{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.25328330206379,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0075046904315197,
      "grad_norm": 0.24946506321430206,
      "learning_rate": 2.02e-06,
      "loss": 1.6195,
      "step": 1
    },
    {
      "epoch": 0.0150093808630394,
      "grad_norm": 0.29706066846847534,
      "learning_rate": 4.04e-06,
      "loss": 1.7606,
      "step": 2
    },
    {
      "epoch": 0.0225140712945591,
      "grad_norm": 0.34257644414901733,
      "learning_rate": 6.06e-06,
      "loss": 1.6976,
      "step": 3
    },
    {
      "epoch": 0.0300187617260788,
      "grad_norm": 0.29441866278648376,
      "learning_rate": 8.08e-06,
      "loss": 1.6113,
      "step": 4
    },
    {
      "epoch": 0.0375234521575985,
      "grad_norm": 0.37839648127555847,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 1.7146,
      "step": 5
    },
    {
      "epoch": 0.0450281425891182,
      "grad_norm": 0.26730063557624817,
      "learning_rate": 1.212e-05,
      "loss": 1.6769,
      "step": 6
    },
    {
      "epoch": 0.0525328330206379,
      "grad_norm": 0.3574599027633667,
      "learning_rate": 1.4140000000000002e-05,
      "loss": 1.7317,
      "step": 7
    },
    {
      "epoch": 0.0600375234521576,
      "grad_norm": 0.3100043833255768,
      "learning_rate": 1.616e-05,
      "loss": 1.636,
      "step": 8
    },
    {
      "epoch": 0.0675422138836773,
      "grad_norm": 0.2719382047653198,
      "learning_rate": 1.818e-05,
      "loss": 1.6261,
      "step": 9
    },
    {
      "epoch": 0.075046904315197,
      "grad_norm": 0.35602274537086487,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 1.6088,
      "step": 10
    },
    {
      "epoch": 0.0825515947467167,
      "grad_norm": 0.2575864791870117,
      "learning_rate": 2.222e-05,
      "loss": 1.7308,
      "step": 11
    },
    {
      "epoch": 0.0900562851782364,
      "grad_norm": 0.2723293602466583,
      "learning_rate": 2.424e-05,
      "loss": 1.5698,
      "step": 12
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 0.24963687360286713,
      "learning_rate": 2.6260000000000003e-05,
      "loss": 1.6378,
      "step": 13
    },
    {
      "epoch": 0.1050656660412758,
      "grad_norm": 0.2475719004869461,
      "learning_rate": 2.8280000000000004e-05,
      "loss": 1.6805,
      "step": 14
    },
    {
      "epoch": 0.1125703564727955,
      "grad_norm": 0.2292100340127945,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 1.6527,
      "step": 15
    },
    {
      "epoch": 0.1200750469043152,
      "grad_norm": 0.22770565748214722,
      "learning_rate": 3.232e-05,
      "loss": 1.6284,
      "step": 16
    },
    {
      "epoch": 0.1275797373358349,
      "grad_norm": 0.24064026772975922,
      "learning_rate": 3.434e-05,
      "loss": 1.6247,
      "step": 17
    },
    {
      "epoch": 0.1350844277673546,
      "grad_norm": 0.2927341163158417,
      "learning_rate": 3.636e-05,
      "loss": 1.6655,
      "step": 18
    },
    {
      "epoch": 0.1425891181988743,
      "grad_norm": 0.2897813618183136,
      "learning_rate": 3.838e-05,
      "loss": 1.6669,
      "step": 19
    },
    {
      "epoch": 0.150093808630394,
      "grad_norm": 0.3220812976360321,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 1.5341,
      "step": 20
    },
    {
      "epoch": 0.1575984990619137,
      "grad_norm": 0.29551008343696594,
      "learning_rate": 4.242e-05,
      "loss": 1.5828,
      "step": 21
    },
    {
      "epoch": 0.1651031894934334,
      "grad_norm": 0.26846417784690857,
      "learning_rate": 4.444e-05,
      "loss": 1.573,
      "step": 22
    },
    {
      "epoch": 0.1726078799249531,
      "grad_norm": 0.20109350979328156,
      "learning_rate": 4.6460000000000006e-05,
      "loss": 1.6138,
      "step": 23
    },
    {
      "epoch": 0.1801125703564728,
      "grad_norm": 0.24828770756721497,
      "learning_rate": 4.848e-05,
      "loss": 1.52,
      "step": 24
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 0.2683457136154175,
      "learning_rate": 5.05e-05,
      "loss": 1.5518,
      "step": 25
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.383939653635025,
      "learning_rate": 5.2520000000000005e-05,
      "loss": 1.45,
      "step": 26
    },
    {
      "epoch": 0.2026266416510319,
      "grad_norm": 0.2534943222999573,
      "learning_rate": 5.454e-05,
      "loss": 1.5009,
      "step": 27
    },
    {
      "epoch": 0.2101313320825516,
      "grad_norm": 0.2871793508529663,
      "learning_rate": 5.656000000000001e-05,
      "loss": 1.4322,
      "step": 28
    },
    {
      "epoch": 0.2176360225140713,
      "grad_norm": 0.20925107598304749,
      "learning_rate": 5.858e-05,
      "loss": 1.5116,
      "step": 29
    },
    {
      "epoch": 0.225140712945591,
      "grad_norm": 0.2413536161184311,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 1.4501,
      "step": 30
    },
    {
      "epoch": 0.2326454033771107,
      "grad_norm": 0.2639910876750946,
      "learning_rate": 6.262000000000001e-05,
      "loss": 1.3605,
      "step": 31
    },
    {
      "epoch": 0.2401500938086304,
      "grad_norm": 0.2476966679096222,
      "learning_rate": 6.464e-05,
      "loss": 1.322,
      "step": 32
    },
    {
      "epoch": 0.24765478424015008,
      "grad_norm": 0.22135838866233826,
      "learning_rate": 6.666e-05,
      "loss": 1.3593,
      "step": 33
    },
    {
      "epoch": 0.2551594746716698,
      "grad_norm": 0.1916331648826599,
      "learning_rate": 6.868e-05,
      "loss": 1.4882,
      "step": 34
    },
    {
      "epoch": 0.2626641651031895,
      "grad_norm": 0.2084406614303589,
      "learning_rate": 7.07e-05,
      "loss": 1.421,
      "step": 35
    },
    {
      "epoch": 0.2701688555347092,
      "grad_norm": 0.1996811330318451,
      "learning_rate": 7.272e-05,
      "loss": 1.3375,
      "step": 36
    },
    {
      "epoch": 0.2776735459662289,
      "grad_norm": 0.20538535714149475,
      "learning_rate": 7.474e-05,
      "loss": 1.3453,
      "step": 37
    },
    {
      "epoch": 0.2851782363977486,
      "grad_norm": 0.2092738300561905,
      "learning_rate": 7.676e-05,
      "loss": 1.3675,
      "step": 38
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 0.17955994606018066,
      "learning_rate": 7.878e-05,
      "loss": 1.365,
      "step": 39
    },
    {
      "epoch": 0.300187617260788,
      "grad_norm": 0.16386628150939941,
      "learning_rate": 8.080000000000001e-05,
      "loss": 1.3524,
      "step": 40
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.1718520224094391,
      "learning_rate": 8.282e-05,
      "loss": 1.1911,
      "step": 41
    },
    {
      "epoch": 0.3151969981238274,
      "grad_norm": 0.1762257218360901,
      "learning_rate": 8.484e-05,
      "loss": 1.3526,
      "step": 42
    },
    {
      "epoch": 0.3227016885553471,
      "grad_norm": 0.16516637802124023,
      "learning_rate": 8.686e-05,
      "loss": 1.4977,
      "step": 43
    },
    {
      "epoch": 0.3302063789868668,
      "grad_norm": 0.1738591492176056,
      "learning_rate": 8.888e-05,
      "loss": 1.348,
      "step": 44
    },
    {
      "epoch": 0.33771106941838647,
