1955,
      "step": 720
    },
    {
      "epoch": 5.6640625,
      "grad_norm": 0.6184900999069214,
      "learning_rate": 0.00010642201834862387,
      "loss": 0.1356,
      "step": 725
    },
    {
      "epoch": 5.703125,
      "grad_norm": 1.0018306970596313,
      "learning_rate": 0.00010576671035386632,
      "loss": 0.2501,
      "step": 730
    },
    {
      "epoch": 5.7421875,
      "grad_norm": 0.9539072513580322,
      "learning_rate": 0.00010511140235910879,
      "loss": 0.3376,
      "step": 735
    },
    {
      "epoch": 5.78125,
      "grad_norm": 0.7880743741989136,
      "learning_rate": 0.00010445609436435124,
      "loss": 0.176,
      "step": 740
    },
    {
      "epoch": 5.8203125,
      "grad_norm": 0.7900522351264954,
      "learning_rate": 0.00010380078636959371,
      "loss": 0.2542,
      "step": 745
    },
    {
      "epoch": 5.859375,
      "grad_norm": 0.7261202931404114,
      "learning_rate": 0.00010314547837483617,
      "loss": 0.1128,
      "step": 750
    },
    {
      "epoch": 5.8984375,
      "grad_norm": 0.8230142593383789,
      "learning_rate": 0.00010249017038007865,
      "loss": 0.2231,
      "step": 755
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.3808448612689972,
      "learning_rate": 0.0001018348623853211,
      "loss": 0.151,
      "step": 760
    },
    {
      "epoch": 5.9765625,
      "grad_norm": 0.4841325581073761,
      "learning_rate": 0.00010117955439056357,
      "loss": 0.1336,
      "step": 765
    },
    {
      "epoch": 6.015625,
      "grad_norm": 0.29059118032455444,
      "learning_rate": 0.00010052424639580605,
      "loss": 0.2318,
      "step": 770
    },
    {
      "epoch": 6.0546875,
      "grad_norm": 0.6378641724586487,
      "learning_rate": 9.986893840104849e-05,
      "loss": 0.1794,
      "step": 775
    },
    {
      "epoch": 6.09375,
      "grad_norm": 1.158392310142517,
      "learning_rate": 9.921363040629096e-05,
      "loss": 0.1575,
      "step": 780
    },
    {
      "epoch": 6.1328125,
      "grad_norm": 0.4778974652290344,
      "learning_rate": 9.855832241153343e-05,
      "loss": 0.1201,
      "step": 785
    },
    {
      "epoch": 6.171875,
      "grad_norm": 0.4302467107772827,
      "learning_rate": 9.790301441677588e-05,
      "loss": 0.0934,
      "step": 790
    },
    {
      "epoch": 6.2109375,
      "grad_norm": 0.5038356781005859,
      "learning_rate": 9.724770642201836e-05,
      "loss": 0.1497,
      "step": 795
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.6529866456985474,
      "learning_rate": 9.659239842726083e-05,
      "loss": 0.1439,
      "step": 800
    },
    {
      "epoch": 6.2890625,
      "grad_norm": 0.8134426474571228,
      "learning_rate": 9.593709043250328e-05,
      "loss": 0.1469,
      "step": 805
    },
    {
      "epoch": 6.328125,
      "grad_norm": 1.0741759538650513,
      "learning_rate": 9.528178243774575e-05,
      "loss": 0.0793,
      "step": 810
    },
    {
      "epoch": 6.3671875,
      "grad_norm": 0.37064865231513977,
      "learning_rate": 9.462647444298822e-05,
      "loss": 0.0891,
      "step": 815
    },
    {
      "epoch": 6.40625,
      "grad_norm": 0.6766513586044312,
      "learning_rate": 9.397116644823067e-05,
      "loss": 0.2092,
      "step": 820
    },
    {
      "epoch": 6.4453125,
      "grad_norm": 0.45673248171806335,
      "learning_rate": 9.331585845347314e-05,