      "grad_norm": 0.15606296062469482,
      "learning_rate": 9.09e-05,
      "loss": 1.2181,
      "step": 45
    },
    {
      "epoch": 0.3452157598499062,
      "grad_norm": 0.14739865064620972,
      "learning_rate": 9.292000000000001e-05,
      "loss": 1.5,
      "step": 46
    },
    {
      "epoch": 0.3527204502814259,
      "grad_norm": 0.14701372385025024,
      "learning_rate": 9.494e-05,
      "loss": 1.2768,
      "step": 47
    },
    {
      "epoch": 0.3602251407129456,
      "grad_norm": 0.16340306401252747,
      "learning_rate": 9.696e-05,
      "loss": 1.3671,
      "step": 48
    },
    {
      "epoch": 0.3677298311444653,
      "grad_norm": 0.15772990882396698,
      "learning_rate": 9.898e-05,
      "loss": 1.3383,
      "step": 49
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 0.15352784097194672,
      "learning_rate": 0.000101,
      "loss": 1.3189,
      "step": 50
    },
    {
      "epoch": 0.3827392120075047,
      "grad_norm": 0.14889372885227203,
      "learning_rate": 0.00010302,
      "loss": 1.2749,
      "step": 51
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 0.150027796626091,
      "learning_rate": 0.00010504000000000001,
      "loss": 1.3128,
      "step": 52
    },
    {
      "epoch": 0.3977485928705441,
      "grad_norm": 0.14402741193771362,
      "learning_rate": 0.00010706000000000001,
      "loss": 1.2489,
      "step": 53
    },
    {
      "epoch": 0.4052532833020638,
      "grad_norm": 0.15602846443653107,
      "learning_rate": 0.00010908,
      "loss": 1.3278,
      "step": 54
    },
    {
      "epoch": 0.41275797373358347,
      "grad_norm": 0.168035089969635,
      "learning_rate": 0.00011110000000000002,
      "loss": 1.3057,
      "step": 55
    },
    {
      "epoch": 0.4202626641651032,
      "grad_norm": 0.14109627902507782,
      "learning_rate": 0.00011312000000000001,
      "loss": 1.3653,
      "step": 56
    },
    {
      "epoch": 0.4277673545966229,
      "grad_norm": 0.16596925258636475,
      "learning_rate": 0.00011514,
      "loss": 1.3145,
      "step": 57
    },
    {
      "epoch": 0.4352720450281426,
      "grad_norm": 0.1616641730070114,
      "learning_rate": 0.00011716,
      "loss": 1.2194,
      "step": 58
    },
    {
      "epoch": 0.4427767354596623,
      "grad_norm": 0.16131196916103363,
      "learning_rate": 0.00011918,
      "loss": 1.3411,
      "step": 59
    },
    {
      "epoch": 0.450281425891182,
      "grad_norm": 0.18199914693832397,
      "learning_rate": 0.00012119999999999999,
      "loss": 1.352,
      "step": 60
    },
    {
      "epoch": 0.45778611632270166,
      "grad_norm": 0.14444635808467865,
      "learning_rate": 0.00012322,
      "loss": 1.3969,
      "step": 61
    },
    {
      "epoch": 0.4652908067542214,
      "grad_norm": 0.13516508042812347,
      "learning_rate": 0.00012524000000000001,
      "loss": 1.257,
      "step": 62
    },
    {
      "epoch": 0.4727954971857411,
      "grad_norm": 0.17646552622318268,
      "learning_rate": 0.00012726,
      "loss": 1.2882,
      "step": 63
    },
    {
      "epoch": 0.4803001876172608,
      "grad_norm": 0.15815891325473785,
      "learning_rate": 0.00012928,
      "loss": 1.3732,
      "step": 64
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.1589154303073883,
      "learning_rate": 0.00013130000000000002,
      "loss": 1.4189,
      "step": 65
    },
    {
      "epoch": 0.49530956848030017,
      "grad_norm": 0.15148110687732697,
      "learning_rate": 0.00013332,
      "loss": 1.3166,
      "step": 66
    },
    {
      "epoch": 0.5028142589118199,
      "grad_norm": 0.16850154101848602,
      "learning_rate": 0.00013534000000000002,
      "loss": 1.3046,
      "step": 67
    },
    {
      "epoch": 0.5103189493433395,
      "grad_norm": 0.14561696350574493,
      "learning_rate": 0.00013736,
      "loss": 1.2719,
      "step": 68
    },
    {
      "epoch": 0.5178236397748592,
      "grad_norm": 0.16271142661571503,
      "learning_rate": 0.00013937999999999998,
      "loss": 1.23,
      "step": 69
    },
    {
      "epoch": 0.525328330206379,
      "grad_norm": 0.17095573246479034,
      "learning_rate": 0.0001414,
      "loss": 1.2615,
      "step": 70
    },
    {
      "epoch": 0.5328330206378987,
      "grad_norm": 0.14449560642242432,
      "learning_rate": 0.00014342,
      "loss": 1.4055,
      "step": 71
    },
    {
      "epoch": 0.5403377110694184,
      "grad_norm": 0.15791203081607819,
      "learning_rate": 0.00014544,
      "loss": 1.2517,
      "step": 72
    },
    {
      "epoch": 0.5478424015009381,
      "grad_norm": 0.16690890491008759,
      "learning_rate": 0.00014746,
      "loss": 1.2306,
      "step": 73
    },
    {
      "epoch": 0.5553470919324578,
      "grad_norm": 0.162491574883461,
      "learning_rate": 0.00014948,
      "loss": 1.2645,
      "step": 74
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 0.13756857812404633,
      "learning_rate": 0.0001515,
      "loss": 1.2418,
      "step": 75
    },
    {
      "epoch": 0.5703564727954972,
      "grad_norm": 0.15838544070720673,
      "learning_rate": 0.00015352,
      "loss": 1.2771,
      "step": 76
    },
    {
      "epoch": 0.5778611632270169,
      "grad_norm": 0.16285070776939392,
      "learning_rate": 0.00015554000000000002,
      "loss": 1.3045,
      "step": 77
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 0.15430669486522675,
      "learning_rate": 0.00015756,
      "loss": 1.3884,
      "step": 78
    },
    {
      "epoch": 0.5928705440900562,
      "grad_norm": 0.14215943217277527,
      "learning_rate": 0.00015958000000000001,
      "loss": 1.2564,
      "step": 79
    },
    {
      "epoch": 0.600375234521576,
      "grad_norm": 0.16277864575386047,
      "learning_rate": 0.00016160000000000002,
      "loss": 1.2688,
      "step": 80
    },
    {
      "epoch": 0.6078799249530957,
      "grad_norm": 0.1476689726114273,
      "learning_rate": 0.00016362,
      "loss": 1.1987,
      "step": 81
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.15069614350795746,
      "learning_rate": 0.00016564,
      "loss": 1.2378,
      "step": 82
    },
    {
      "epoch": 0.6228893058161351,
      "grad_norm": 0.1492914855480194,
      "learning_rate": 0.00016766,
      "loss": 1.3346,
      "step": 83
    },
    {
      "epoch": 0.6303939962476548,
      "grad_norm": 0.15325866639614105,
      "learning_rate": 0.00016968,
      "loss": 1.3019,
      "step": 84
    },
    {
      "epoch": 0.6378986866791745,
      "grad_norm": 0.15274472534656525,
      "learning_rate": 0.0001717,
      "loss": 1.2173,
      "step": 85
    },
    {
      "epoch": 0.6454033771106942,
      "grad_norm": 0.17855481803417206,
      "learning_rate": 0.00017372,
      "loss": 1.5073,
      "step": 86
    },
    {
      "epoch": 0.6529080675422139,
      "grad_norm": 0.14481309056282043,
      "learning_rate": 0.00017574,
      "loss": 1.3806,
      "step": 87
    },
    {
      "epoch": 0.6604127579737336,
      "grad_norm": 0.1616334468126297,
      "learning_rate": 0.00017776,
      "loss": 1.3155,
      "step": 88
    },
    {
      "epoch": 0.6679174484052532,
      "grad_norm": 0.16004852950572968,
      "learning_rate": 0.00017978000000000002,
      "loss": 1.2397,
      "step": 89
    },
    {
      "epoch": 0.6754221388367729,
      "grad_norm": 0.164125457406044,
      "learning_rate": 0.0001818,
      "loss": 1.3199,
      "step": 90
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 0.14032438397407532,
      "learning_rate": 0.00018382,
      "loss": 1.2188,
      "step": 91
    },
    {
      "epoch": 0.6904315196998124,
      "grad_norm": 0.15741504728794098,
      "learning_rate": 0.00018584000000000002,
      "loss": 1.3459,
      "step": 92
    },
    {
      "epoch": 0.6979362101313321,
      "grad_norm": 0.15716086328029633,
      "learning_rate": 0.00018786,
      "loss": 1.2176,
      "step": 93
    },
    {
      "epoch": 0.7054409005628518,
      "grad_norm": 0.15708261728286743,
      "learning_rate": 0.00018988,
      "loss": 1.296,
      "step": 94
    },
    {
      "epoch": 0.7129455909943715,
      "grad_norm": 0.1511552631855011,
      "learning_rate": 0.0001919,
      "loss": 1.2859,
      "step": 95
    },
    {
      "epoch": 0.7204502814258912,
      "grad_norm": 0.14795956015586853,
      "learning_rate": 0.00019392,
      "loss": 1.3124,
      "step": 96
    },
    {
      "epoch": 0.7279549718574109,
      "grad_norm": 0.17193551361560822,
      "learning_rate": 0.00019594,
      "loss": 1.2973,
      "step": 97
    },
    {
      "epoch": 0.7354596622889306,
      "grad_norm": 0.15188397467136383,
      "learning_rate": 0.00019796,
      "loss": 1.3048,
      "step": 98
    },
    {
      "epoch": 0.7429643527204502,
      "grad_norm": 0.15585751831531525,
      "learning_rate": 0.00019998,
      "loss": 1.3006,
      "step": 99
    },
    {
      "epoch": 0.7504690431519699,
      "grad_norm": 0.15665537118911743,
      "learning_rate": 0.000202,
      "loss": 1.4575,
      "step": 100
    },
    {
      "epoch": 0.7579737335834896,
      "grad_norm": 0.14343960583209991,
      "learning_rate": 0.0002019986155169901,
      "loss": 1.4348,
      "step": 101
    },
    {
      "epoch": 0.7654784240150094,
      "grad_norm": 0.1558350771665573,
      "learning_rate": 0.00020199446210591673,
      "loss": 1.223,
      "step": 102
    },
    {
      "epoch": 0.7729831144465291,
      "grad_norm": 0.1505260467529297,
      "learning_rate": 0.00020198753988064772,
      "loss": 1.1713,
      "step": 103
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 0.15138494968414307,
      "learning_rate": 0.0002019778490309594,
      "loss": 1.3375,
      "step": 104
    },
    {
      "epoch": 0.7879924953095685,
      "grad_norm": 0.15843193233013153,
      "learning_rate": 0.00020196538982253126,
      "loss": 1.21,
      "step": 105
    },
    {
      "epoch": 0.7954971857410882,
      "grad_norm": 0.14810356497764587,
      "learning_rate": 0.0002019501625969389,
      "loss": 1.2772,
      "step": 106
    },
    {
      "epoch": 0.8030018761726079,
      "grad_norm": 0.14969675242900848,
      "learning_rate": 0.00020193216777164428,
      "loss": 1.2787,
      "step": 107
    },
    {
      "epoch": 0.8105065666041276,
      "grad_norm": 0.147436261177063,
      "learning_rate": 0.0002019114058399847,
      "loss": 1.3535,
      "step": 108
    },
    {
      "epoch": 0.8180112570356473,
      "grad_norm": 0.151942178606987,
      "learning_rate": 0.00020188787737115897,
      "loss": 1.2255,
      "step": 109
    },
    {
      "epoch": 0.8255159474671669,
      "grad_norm": 0.1577390879392624,
      "learning_rate": 0.00020186158301021195,
      "loss": 1.2792,
      "step": 110
    },
    {
      "epoch": 0.8330206378986866,
      "grad_norm": 0.171566441655159,
      "learning_rate": 0.00020183252347801686,
      "loss": 1.2796,
      "step": 111
    },
    {
      "epoch": 0.8405253283302064,
      "grad_norm": 0.1481330841779709,
      "learning_rate": 0.00020180069957125544,
      "loss": 1.2864,
      "step": 112
    },
    {
      "epoch": 0.8480300187617261,
      "grad_norm": 0.15127785503864288,
      "learning_rate": 0.00020176611216239613,
      "loss": 1.2473,
      "step": 113
    },
    {
      "epoch": 0.8555347091932458,
      "grad_norm": 0.13949428498744965,
      "learning_rate": 0.00020172876219967027,
      "loss": 1.3336,
      "step": 114
    },
    {
      "epoch": 0.8630393996247655,
      "grad_norm": 0.15546847879886627,
      "learning_rate": 0.00020168865070704594,
      "loss": 1.2505,
      "step": 115
    },
    {
      "epoch": 0.8705440900562852,
      "grad_norm": 0.14151915907859802,
      "learning_rate": 0.00020164577878419994,
      "loss": 1.3321,
      "step": 116
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 0.16247083246707916,
      "learning_rate": 0.00020160014760648774,
      "loss": 1.1039,
      "step": 117
    },
    {
      "epoch": 0.8855534709193246,
      "grad_norm": 0.1587982028722763,
      "learning_rate": 0.00020155175842491107,
      "loss": 1.28,
      "step": 118
    },
    {
      "epoch": 0.8930581613508443,
      "grad_norm": 0.14681552350521088,
      "learning_rate": 0.00020150061256608387,
      "loss": 1.1579,
      "step": 119
    },
    {
      "epoch": 0.900562851782364,
      "grad_norm": 0.14857684075832367,
      "learning_rate": 0.0002014467114321956,
      "loss": 1.276,
      "step": 120
    },
    {
      "epoch": 0.9080675422138836,
      "grad_norm": 0.1468636840581894,
      "learning_rate": 0.00020139005650097317,
      "loss": 1.2918,
      "step": 121
    },
    {
      "epoch": 0.9155722326454033,
      "grad_norm": 0.15299977362155914,
      "learning_rate": 0.00020133064932564002,
      "loss": 1.196,
      "step": 122
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.14323540031909943,
      "learning_rate": 0.000201268491534874,
      "loss": 1.3483,
      "step": 123
    },
    {
      "epoch": 0.9305816135084428,
      "grad_norm": 0.16411356627941132,
      "learning_rate": 0.00020120358483276227,
      "loss": 1.2615,
      "step": 124
    },
    {
      "epoch": 0.9380863039399625,
      "grad_norm": 0.14410175383090973,
      "learning_rate": 0.00020113593099875486,
      "loss": 1.2669,
      "step": 125
    },
    {
      "epoch": 0.9455909943714822,
      "grad_norm": 0.16292551159858704,
      "learning_rate": 0.00020106553188761582,
      "loss": 1.2082,
      "step": 126
    },
    {
      "epoch": 0.9530956848030019,
      "grad_norm": 0.1520552784204483,
      "learning_rate": 0.0002009923894293723,
      "loss": 1.1822,
      "step": 127
    },
    {
      "epoch": 0.9606003752345216,
      "grad_norm": 0.16175416111946106,
      "learning_rate": 0.00020091650562926183,
      "loss": 1.2638,
      "step": 128
    },
    {
      "epoch": 0.9681050656660413,
      "grad_norm": 0.15966331958770752,
      "learning_rate": 0.00020083788256767702,
      "loss": 1.221,
      "step": 129
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.1437453031539917,
      "learning_rate": 0.00020075652240010892,
      "loss": 1.3926,
      "step": 130
    },
    {
      "epoch": 0.9831144465290806,
      "grad_norm": 0.141871839761734,
      "learning_rate": 0.00020067242735708754,
      "loss": 1.4327,
      "step": 131
    },
    {
      "epoch": 0.9906191369606003,