      "loss": 0.1063,
      "step": 825
    },
    {
      "epoch": 6.484375,
      "grad_norm": 0.6083511710166931,
      "learning_rate": 9.266055045871561e-05,
      "loss": 0.0958,
      "step": 830
    },
    {
      "epoch": 6.5234375,
      "grad_norm": 0.954582691192627,
      "learning_rate": 9.200524246395806e-05,
      "loss": 0.1522,
      "step": 835
    },
    {
      "epoch": 6.5625,
      "grad_norm": 0.6275842785835266,
      "learning_rate": 9.134993446920053e-05,
      "loss": 0.1387,
      "step": 840
    },
    {
      "epoch": 6.6015625,
      "grad_norm": 0.826816976070404,
      "learning_rate": 9.069462647444298e-05,
      "loss": 0.1997,
      "step": 845
    },
    {
      "epoch": 6.640625,
      "grad_norm": 0.5855023264884949,
      "learning_rate": 9.003931847968545e-05,
      "loss": 0.1423,
      "step": 850
    },
    {
      "epoch": 6.6796875,
      "grad_norm": 0.37608104944229126,
      "learning_rate": 8.938401048492792e-05,
      "loss": 0.1563,
      "step": 855
    },
    {
      "epoch": 6.71875,
      "grad_norm": 1.1068248748779297,
      "learning_rate": 8.872870249017037e-05,
      "loss": 0.1066,
      "step": 860
    },
    {
      "epoch": 6.7578125,
      "grad_norm": 0.8714601397514343,
      "learning_rate": 8.807339449541285e-05,
      "loss": 0.1076,
      "step": 865
    },
    {
      "epoch": 6.796875,
      "grad_norm": 0.6995155215263367,
      "learning_rate": 8.741808650065532e-05,
      "loss": 0.0935,
      "step": 870
    },
    {
      "epoch": 6.8359375,
      "grad_norm": 0.895413875579834,
      "learning_rate": 8.676277850589778e-05,
      "loss": 0.1802,
      "step": 875
    },
    {
      "epoch": 6.875,
      "grad_norm": 0.8599961400032043,
      "learning_rate": 8.610747051114024e-05,
      "loss": 0.2149,
      "step": 880
    },
    {
      "epoch": 6.9140625,
      "grad_norm": 0.6649323105812073,
      "learning_rate": 8.545216251638271e-05,
      "loss": 0.1272,
      "step": 885
    },
    {
      "epoch": 6.953125,
      "grad_norm": 0.6272252798080444,
      "learning_rate": 8.479685452162516e-05,
      "loss": 0.1269,
      "step": 890
    },
    {
      "epoch": 6.9921875,
      "grad_norm": 0.837714672088623,
      "learning_rate": 8.414154652686763e-05,
      "loss": 0.1932,
      "step": 895
    },
    {
      "epoch": 7.03125,
      "grad_norm": 0.7026847004890442,
      "learning_rate": 8.34862385321101e-05,
      "loss": 0.1475,
      "step": 900
    },
    {
      "epoch": 7.0703125,
      "grad_norm": 0.24409687519073486,
      "learning_rate": 8.283093053735255e-05,
      "loss": 0.0635,
      "step": 905
    },
    {
      "epoch": 7.109375,
      "grad_norm": 0.26595592498779297,
      "learning_rate": 8.217562254259502e-05,
      "loss": 0.0448,
      "step": 910
    },
    {
      "epoch": 7.1484375,
      "grad_norm": 0.6503292322158813,
      "learning_rate": 8.152031454783749e-05,
      "loss": 0.0652,
      "step": 915
    },
    {
      "epoch": 7.1875,
      "grad_norm": 1.0240068435668945,
      "learning_rate": 8.086500655307994e-05,
      "loss": 0.077,
      "step": 920
    },
    {
      "epoch": 7.2265625,
      "grad_norm": 0.36204642057418823,
      "learning_rate": 8.020969855832241e-05,
      "loss": 0.0827,
      "step": 925
    },
    {
      "epoch": 7.265625,
      "grad_norm": 0.8305175304412842,