      "grad_norm": 0.15024764835834503,
      "learning_rate": 0.00020058559974412102,
      "loss": 1.2713,
      "step": 132
    },
    {
      "epoch": 0.99812382739212,
      "grad_norm": 0.14713317155838013,
      "learning_rate": 0.00020049604194163217,
      "loss": 1.2348,
      "step": 133
    },
    {
      "epoch": 1.0056285178236397,
      "grad_norm": 0.34685394167900085,
      "learning_rate": 0.00020040375640489343,
      "loss": 1.97,
      "step": 134
    },
    {
      "epoch": 1.0131332082551594,
      "grad_norm": 0.1429501622915268,
      "learning_rate": 0.00020030874566395943,
      "loss": 1.242,
      "step": 135
    },
    {
      "epoch": 1.020637898686679,
      "grad_norm": 0.15166796743869781,
      "learning_rate": 0.00020021101232359757,
      "loss": 1.2458,
      "step": 136
    },
    {
      "epoch": 1.0281425891181988,
      "grad_norm": 0.165684312582016,
      "learning_rate": 0.00020011055906321676,
      "loss": 1.2891,
      "step": 137
    },
    {
      "epoch": 1.0356472795497185,
      "grad_norm": 0.15929703414440155,
      "learning_rate": 0.0002000073886367939,
      "loss": 1.2887,
      "step": 138
    },
    {
      "epoch": 1.0431519699812384,
      "grad_norm": 0.15522639453411102,
      "learning_rate": 0.00019990150387279835,
      "loss": 1.1787,
      "step": 139
    },
    {
      "epoch": 1.050656660412758,
      "grad_norm": 0.17259898781776428,
      "learning_rate": 0.00019979290767411438,
      "loss": 1.1459,
      "step": 140
    },
    {
      "epoch": 1.0581613508442778,
      "grad_norm": 0.1543988436460495,
      "learning_rate": 0.00019968160301796163,
      "loss": 1.2485,
      "step": 141
    },
    {
      "epoch": 1.0656660412757974,
      "grad_norm": 0.16396988928318024,
      "learning_rate": 0.0001995675929558135,
      "loss": 1.3666,
      "step": 142
    },
    {
      "epoch": 1.0731707317073171,
      "grad_norm": 0.14671728014945984,
      "learning_rate": 0.0001994508806133134,
      "loss": 1.2105,
      "step": 143
    },
    {
      "epoch": 1.0806754221388368,
      "grad_norm": 0.1734577715396881,
      "learning_rate": 0.0001993314691901892,
      "loss": 1.1673,
      "step": 144
    },
    {
      "epoch": 1.0881801125703565,
      "grad_norm": 0.16767092049121857,
      "learning_rate": 0.00019920936196016534,
      "loss": 1.2222,
      "step": 145
    },
    {
      "epoch": 1.0956848030018762,
      "grad_norm": 0.16060315072536469,
      "learning_rate": 0.00019908456227087326,
      "loss": 1.2466,
      "step": 146
    },
    {
      "epoch": 1.103189493433396,
      "grad_norm": 0.15886545181274414,
      "learning_rate": 0.00019895707354375945,
      "loss": 1.1378,
      "step": 147
    },
    {
      "epoch": 1.1106941838649156,
      "grad_norm": 0.16979484260082245,
      "learning_rate": 0.00019882689927399174,
      "loss": 1.2946,
      "step": 148
    },
    {
      "epoch": 1.1181988742964353,
      "grad_norm": 0.16586163640022278,
      "learning_rate": 0.00019869404303036355,
      "loss": 1.2014,
      "step": 149
    },
    {
      "epoch": 1.125703564727955,
      "grad_norm": 0.16295990347862244,
      "learning_rate": 0.00019855850845519588,
      "loss": 1.2655,
      "step": 150
    },
    {
      "epoch": 1.1332082551594747,
      "grad_norm": 0.15378952026367188,
      "learning_rate": 0.00019842029926423762,
      "loss": 1.2853,
      "step": 151
    },
    {
      "epoch": 1.1407129455909943,
      "grad_norm": 0.15250574052333832,
      "learning_rate": 0.00019827941924656348,
      "loss": 1.2408,
      "step": 152
    },
    {
      "epoch": 1.148217636022514,
      "grad_norm": 0.15639042854309082,
      "learning_rate": 0.00019813587226447034,
      "loss": 1.2989,
      "step": 153
    },
    {
      "epoch": 1.1557223264540337,
      "grad_norm": 0.15901166200637817,
      "learning_rate": 0.00019798966225337126,
      "loss": 1.0916,
      "step": 154
    },
    {
      "epoch": 1.1632270168855534,
      "grad_norm": 0.15720036625862122,
      "learning_rate": 0.00019784079322168752,
      "loss": 1.2374,
      "step": 155
    },
    {
      "epoch": 1.170731707317073,
      "grad_norm": 0.15723657608032227,
      "learning_rate": 0.00019768926925073878,
      "loss": 1.2457,
      "step": 156
    },
    {
      "epoch": 1.1782363977485928,
      "grad_norm": 0.1488468050956726,
      "learning_rate": 0.00019753509449463134,
      "loss": 1.2046,
      "step": 157
    },
    {
      "epoch": 1.1857410881801125,
      "grad_norm": 0.1584654003381729,
      "learning_rate": 0.00019737827318014396,
      "loss": 1.2073,
      "step": 158
    },
    {
      "epoch": 1.1932457786116322,
      "grad_norm": 0.16308680176734924,
      "learning_rate": 0.00019721880960661223,
      "loss": 1.1337,
      "step": 159
    },
    {
      "epoch": 1.200750469043152,
      "grad_norm": 0.17071057856082916,
      "learning_rate": 0.00019705670814581052,
      "loss": 1.3404,
      "step": 160
    },
    {
      "epoch": 1.2082551594746715,
      "grad_norm": 0.17056912183761597,
      "learning_rate": 0.0001968919732418323,
      "loss": 1.17,
      "step": 161
    },
    {
      "epoch": 1.2157598499061915,
      "grad_norm": 0.16791465878486633,
      "learning_rate": 0.00019672460941096818,
      "loss": 1.1963,
      "step": 162
    },
    {
      "epoch": 1.2232645403377111,
      "grad_norm": 0.14966829121112823,
      "learning_rate": 0.0001965546212415821,
      "loss": 1.1704,
      "step": 163
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.1571626216173172,
      "learning_rate": 0.0001963820133939856,
      "loss": 1.3114,
      "step": 164
    },
    {
      "epoch": 1.2382739212007505,
      "grad_norm": 0.1762620061635971,
      "learning_rate": 0.00019620679060031003,
      "loss": 1.4425,
      "step": 165
    },
    {
      "epoch": 1.2457786116322702,
      "grad_norm": 0.15745310485363007,
      "learning_rate": 0.00019602895766437678,
      "loss": 1.0596,
      "step": 166
    },
    {
      "epoch": 1.25328330206379,
      "grad_norm": 0.15808038413524628,
      "learning_rate": 0.0001958485194615656,
      "loss": 1.1587,
      "step": 167
    },
    {
      "epoch": 1.2607879924953096,
      "grad_norm": 0.17968399822711945,
      "learning_rate": 0.00019566548093868106,
      "loss": 1.3376,
      "step": 168
    },
    {
      "epoch": 1.2682926829268293,
      "grad_norm": 0.1647973507642746,
      "learning_rate": 0.00019547984711381662,
      "loss": 1.224,
      "step": 169
    },
    {
      "epoch": 1.275797373358349,
      "grad_norm": 0.16712307929992676,
      "learning_rate": 0.00019529162307621738,
      "loss": 1.2078,
      "step": 170
    },
    {
      "epoch": 1.2833020637898687,
      "grad_norm": 0.168698251247406,
      "learning_rate": 0.00019510081398614045,
      "loss": 1.2032,
      "step": 171
    },
    {
      "epoch": 1.2908067542213884,
      "grad_norm": 0.16375304758548737,
      "learning_rate": 0.00019490742507471338,
      "loss": 1.1734,
      "step": 172
    },
    {
      "epoch": 1.298311444652908,
      "grad_norm": 0.17059938609600067,
      "learning_rate": 0.00019471146164379093,
      "loss": 1.2343,
      "step": 173
    },
    {
      "epoch": 1.3058161350844277,
      "grad_norm": 0.16624243557453156,
      "learning_rate": 0.00019451292906580948,
      "loss": 1.2059,
      "step": 174
    },
    {
      "epoch": 1.3133208255159474,
      "grad_norm": 0.17976044118404388,
      "learning_rate": 0.00019431183278363997,
      "loss": 1.2209,
      "step": 175
    },
    {
      "epoch": 1.320825515947467,
      "grad_norm": 0.1557455062866211,
      "learning_rate": 0.00019410817831043856,
      "loss": 1.169,
      "step": 176
    },
    {
      "epoch": 1.3283302063789868,
      "grad_norm": 0.1650816947221756,
      "learning_rate": 0.00019390197122949552,
      "loss": 1.3046,
      "step": 177
    },
    {
      "epoch": 1.3358348968105065,
      "grad_norm": 0.17368149757385254,
      "learning_rate": 0.0001936932171940821,
      "loss": 1.2899,
      "step": 178
    },
    {
      "epoch": 1.3433395872420264,
      "grad_norm": 0.14871825277805328,
      "learning_rate": 0.0001934819219272957,
      "loss": 1.1832,
      "step": 179
    },
    {
      "epoch": 1.3508442776735459,
      "grad_norm": 0.1574486494064331,
      "learning_rate": 0.0001932680912219027,
      "loss": 1.1538,
      "step": 180
    },
    {
      "epoch": 1.3583489681050658,
      "grad_norm": 0.16956180334091187,
      "learning_rate": 0.00019305173094017996,
      "loss": 1.3534,
      "step": 181
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 0.1664501577615738,
      "learning_rate": 0.00019283284701375393,
      "loss": 1.0876,
      "step": 182
    },
    {
      "epoch": 1.3733583489681052,
      "grad_norm": 0.17235621809959412,
      "learning_rate": 0.00019261144544343794,
      "loss": 1.1503,
      "step": 183
    },
    {
      "epoch": 1.3808630393996248,
      "grad_norm": 0.17231658101081848,
      "learning_rate": 0.00019238753229906797,
      "loss": 1.1031,
      "step": 184
    },
    {
      "epoch": 1.3883677298311445,
      "grad_norm": 0.15035244822502136,
      "learning_rate": 0.00019216111371933594,
      "loss": 1.0518,
      "step": 185
    },
    {
      "epoch": 1.3958724202626642,
      "grad_norm": 0.21291190385818481,
      "learning_rate": 0.00019193219591162155,
      "loss": 1.3352,
      "step": 186
    },
    {
      "epoch": 1.403377110694184,
      "grad_norm": 0.16924338042736053,
      "learning_rate": 0.00019170078515182216,
      "loss": 1.1968,
      "step": 187
    },
    {
      "epoch": 1.4108818011257036,
      "grad_norm": 0.17227092385292053,
      "learning_rate": 0.0001914668877841807,
      "loss": 1.3173,
      "step": 188
    },
    {
      "epoch": 1.4183864915572233,
      "grad_norm": 0.16759388148784637,
      "learning_rate": 0.0001912305102211116,
      "loss": 1.2863,
      "step": 189
    },
    {
      "epoch": 1.425891181988743,
      "grad_norm": 0.16890232264995575,
      "learning_rate": 0.00019099165894302515,
      "loss": 1.1406,
      "step": 190
    },
    {
      "epoch": 1.4333958724202627,
      "grad_norm": 0.16948051750659943,
      "learning_rate": 0.00019075034049814983,
      "loss": 1.2759,
      "step": 191
    },
    {
      "epoch": 1.4409005628517824,
      "grad_norm": 0.16042958199977875,
      "learning_rate": 0.00019050656150235268,
      "loss": 1.2761,
      "step": 192
    },
    {
      "epoch": 1.448405253283302,
      "grad_norm": 0.17444923520088196,
      "learning_rate": 0.00019026032863895805,
      "loss": 1.2717,
      "step": 193
    },
    {
      "epoch": 1.4559099437148217,
      "grad_norm": 0.18557068705558777,
      "learning_rate": 0.0001900116486585642,
      "loss": 1.296,
      "step": 194
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 0.1767808198928833,
      "learning_rate": 0.0001897605283788585,
      "loss": 1.338,
      "step": 195
    },
    {
      "epoch": 1.4709193245778611,
      "grad_norm": 0.16617947816848755,
      "learning_rate": 0.0001895069746844302,
      "loss": 1.1628,
      "step": 196
    },
    {
      "epoch": 1.4784240150093808,
      "grad_norm": 0.18191847205162048,
      "learning_rate": 0.000189250994526582,
      "loss": 1.2367,
      "step": 197
    },
    {
      "epoch": 1.4859287054409005,
      "grad_norm": 0.16195711493492126,
      "learning_rate": 0.00018899259492313915,
      "loss": 1.1815,
      "step": 198
    },
    {
      "epoch": 1.4934333958724202,
      "grad_norm": 0.1713341623544693,
      "learning_rate": 0.00018873178295825732,
      "loss": 1.2204,
      "step": 199
    },
    {
      "epoch": 1.50093808630394,
      "grad_norm": 0.17082621157169342,
      "learning_rate": 0.00018846856578222832,
      "loss": 1.2357,
      "step": 200
    },
    {
      "epoch": 1.5084427767354596,
      "grad_norm": 0.164811909198761,
      "learning_rate": 0.00018820295061128394,
      "loss": 1.3,
      "step": 201
    },
    {
      "epoch": 1.5159474671669795,
      "grad_norm": 0.16022959351539612,
      "learning_rate": 0.00018793494472739831,
      "loss": 1.2462,
      "step": 202
    },
    {
      "epoch": 1.523452157598499,
      "grad_norm": 0.14652663469314575,
      "learning_rate": 0.00018766455547808813,
      "loss": 1.1102,
      "step": 203
    },
    {
      "epoch": 1.5309568480300189,
      "grad_norm": 0.17688073217868805,
      "learning_rate": 0.0001873917902762112,
      "loss": 1.38,
      "step": 204
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.16433537006378174,
      "learning_rate": 0.0001871166565997633,
      "loss": 1.2009,
      "step": 205
    },
    {
      "epoch": 1.5459662288930582,
      "grad_norm": 0.16867071390151978,
      "learning_rate": 0.00018683916199167325,
      "loss": 1.1163,
      "step": 206
    },
    {
      "epoch": 1.5534709193245777,
      "grad_norm": 0.1684701293706894,
      "learning_rate": 0.00018655931405959586,
      "loss": 1.2663,
      "step": 207
    },
    {
      "epoch": 1.5609756097560976,
      "grad_norm": 0.1684935986995697,
      "learning_rate": 0.00018627712047570352,
      "loss": 1.2855,
      "step": 208
    },
    {
      "epoch": 1.5684803001876173,
      "grad_norm": 0.17889899015426636,
      "learning_rate": 0.00018599258897647594,
      "loss": 1.2835,
      "step": 209
    },
    {
      "epoch": 1.575984990619137,
      "grad_norm": 0.1752469539642334,
      "learning_rate": 0.00018570572736248782,
      "loss": 1.1292,
      "step": 210
    },
    {
      "epoch": 1.5834896810506567,
      "grad_norm": 0.15449506044387817,
      "learning_rate": 0.0001854165434981953,
      "loss": 1.1301,
      "step": 211
    },
    {
      "epoch": 1.5909943714821764,
      "grad_norm": 0.16106131672859192,
      "learning_rate": 0.00018512504531172005,
      "loss": 1.2506,
      "step": 212
    },
    {
      "epoch": 1.598499061913696,
      "grad_norm": 0.16306908428668976,
      "learning_rate": 0.0001848312407946321,
      "loss": 1.2689,
      "step": 213
    },
    {
      "epoch": 1.6060037523452158,
      "grad_norm": 0.16371643543243408,
      "learning_rate": 0.00018453513800173072,
      "loss": 1.3472,
      "step": 214
    },
    {
      "epoch": 1.6135084427767354,
      "grad_norm": 0.16967421770095825,