      "learning_rate": 7.955439056356488e-05,
      "loss": 0.1117,
      "step": 930
    },
    {
      "epoch": 7.3046875,
      "grad_norm": 0.31086069345474243,
      "learning_rate": 7.889908256880735e-05,
      "loss": 0.0912,
      "step": 935
    },
    {
      "epoch": 7.34375,
      "grad_norm": 0.5125362873077393,
      "learning_rate": 7.824377457404981e-05,
      "loss": 0.0819,
      "step": 940
    },
    {
      "epoch": 7.3828125,
      "grad_norm": 0.6713749766349792,
      "learning_rate": 7.758846657929227e-05,
      "loss": 0.1059,
      "step": 945
    },
    {
      "epoch": 7.421875,
      "grad_norm": 0.6156826615333557,
      "learning_rate": 7.693315858453474e-05,
      "loss": 0.1326,
      "step": 950
    },
    {
      "epoch": 7.4609375,
      "grad_norm": 0.7549245953559875,
      "learning_rate": 7.62778505897772e-05,
      "loss": 0.0854,
      "step": 955
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.8916281461715698,
      "learning_rate": 7.562254259501966e-05,
      "loss": 0.0898,
      "step": 960
    },
    {
      "epoch": 7.5390625,
      "grad_norm": 0.7383102178573608,
      "learning_rate": 7.496723460026212e-05,
      "loss": 0.0699,
      "step": 965
    },
    {
      "epoch": 7.578125,
      "grad_norm": 0.7137540578842163,
      "learning_rate": 7.431192660550459e-05,
      "loss": 0.0991,
      "step": 970
    },
    {
      "epoch": 7.6171875,
      "grad_norm": 0.3973597586154938,
      "learning_rate": 7.365661861074705e-05,
      "loss": 0.089,
      "step": 975
    },
    {
      "epoch": 7.65625,
      "grad_norm": 0.8355888724327087,
      "learning_rate": 7.300131061598951e-05,
      "loss": 0.0913,
      "step": 980
    },
    {
      "epoch": 7.6953125,
      "grad_norm": 0.5336706042289734,
      "learning_rate": 7.234600262123198e-05,
      "loss": 0.1516,
      "step": 985
    },
    {
      "epoch": 7.734375,
      "grad_norm": 0.8123258352279663,
      "learning_rate": 7.169069462647444e-05,
      "loss": 0.1144,
      "step": 990
    },
    {
      "epoch": 7.7734375,
      "grad_norm": 0.4773609936237335,
      "learning_rate": 7.103538663171692e-05,
      "loss": 0.0785,
      "step": 995
    },
    {
      "epoch": 7.8125,
      "grad_norm": 0.8169093132019043,
      "learning_rate": 7.038007863695938e-05,
      "loss": 0.1227,
      "step": 1000
    },
    {
      "epoch": 7.8515625,
      "grad_norm": 0.27247804403305054,
      "learning_rate": 6.972477064220184e-05,
      "loss": 0.086,
      "step": 1005
    },
    {
      "epoch": 7.890625,
      "grad_norm": 0.5810950398445129,
      "learning_rate": 6.90694626474443e-05,
      "loss": 0.0948,
      "step": 1010
    },
    {
      "epoch": 7.9296875,
      "grad_norm": 0.9459575414657593,
      "learning_rate": 6.841415465268677e-05,
      "loss": 0.0889,
      "step": 1015
    },
    {
      "epoch": 7.96875,
      "grad_norm": 0.127482607960701,
      "learning_rate": 6.775884665792923e-05,
      "loss": 0.0981,
      "step": 1020
    },
    {
      "epoch": 8.0078125,
      "grad_norm": 0.3091827630996704,
      "learning_rate": 6.71035386631717e-05,
      "loss": 0.1035,
      "step": 1025
    },
    {
      "epoch": 8.046875,
      "grad_norm": 0.4560360610485077,
      "learning_rate": 6.644823066841416e-05,
      "loss": 0.0834,
      "step": 1030
    },
    {
      "epoch": 8.0859375,
      "grad_norm": 0.6077558398246765,