      "learning_rate": 0.00018423674505082356,
      "loss": 1.2062,
      "step": 215
    },
    {
      "epoch": 1.6210131332082551,
      "grad_norm": 0.16654905676841736,
      "learning_rate": 0.0001839360701225041,
      "loss": 1.1841,
      "step": 216
    },
    {
      "epoch": 1.6285178236397748,
      "grad_norm": 0.17192760109901428,
      "learning_rate": 0.00018363312145992737,
      "loss": 1.3917,
      "step": 217
    },
    {
      "epoch": 1.6360225140712945,
      "grad_norm": 0.17148400843143463,
      "learning_rate": 0.00018332790736858397,
      "loss": 1.1822,
      "step": 218
    },
    {
      "epoch": 1.6435272045028144,
      "grad_norm": 0.16966994106769562,
      "learning_rate": 0.00018302043621607245,
      "loss": 1.2164,
      "step": 219
    },
    {
      "epoch": 1.6510318949343339,
      "grad_norm": 0.1766677051782608,
      "learning_rate": 0.00018271071643186968,
      "loss": 1.3093,
      "step": 220
    },
    {
      "epoch": 1.6585365853658538,
      "grad_norm": 0.17478987574577332,
      "learning_rate": 0.0001823987565071001,
      "loss": 1.159,
      "step": 221
    },
    {
      "epoch": 1.6660412757973733,
      "grad_norm": 0.15623775124549866,
      "learning_rate": 0.00018208456499430256,
      "loss": 1.0533,
      "step": 222
    },
    {
      "epoch": 1.6735459662288932,
      "grad_norm": 0.16679462790489197,
      "learning_rate": 0.00018176815050719615,
      "loss": 1.1417,
      "step": 223
    },
    {
      "epoch": 1.6810506566604126,
      "grad_norm": 0.17018449306488037,
      "learning_rate": 0.00018144952172044381,
      "loss": 1.1481,
      "step": 224
    },
    {
      "epoch": 1.6885553470919326,
      "grad_norm": 0.17774926126003265,
      "learning_rate": 0.00018112868736941477,
      "loss": 1.0913,
      "step": 225
    },
    {
      "epoch": 1.696060037523452,
      "grad_norm": 0.16548708081245422,
      "learning_rate": 0.00018080565624994474,
      "loss": 1.2371,
      "step": 226
    },
    {
      "epoch": 1.703564727954972,
      "grad_norm": 0.1749250739812851,
      "learning_rate": 0.00018048043721809507,
      "loss": 1.2312,
      "step": 227
    },
    {
      "epoch": 1.7110694183864914,
      "grad_norm": 0.17566196620464325,
      "learning_rate": 0.00018015303918990982,
      "loss": 1.1096,
      "step": 228
    },
    {
      "epoch": 1.7185741088180113,
      "grad_norm": 0.18237973749637604,
      "learning_rate": 0.0001798234711411713,
      "loss": 1.2085,
      "step": 229
    },
    {
      "epoch": 1.726078799249531,
      "grad_norm": 0.1600469946861267,
      "learning_rate": 0.00017949174210715407,
      "loss": 1.2392,
      "step": 230
    },
    {
      "epoch": 1.7335834896810507,
      "grad_norm": 0.17133712768554688,
      "learning_rate": 0.00017915786118237714,
      "loss": 1.2948,
      "step": 231
    },
    {
      "epoch": 1.7410881801125704,
      "grad_norm": 0.16963626444339752,
      "learning_rate": 0.0001788218375203547,
      "loss": 1.1743,
      "step": 232
    },
    {
      "epoch": 1.74859287054409,
      "grad_norm": 0.16002561151981354,
      "learning_rate": 0.00017848368033334528,
      "loss": 1.2923,
      "step": 233
    },
    {
      "epoch": 1.7560975609756098,
      "grad_norm": 0.18220892548561096,
      "learning_rate": 0.00017814339889209887,
      "loss": 1.279,
      "step": 234
    },
    {
      "epoch": 1.7636022514071295,
      "grad_norm": 0.17350544035434723,
      "learning_rate": 0.00017780100252560313,
      "loss": 1.2086,
      "step": 235
    },
    {
      "epoch": 1.7711069418386491,
      "grad_norm": 0.1693098396062851,
      "learning_rate": 0.0001774565006208274,
      "loss": 1.253,
      "step": 236
    },
    {
      "epoch": 1.7786116322701688,
      "grad_norm": 0.16078263521194458,
      "learning_rate": 0.00017710990262246543,
      "loss": 1.1813,
      "step": 237
    },
    {
      "epoch": 1.7861163227016885,
      "grad_norm": 0.16030295193195343,
      "learning_rate": 0.0001767612180326764,
      "loss": 1.246,
      "step": 238
    },
    {
      "epoch": 1.7936210131332082,
      "grad_norm": 0.17168748378753662,
      "learning_rate": 0.00017641045641082453,
      "loss": 1.2802,
      "step": 239
    },
    {
      "epoch": 1.8011257035647281,
      "grad_norm": 0.17870783805847168,
      "learning_rate": 0.00017605762737321683,
      "loss": 1.1422,
      "step": 240
    },
    {
      "epoch": 1.8086303939962476,
      "grad_norm": 0.17350563406944275,
      "learning_rate": 0.0001757027405928396,
      "loss": 1.3117,
      "step": 241
    },
    {
      "epoch": 1.8161350844277675,
      "grad_norm": 0.1645297408103943,
      "learning_rate": 0.0001753458057990932,
      "loss": 1.058,
      "step": 242
    },
    {
      "epoch": 1.823639774859287,
      "grad_norm": 0.16183708608150482,
      "learning_rate": 0.00017498683277752527,
      "loss": 1.1449,
      "step": 243
    },
    {
      "epoch": 1.8311444652908069,
      "grad_norm": 0.16217318177223206,
      "learning_rate": 0.00017462583136956258,
      "loss": 1.3132,
      "step": 244
    },
    {
      "epoch": 1.8386491557223263,
      "grad_norm": 0.17231813073158264,
      "learning_rate": 0.00017426281147224105,
      "loss": 1.1858,
      "step": 245
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.1805432140827179,
      "learning_rate": 0.00017389778303793457,
      "loss": 1.2112,
      "step": 246
    },
    {
      "epoch": 1.8536585365853657,
      "grad_norm": 0.175805926322937,
      "learning_rate": 0.00017353075607408209,
      "loss": 1.2624,
      "step": 247
    },
    {
      "epoch": 1.8611632270168856,
      "grad_norm": 0.17483623325824738,
      "learning_rate": 0.00017316174064291315,
      "loss": 1.3083,
      "step": 248
    },
    {
      "epoch": 1.868667917448405,
      "grad_norm": 0.17274804413318634,
      "learning_rate": 0.00017279074686117225,
      "loss": 1.1192,
      "step": 249
    },
    {
      "epoch": 1.876172607879925,
      "grad_norm": 0.1718122363090515,
      "learning_rate": 0.0001724177848998413,
      "loss": 1.1241,
      "step": 250
    },
    {
      "epoch": 1.8836772983114447,
      "grad_norm": 0.16955551505088806,
      "learning_rate": 0.0001720428649838609,
      "loss": 1.1583,
      "step": 251
    },
    {
      "epoch": 1.8911819887429644,
      "grad_norm": 0.1618235558271408,
      "learning_rate": 0.0001716659973918499,
      "loss": 1.1299,
      "step": 252
    },
    {
      "epoch": 1.898686679174484,
      "grad_norm": 0.17560751736164093,
      "learning_rate": 0.00017128719245582374,
      "loss": 1.292,
      "step": 253
    },
    {
      "epoch": 1.9061913696060038,
      "grad_norm": 0.16492795944213867,
      "learning_rate": 0.0001709064605609111,
      "loss": 1.156,
      "step": 254
    },
    {
      "epoch": 1.9136960600375235,
      "grad_norm": 0.17323772609233856,
      "learning_rate": 0.00017052381214506914,
      "loss": 1.2038,
      "step": 255
    },
    {
      "epoch": 1.9212007504690432,
      "grad_norm": 0.17622260749340057,
      "learning_rate": 0.00017013925769879755,
      "loss": 1.3161,
      "step": 256
    },
    {
      "epoch": 1.9287054409005628,