      "learning_rate": 6.579292267365662e-05,
      "loss": 0.0711,
      "step": 1035
    },
    {
      "epoch": 8.125,
      "grad_norm": 0.7591610550880432,
      "learning_rate": 6.513761467889909e-05,
      "loss": 0.0877,
      "step": 1040
    },
    {
      "epoch": 8.1640625,
      "grad_norm": 0.38261088728904724,
      "learning_rate": 6.448230668414155e-05,
      "loss": 0.0422,
      "step": 1045
    },
    {
      "epoch": 8.203125,
      "grad_norm": 0.8100435137748718,
      "learning_rate": 6.382699868938401e-05,
      "loss": 0.0883,
      "step": 1050
    },
    {
      "epoch": 8.2421875,
      "grad_norm": 0.25259244441986084,
      "learning_rate": 6.317169069462647e-05,
      "loss": 0.0696,
      "step": 1055
    },
    {
      "epoch": 8.28125,
      "grad_norm": 0.7604616284370422,
      "learning_rate": 6.251638269986894e-05,
      "loss": 0.0761,
      "step": 1060
    },
    {
      "epoch": 8.3203125,
      "grad_norm": 0.395271897315979,
      "learning_rate": 6.186107470511141e-05,
      "loss": 0.0769,
      "step": 1065
    },
    {
      "epoch": 8.359375,
      "grad_norm": 0.6258074045181274,
      "learning_rate": 6.120576671035388e-05,
      "loss": 0.0543,
      "step": 1070
    },
    {
      "epoch": 8.3984375,
      "grad_norm": 0.13846486806869507,
      "learning_rate": 6.055045871559634e-05,
      "loss": 0.0526,
      "step": 1075
    },
    {
      "epoch": 8.4375,
      "grad_norm": 0.7763333916664124,
      "learning_rate": 5.98951507208388e-05,
      "loss": 0.0649,
      "step": 1080
    },
    {
      "epoch": 8.4765625,
      "grad_norm": 0.6369220614433289,
      "learning_rate": 5.923984272608126e-05,
      "loss": 0.043,
      "step": 1085
    },
    {
      "epoch": 8.515625,
      "grad_norm": 0.6248875856399536,
      "learning_rate": 5.858453473132373e-05,
      "loss": 0.0648,
      "step": 1090
    },
    {
      "epoch": 8.5546875,
      "grad_norm": 0.1796266883611679,
      "learning_rate": 5.792922673656619e-05,
      "loss": 0.0487,
      "step": 1095
    },
    {
      "epoch": 8.59375,
      "grad_norm": 0.7085462212562561,
      "learning_rate": 5.727391874180865e-05,
      "loss": 0.0745,
      "step": 1100
    },
    {
      "epoch": 8.6328125,
      "grad_norm": 0.27906715869903564,
      "learning_rate": 5.661861074705112e-05,
      "loss": 0.0382,
      "step": 1105
    },
    {
      "epoch": 8.671875,
      "grad_norm": 0.4836632013320923,
      "learning_rate": 5.596330275229358e-05,
      "loss": 0.0714,
      "step": 1110
    },
    {
      "epoch": 8.7109375,
      "grad_norm": 0.5871438384056091,
      "learning_rate": 5.530799475753604e-05,
      "loss": 0.0427,
      "step": 1115
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.3616584241390228,
      "learning_rate": 5.4652686762778507e-05,
      "loss": 0.0458,
      "step": 1120
    },
    {
      "epoch": 8.7890625,
      "grad_norm": 0.9580535292625427,
      "learning_rate": 5.399737876802097e-05,
      "loss": 0.0759,
      "step": 1125
    },
    {
      "epoch": 8.828125,
      "grad_norm": 0.6458576321601868,
      "learning_rate": 5.334207077326344e-05,
      "loss": 0.069,
      "step": 1130
    },
    {
      "epoch": 8.8671875,
      "grad_norm": 0.21817967295646667,
      "learning_rate": 5.26867627785059e-05,
      "loss": 0.0511,
      "step": 1135
    },
    {
      "epoch": 8.90625,
      "grad_norm": 0.6552639603614807,
      "learning_rate": 5.203145478374837e-05,
      "loss": 0.0707,
      "step": 1140
    },
    {
      "epoch": 8.9453125,
      "grad_norm": 0.5542663931846619,
      "learning_rate": 5.137614678899083e-05,
      "loss": 0.0811,
      "step": 1145
    },
    {
      "epoch": 8.984375,
      "grad_norm": 0.2486066222190857,
      "learning_rate": 5.072083879423329e-05,
      "loss": 0.0491,
      "step": 1150
    },
    {
      "epoch": 9.0234375,
      "grad_norm": 0.10281497240066528,
      "learning_rate": 5.006553079947576e-05,
      "loss": 0.0513,
      "step": 1155
    },
    {
      "epoch": 9.0625,
      "grad_norm": 0.5462967753410339,
      "learning_rate": 4.941022280471822e-05,
      "loss": 0.0444,
      "step": 1160
    },
    {
      "epoch": 9.1015625,
      "grad_norm": 0.32824379205703735,
      "learning_rate": 4.875491480996068e-05,
      "loss": 0.0419,
      "step": 1165
    },
    {
      "epoch": 9.140625,
      "grad_norm": 0.15365761518478394,
      "learning_rate": 4.809960681520315e-05,
      "loss": 0.0532,
      "step": 1170
    },
    {
      "epoch": 9.1796875,
      "grad_norm": 0.4261007308959961,
      "learning_rate": 4.744429882044561e-05,
      "loss": 0.0476,
      "step": 1175
    },
    {
      "epoch": 9.21875,
      "grad_norm": 0.5910694599151611,
      "learning_rate": 4.678899082568808e-05,
      "loss": 0.0437,
      "step": 1180
    },
    {
      "epoch": 9.2578125,
      "grad_norm": 0.30444568395614624,
      "learning_rate": 4.613368283093054e-05,
      "loss": 0.0485,
      "step": 1185
    },
    {
      "epoch": 9.296875,
      "grad_norm": 0.21978724002838135,
      "learning_rate": 4.5478374836173006e-05,
      "loss": 0.0422,
      "step": 1190
    },
    {
      "epoch": 9.3359375,
      "grad_norm": 0.754964292049408,
      "learning_rate": 4.482306684141547e-05,
      "loss": 0.0474,
      "step": 1195
    },
    {
      "epoch": 9.375,
      "grad_norm": 0.08515404164791107,
      "learning_rate": 4.416775884665793e-05,
      "loss": 0.0333,
      "step": 1200
    },
    {
      "epoch": 9.4140625,
      "grad_norm": 0.22733353078365326,
      "learning_rate": 4.3512450851900395e-05,
      "loss": 0.0408,
      "step": 1205
    },
    {
      "epoch": 9.453125,
      "grad_norm": 0.2974601089954376,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0493,
      "step": 1210
    },
    {
      "epoch": 9.4921875,
      "grad_norm": 0.7275934815406799,
      "learning_rate": 4.2201834862385324e-05,
      "loss": 0.0538,
      "step": 1215
    },
    {
      "epoch": 9.53125,
      "grad_norm": 0.4996713101863861,
      "learning_rate": 4.154652686762779e-05,
      "loss": 0.0383,
      "step": 1220
    },
    {
      "epoch": 9.5703125,
      "grad_norm": 0.4535525441169739,
      "learning_rate": 4.089121887287025e-05,
      "loss": 0.0654,
      "step": 1225
    },
    {
      "epoch": 9.609375,
      "grad_norm": 0.44223669171333313,
      "learning_rate": 4.023591087811271e-05,
      "loss": 0.0369,
      "step": 1230
    },
    {
      "epoch": 9.6484375,
      "grad_norm": 0.08399149775505066,
      "learning_rate": 3.958060288335518e-05,
      "loss": 0.0374,
      "step": 1235
    },
    {
      "epoch": 9.6875,
      "grad_norm": 0.4127291142940521,
      "learning_rate": 3.892529488859764e-05,
      "loss": 0.0379,
      "step": 1240