      "grad_norm": 0.17470543086528778,
      "learning_rate": 0.0001697528077648507,
      "loss": 1.2145,
      "step": 257
    },
    {
      "epoch": 1.9362101313320825,
      "grad_norm": 0.17390191555023193,
      "learning_rate": 0.00016936447293794873,
      "loss": 1.1653,
      "step": 258
    },
    {
      "epoch": 1.9437148217636022,
      "grad_norm": 0.166363924741745,
      "learning_rate": 0.0001689742638644871,
      "loss": 1.1886,
      "step": 259
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.1542292982339859,
      "learning_rate": 0.0001685821912422447,
      "loss": 1.0965,
      "step": 260
    },
    {
      "epoch": 1.9587242026266416,
      "grad_norm": 0.17397943139076233,
      "learning_rate": 0.00016818826582009044,
      "loss": 1.3139,
      "step": 261
    },
    {
      "epoch": 1.9662288930581613,
      "grad_norm": 0.16641609370708466,
      "learning_rate": 0.00016779249839768884,
      "loss": 1.2891,
      "step": 262
    },
    {
      "epoch": 1.9737335834896812,
      "grad_norm": 0.16674664616584778,
      "learning_rate": 0.00016739489982520368,
      "loss": 1.3371,
      "step": 263
    },
    {
      "epoch": 1.9812382739212007,
      "grad_norm": 0.17100049555301666,
      "learning_rate": 0.00016699548100300066,
      "loss": 1.2478,
      "step": 264
    },
    {
      "epoch": 1.9887429643527206,
      "grad_norm": 0.16496621072292328,
      "learning_rate": 0.00016659425288134854,
      "loss": 1.1691,
      "step": 265
    },
    {
      "epoch": 1.99624765478424,
      "grad_norm": 0.17649468779563904,
      "learning_rate": 0.00016619122646011902,
      "loss": 1.1012,
      "step": 266
    },
    {
      "epoch": 2.00375234521576,
      "grad_norm": 0.41986083984375,
      "learning_rate": 0.00016578641278848497,
      "loss": 2.1257,
      "step": 267
    },
    {
      "epoch": 2.0112570356472794,
      "grad_norm": 0.1715676486492157,
      "learning_rate": 0.00016537982296461768,
      "loss": 1.1251,
      "step": 268
    },
    {
      "epoch": 2.0187617260787993,
      "grad_norm": 0.176234632730484,
      "learning_rate": 0.00016497146813538257,
      "loss": 1.3606,
      "step": 269
    },
    {
      "epoch": 2.026266416510319,
      "grad_norm": 0.1678568720817566,
      "learning_rate": 0.00016456135949603358,
      "loss": 1.2104,
      "step": 270
    },
    {
      "epoch": 2.0337711069418387,
      "grad_norm": 0.18491528928279877,
      "learning_rate": 0.00016414950828990625,
      "loss": 1.1191,
      "step": 271
    },
    {
      "epoch": 2.041275797373358,
      "grad_norm": 0.18430155515670776,
      "learning_rate": 0.00016373592580810935,
      "loss": 1.1588,
      "step": 272
    },
    {
      "epoch": 2.048780487804878,
      "grad_norm": 0.17635607719421387,
      "learning_rate": 0.00016332062338921563,
      "loss": 1.0215,
      "step": 273
    },
    {
      "epoch": 2.0562851782363976,
      "grad_norm": 0.1976105123758316,
      "learning_rate": 0.00016290361241895064,
      "loss": 1.2266,
      "step": 274
    },
    {
      "epoch": 2.0637898686679175,
      "grad_norm": 0.17706206440925598,
      "learning_rate": 0.0001624849043298808,
      "loss": 1.1343,
      "step": 275
    },
    {
      "epoch": 2.071294559099437,
      "grad_norm": 0.17901922762393951,
      "learning_rate": 0.00016206451060109988,
      "loss": 1.0875,
      "step": 276
    },
    {
      "epoch": 2.078799249530957,
      "grad_norm": 0.18593402206897736,
      "learning_rate": 0.0001616424427579143,
      "loss": 1.2195,
      "step": 277
    },
    {
      "epoch": 2.0863039399624768,
      "grad_norm": 0.18124711513519287,
      "learning_rate": 0.0001612187123715272,
      "loss": 1.1383,
      "step": 278
    },
    {
      "epoch": 2.0938086303939962,
      "grad_norm": 0.18754807114601135,
      "learning_rate": 0.00016079333105872118,
      "loss": 1.2342,
      "step": 279
    },
    {
      "epoch": 2.101313320825516,
      "grad_norm": 0.1743268072605133,
      "learning_rate": 0.00016036631048153979,
      "loss": 1.0966,
      "step": 280
    },
    {
      "epoch": 2.1088180112570356,
      "grad_norm": 0.18791033327579498,
      "learning_rate": 0.00015993766234696785,
      "loss": 1.0818,
      "step": 281
    },
    {
      "epoch": 2.1163227016885555,
      "grad_norm": 0.18581430613994598,
      "learning_rate": 0.00015950739840661055,
      "loss": 1.1301,
      "step": 282
    },
    {
      "epoch": 2.123827392120075,
      "grad_norm": 0.19216033816337585,
      "learning_rate": 0.00015907553045637116,
      "loss": 1.3461,
      "step": 283
    },
    {
      "epoch": 2.131332082551595,
      "grad_norm": 0.1855672150850296,
      "learning_rate": 0.00015864207033612762,
      "loss": 1.1279,
      "step": 284
    },
    {
      "epoch": 2.1388367729831144,
      "grad_norm": 0.17570380866527557,
      "learning_rate": 0.00015820702992940813,
      "loss": 1.0614,
      "step": 285
    },
    {
      "epoch": 2.1463414634146343,
      "grad_norm": 0.19156311452388763,
      "learning_rate": 0.0001577704211630652,
      "loss": 1.2225,
      "step": 286
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 0.1879376322031021,
      "learning_rate": 0.00015733225600694873,
      "loss": 1.13,
      "step": 287
    },
    {
      "epoch": 2.1613508442776737,
      "grad_norm": 0.18353459239006042,
      "learning_rate": 0.00015689254647357776,
      "loss": 1.1511,
      "step": 288
    },
    {
      "epoch": 2.168855534709193,
      "grad_norm": 0.17921237647533417,
      "learning_rate": 0.0001564513046178113,
      "loss": 1.0845,
      "step": 289
    },
    {
      "epoch": 2.176360225140713,
      "grad_norm": 0.21026204526424408,
      "learning_rate": 0.00015600854253651776,
      "loss": 1.1584,
      "step": 290
    },
    {
      "epoch": 2.1838649155722325,
      "grad_norm": 0.18646669387817383,
      "learning_rate": 0.00015556427236824318,
      "loss": 1.0322,
      "step": 291
    },
    {
      "epoch": 2.1913696060037524,
      "grad_norm": 0.20405703783035278,
      "learning_rate": 0.00015511850629287865,
      "loss": 1.4473,
      "step": 292
    },
    {
      "epoch": 2.198874296435272,
      "grad_norm": 0.18749438226222992,
      "learning_rate": 0.00015467125653132637,
      "loss": 1.034,
      "step": 293
    },
    {
      "epoch": 2.206378986866792,
      "grad_norm": 0.20728205144405365,
      "learning_rate": 0.00015422253534516444,
      "loss": 1.2028,
      "step": 294
    },
    {
      "epoch": 2.2138836772983113,
      "grad_norm": 0.20661289989948273,
      "learning_rate": 0.00015377235503631083,
      "loss": 1.0371,
      "step": 295
    },
    {
      "epoch": 2.221388367729831,
      "grad_norm": 0.19144625961780548,
      "learning_rate": 0.00015332072794668617,
      "loss": 1.0591,
      "step": 296
    },
    {
      "epoch": 2.2288930581613506,
      "grad_norm": 0.2048298567533493,
      "learning_rate": 0.0001528676664578752,
      "loss": 1.33,
      "step": 297
    },
    {
      "epoch": 2.2363977485928705,
      "grad_norm": 0.19968393445014954,
      "learning_rate": 0.00015241318299078751,
      "loss": 1.1372,