    },
    {
      "epoch": 9.7265625,
      "grad_norm": 0.21898190677165985,
      "learning_rate": 3.82699868938401e-05,
      "loss": 0.0323,
      "step": 1245
    },
    {
      "epoch": 9.765625,
      "grad_norm": 0.1628551185131073,
      "learning_rate": 3.761467889908257e-05,
      "loss": 0.0438,
      "step": 1250
    },
    {
      "epoch": 9.8046875,
      "grad_norm": 0.42952653765678406,
      "learning_rate": 3.695937090432504e-05,
      "loss": 0.0358,
      "step": 1255
    },
    {
      "epoch": 9.84375,
      "grad_norm": 0.23393145203590393,
      "learning_rate": 3.63040629095675e-05,
      "loss": 0.0424,
      "step": 1260
    },
    {
      "epoch": 9.8828125,
      "grad_norm": 0.3994542956352234,
      "learning_rate": 3.564875491480996e-05,
      "loss": 0.0416,
      "step": 1265
    },
    {
      "epoch": 9.921875,
      "grad_norm": 0.32643911242485046,
      "learning_rate": 3.499344692005243e-05,
      "loss": 0.0378,
      "step": 1270
    },
    {
      "epoch": 9.9609375,
      "grad_norm": 0.13228672742843628,
      "learning_rate": 3.433813892529489e-05,
      "loss": 0.0302,
      "step": 1275
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.5824608206748962,
      "learning_rate": 3.3682830930537356e-05,
      "loss": 0.0534,
      "step": 1280
    },
    {
      "epoch": 10.0390625,
      "grad_norm": 0.36108893156051636,
      "learning_rate": 3.302752293577982e-05,
      "loss": 0.0316,
      "step": 1285
    },
    {
      "epoch": 10.078125,
      "grad_norm": 0.21287214756011963,
      "learning_rate": 3.2372214941022284e-05,
      "loss": 0.0325,
      "step": 1290
    },
    {
      "epoch": 10.1171875,
      "grad_norm": 0.3202109932899475,
      "learning_rate": 3.1716906946264745e-05,
      "loss": 0.0299,
      "step": 1295
    },
    {
      "epoch": 10.15625,
      "grad_norm": 0.21489982306957245,
      "learning_rate": 3.1061598951507206e-05,
      "loss": 0.0272,
      "step": 1300
    },
    {
      "epoch": 10.1953125,
      "grad_norm": 0.1839279681444168,
      "learning_rate": 3.0406290956749674e-05,
      "loss": 0.0284,
      "step": 1305
    },
    {
      "epoch": 10.234375,
      "grad_norm": 0.36773788928985596,
      "learning_rate": 2.9750982961992135e-05,
      "loss": 0.0304,
      "step": 1310
    },
    {
      "epoch": 10.2734375,
      "grad_norm": 0.13723714649677277,
      "learning_rate": 2.9095674967234606e-05,
      "loss": 0.0367,
      "step": 1315
    },
    {
      "epoch": 10.3125,
      "grad_norm": 0.40129488706588745,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 0.04,
      "step": 1320
    },
    {
      "epoch": 10.3515625,
      "grad_norm": 0.08511374145746231,
      "learning_rate": 2.778505897771953e-05,
      "loss": 0.0246,
      "step": 1325
    },
    {
      "epoch": 10.390625,
      "grad_norm": 0.43909040093421936,
      "learning_rate": 2.7129750982961995e-05,
      "loss": 0.0318,
      "step": 1330
    },
    {
      "epoch": 10.4296875,
      "grad_norm": 0.11963684856891632,
      "learning_rate": 2.6474442988204456e-05,
      "loss": 0.0313,
      "step": 1335
    },
    {
      "epoch": 10.46875,
      "grad_norm": 0.07002709805965424,
      "learning_rate": 2.581913499344692e-05,
      "loss": 0.0234,
      "step": 1340
    },
    {
      "epoch": 10.5078125,
      "grad_norm": 0.5138005614280701,
      "learning_rate": 2.5163826998689384e-05,
      "loss": 0.0333,
      "step": 1345
    },
    {
      "epoch": 10.546875,
      "grad_norm": 0.5884416699409485,
      "learning_rate": 2.450851900393185e-05,
      "loss": 0.0334,
      "step": 1350
    },
    {
      "epoch": 10.5859375,
      "grad_norm": 0.2413049191236496,
      "learning_rate": 2.3853211009174313e-05,
      "loss": 0.029,
      "step": 1355
    },
    {
      "epoch": 10.625,
      "grad_norm": 0.07464331388473511,
      "learning_rate": 2.3197903014416777e-05,
      "loss": 0.0294,
      "step": 1360
    },
    {
      "epoch": 10.6640625,
      "grad_norm": 0.38783198595046997,
      "learning_rate": 2.254259501965924e-05,
      "loss": 0.0304,
      "step": 1365
    },
    {
      "epoch": 10.703125,
      "grad_norm": 0.46144524216651917,
      "learning_rate": 2.1887287024901702e-05,
      "loss": 0.0321,
      "step": 1370
    },
    {
      "epoch": 10.7421875,
      "grad_norm": 0.10308784991502762,
      "learning_rate": 2.123197903014417e-05,
      "loss": 0.0329,
      "step": 1375
    },
    {
      "epoch": 10.78125,
      "grad_norm": 0.33099478483200073,
      "learning_rate": 2.0576671035386634e-05,
      "loss": 0.0323,
      "step": 1380
    },
    {
      "epoch": 10.8203125,
      "grad_norm": 0.381788045167923,
      "learning_rate": 1.9921363040629095e-05,
      "loss": 0.0268,
      "step": 1385
    },
    {
      "epoch": 10.859375,
      "grad_norm": 0.08423493057489395,
      "learning_rate": 1.9266055045871563e-05,
      "loss": 0.0304,
      "step": 1390
    },
    {
      "epoch": 10.8984375,
      "grad_norm": 0.5857261419296265,
      "learning_rate": 1.8610747051114023e-05,
      "loss": 0.0327,
      "step": 1395
    },
    {
      "epoch": 10.9375,
      "grad_norm": 0.12234900146722794,
      "learning_rate": 1.7955439056356488e-05,
      "loss": 0.0298,
      "step": 1400
    },
    {
      "epoch": 10.9765625,
      "grad_norm": 0.4933612048625946,
      "learning_rate": 1.7300131061598955e-05,
      "loss": 0.0329,
      "step": 1405
    },
    {
      "epoch": 11.015625,
      "grad_norm": 0.21089200675487518,
      "learning_rate": 1.6644823066841416e-05,
      "loss": 0.0274,
      "step": 1410
    },
    {
      "epoch": 11.0546875,
      "grad_norm": 0.10511160641908646,
      "learning_rate": 1.598951507208388e-05,
      "loss": 0.0254,
      "step": 1415
    },
    {
      "epoch": 11.09375,
      "grad_norm": 0.10039519518613815,
      "learning_rate": 1.5334207077326345e-05,
      "loss": 0.0211,
      "step": 1420
    },
    {
      "epoch": 11.1328125,
      "grad_norm": 0.17804010212421417,
      "learning_rate": 1.4678899082568809e-05,
      "loss": 0.0239,
      "step": 1425
    },
    {
      "epoch": 11.171875,
      "grad_norm": 0.21066808700561523,
      "learning_rate": 1.4023591087811271e-05,
      "loss": 0.0302,
      "step": 1430
    },
    {
      "epoch": 11.2109375,
      "grad_norm": 0.14561991393566132,
      "learning_rate": 1.3368283093053736e-05,
      "loss": 0.0204,
      "step": 1435
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.0932135209441185,
      "learning_rate": 1.2712975098296202e-05,
      "loss": 0.0284,
      "step": 1440
    },
    {
      "epoch": 11.2890625,
      "grad_norm": 0.2561885416507721,
      "learning_rate": 1.2057667103538664e-05,
      "loss": 0.0246,
      "step": 1445
    },
    {
      "epoch": 11.328125,
      "grad_norm": 0.20448650419712067,
      "learning_rate": 1.1402359108781127e-05,
      "loss": 0.0282,
      "step": 1450
    },
    {
      "epoch": 11.3671875,
      "grad_norm": 0.11967150866985321,
      "learning_rate": 1.0747051114023591e-05,
      "loss": 0.0271,
      "step": 1455
    },
    {
      "epoch": 11.40625,
      "grad_norm": 0.12555184960365295,
      "learning_rate": 1.0091743119266055e-05,
      "loss": 0.0248,
      "step": 1460
    },
    {
      "epoch": 11.4453125,
      "grad_norm": 0.2812488079071045,
      "learning_rate": 9.43643512450852e-06,
      "loss": 0.0312,
      "step": 1465
    },
    {
      "epoch": 11.484375,
      "grad_norm": 0.06791621446609497,
      "learning_rate": 8.781127129750984e-06,
      "loss": 0.0246,
      "step": 1470
    },
    {
      "epoch": 11.5234375,
      "grad_norm": 0.08099279552698135,
      "learning_rate": 8.125819134993446e-06,
      "loss": 0.0243,
      "step": 1475
    },
    {
      "epoch": 11.5625,
      "grad_norm": 0.1286236196756363,
      "learning_rate": 7.4705111402359114e-06,
      "loss": 0.0215,
      "step": 1480
    },
    {
      "epoch": 11.6015625,
      "grad_norm": 0.2546003460884094,
      "learning_rate": 6.815203145478376e-06,
      "loss": 0.0248,
      "step": 1485
    },
    {
      "epoch": 11.640625,
      "grad_norm": 0.18944767117500305,
      "learning_rate": 6.159895150720839e-06,
      "loss": 0.024,
      "step": 1490
    },
    {
      "epoch": 11.6796875,
      "grad_norm": 0.20657788217067719,
      "learning_rate": 5.504587155963303e-06,
      "loss": 0.0206,
      "step": 1495
    },
    {
      "epoch": 11.71875,
      "grad_norm": 0.2377331256866455,
      "learning_rate": 4.849279161205767e-06,
      "loss": 0.0254,
      "step": 1500
    },
    {
      "epoch": 11.7578125,
      "grad_norm": 0.10646895319223404,
      "learning_rate": 4.193971166448231e-06,
      "loss": 0.0214,
      "step": 1505
    },
    {
      "epoch": 11.796875,
      "grad_norm": 0.3463532328605652,
      "learning_rate": 3.538663171690695e-06,
      "loss": 0.0252,
      "step": 1510
    },
    {
      "epoch": 11.8359375,
      "grad_norm": 0.17272751033306122,
      "learning_rate": 2.8833551769331587e-06,
      "loss": 0.0316,
      "step": 1515
    },
    {
      "epoch": 11.875,
      "grad_norm": 0.3141430914402008,
      "learning_rate": 2.2280471821756225e-06,
      "loss": 0.0251,
      "step": 1520
    },
    {
      "epoch": 11.9140625,
      "grad_norm": 0.28511035442352295,
      "learning_rate": 1.5727391874180865e-06,
      "loss": 0.0267,
      "step": 1525
    },
    {
      "epoch": 11.953125,
      "grad_norm": 0.10313425958156586,
      "learning_rate": 9.174311926605506e-07,
      "loss": 0.0196,
      "step": 1530
    },
    {
      "epoch": 11.9921875,
      "grad_norm": 0.28020963072776794,
      "learning_rate": 2.6212319790301444e-07,
      "loss": 0.024,
      "step": 1535
    },
    {
      "epoch": 12.0,
      "step": 1536,
      "total_flos": 8.765889328981094e+16,
      "train_loss": 0.2845887634175597,
      "train_runtime": 2956.9183,
      "train_samples_per_second": 4.152,
      "train_steps_per_second": 0.519
    }
  ],
  "logging_steps": 5,
  "max_steps": 1536,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.765889328981094e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}