      "step": 298
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.1912514716386795,
      "learning_rate": 0.00015195729000531694,
      "loss": 1.1152,
      "step": 299
    },
    {
      "epoch": 2.25140712945591,
      "grad_norm": 0.20684240758419037,
      "learning_rate": 0.0001515,
      "loss": 1.1317,
      "step": 300
    },
    {
      "epoch": 2.2589118198874294,
      "grad_norm": 0.1938362568616867,
      "learning_rate": 0.00015104132551167318,
      "loss": 1.1085,
      "step": 301
    },
    {
      "epoch": 2.2664165103189493,
      "grad_norm": 0.22417286038398743,
      "learning_rate": 0.00015058127911512923,
      "loss": 1.1513,
      "step": 302
    },
    {
      "epoch": 2.273921200750469,
      "grad_norm": 0.2011910080909729,
      "learning_rate": 0.00015011987342277255,
      "loss": 1.1317,
      "step": 303
    },
    {
      "epoch": 2.2814258911819887,
      "grad_norm": 0.18712817132472992,
      "learning_rate": 0.00014965712108427323,
      "loss": 1.0704,
      "step": 304
    },
    {
      "epoch": 2.2889305816135086,
      "grad_norm": 0.19896912574768066,
      "learning_rate": 0.00014919303478622045,
      "loss": 1.1892,
      "step": 305
    },
    {
      "epoch": 2.296435272045028,
      "grad_norm": 0.2061881124973297,
      "learning_rate": 0.00014872762725177447,
      "loss": 1.0772,
      "step": 306
    },
    {
      "epoch": 2.303939962476548,
      "grad_norm": 0.20177169144153595,
      "learning_rate": 0.00014826091124031792,
      "loss": 1.104,
      "step": 307
    },
    {
      "epoch": 2.3114446529080674,
      "grad_norm": 0.20972506701946259,
      "learning_rate": 0.00014779289954710604,
      "loss": 1.1659,
      "step": 308
    },
    {
      "epoch": 2.3189493433395874,
      "grad_norm": 0.1818305253982544,
      "learning_rate": 0.00014732360500291583,
      "loss": 0.9569,
      "step": 309
    },
    {
      "epoch": 2.326454033771107,
      "grad_norm": 0.20779147744178772,
      "learning_rate": 0.00014685304047369423,
      "loss": 1.1488,
      "step": 310
    },
    {
      "epoch": 2.3339587242026267,
      "grad_norm": 0.20816084742546082,
      "learning_rate": 0.00014638121886020555,
      "loss": 1.1104,
      "step": 311
    },
    {
      "epoch": 2.341463414634146,
      "grad_norm": 0.21499446034431458,
      "learning_rate": 0.00014590815309767767,
      "loss": 1.1837,
      "step": 312
    },
    {
      "epoch": 2.348968105065666,
      "grad_norm": 0.19324234127998352,
      "learning_rate": 0.00014543385615544744,
      "loss": 1.1552,
      "step": 313
    },
    {
      "epoch": 2.3564727954971856,
      "grad_norm": 0.20301885902881622,
      "learning_rate": 0.0001449583410366051,
      "loss": 1.2188,
      "step": 314
    },
    {
      "epoch": 2.3639774859287055,
      "grad_norm": 0.20692458748817444,
      "learning_rate": 0.00014448162077763783,
      "loss": 1.1173,
      "step": 315
    },
    {
      "epoch": 2.371482176360225,
      "grad_norm": 0.20142214000225067,
      "learning_rate": 0.00014400370844807234,
      "loss": 1.1568,
      "step": 316
    },
    {
      "epoch": 2.378986866791745,
      "grad_norm": 0.21007882058620453,
      "learning_rate": 0.0001435246171501166,
      "loss": 1.0795,
      "step": 317
    },
    {
      "epoch": 2.3864915572232643,
      "grad_norm": 0.2005329579114914,
      "learning_rate": 0.00014304436001830054,
      "loss": 1.1222,
      "step": 318
    },
    {
      "epoch": 2.3939962476547842,
      "grad_norm": 0.18658891320228577,
      "learning_rate": 0.000142562950219116,
      "loss": 1.067,
      "step": 319
    },
    {
      "epoch": 2.401500938086304,
      "grad_norm": 0.21223872900009155,
      "learning_rate": 0.00014208040095065584,
      "loss": 1.2728,
      "step": 320
    },
    {
      "epoch": 2.4090056285178236,
      "grad_norm": 0.19328981637954712,
      "learning_rate": 0.000141596725442252,
      "loss": 1.0725,
      "step": 321
    },
    {
      "epoch": 2.416510318949343,
      "grad_norm": 0.2031588852405548,
      "learning_rate": 0.00014111193695411285,
      "loss": 1.0599,
      "step": 322
    },
    {
      "epoch": 2.424015009380863,
      "grad_norm": 0.21044471859931946,
      "learning_rate": 0.00014062604877695972,
      "loss": 1.2219,
      "step": 323
    },
    {
      "epoch": 2.431519699812383,
      "grad_norm": 0.19686193764209747,
      "learning_rate": 0.0001401390742316624,
      "loss": 1.1031,
      "step": 324
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.1939302682876587,
      "learning_rate": 0.00013965102666887408,
      "loss": 1.0597,
      "step": 325
    },
    {
      "epoch": 2.4465290806754223,
      "grad_norm": 0.20583440363407135,
      "learning_rate": 0.0001391619194686652,
      "loss": 1.1094,
      "step": 326
    },
    {
      "epoch": 2.4540337711069418,
      "grad_norm": 0.20032913982868195,
      "learning_rate": 0.00013867176604015672,
      "loss": 1.155,
      "step": 327
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 0.20888358354568481,
      "learning_rate": 0.0001381805798211525,
      "loss": 1.0728,
      "step": 328
    },
    {
      "epoch": 2.469043151969981,
      "grad_norm": 0.23921048641204834,
      "learning_rate": 0.00013768837427777082,
      "loss": 1.2459,
      "step": 329
    },
    {
      "epoch": 2.476547842401501,
      "grad_norm": 0.1988176703453064,
      "learning_rate": 0.0001371951629040753,
      "loss": 1.1937,
      "step": 330
    },
    {
      "epoch": 2.4840525328330205,
      "grad_norm": 0.2069939821958542,
      "learning_rate": 0.00013670095922170498,
      "loss": 1.309,
      "step": 331
    },
    {
      "epoch": 2.4915572232645404,
      "grad_norm": 0.19495587050914764,
      "learning_rate": 0.00013620577677950335,
      "loss": 1.1892,
      "step": 332
    },
    {
      "epoch": 2.49906191369606,
      "grad_norm": 0.19906669855117798,
      "learning_rate": 0.00013570962915314725,
      "loss": 1.2079,
      "step": 333
    },
    {
      "epoch": 2.50656660412758,
      "grad_norm": 0.2080148458480835,
      "learning_rate": 0.00013521252994477446,
      "loss": 1.2381,
      "step": 334
    },
    {
      "epoch": 2.5140712945590993,
      "grad_norm": 0.20452843606472015,
      "learning_rate": 0.00013471449278261086,
      "loss": 1.1872,
      "step": 335
    },
    {
      "epoch": 2.521575984990619,
      "grad_norm": 0.2072499543428421,
      "learning_rate": 0.0001342155313205969,
      "loss": 1.1783,
      "step": 336
    },
    {
      "epoch": 2.529080675422139,
      "grad_norm": 0.18055139482021332,
      "learning_rate": 0.0001337156592380131,
      "loss": 0.9911,
      "step": 337
    },
    {
      "epoch": 2.5365853658536586,
      "grad_norm": 0.21083617210388184,
      "learning_rate": 0.00013321489023910508,
      "loss": 1.207,
      "step": 338
    },
    {
      "epoch": 2.544090056285178,
      "grad_norm": 0.20006054639816284,
      "learning_rate": 0.0001327132380527079,
      "loss": 1.1114,
      "step": 339
    },
    {
      "epoch": 2.551594746716698,
      "grad_norm": 0.2103322297334671,
      "learning_rate": 0.0001322107164318697,
      "loss": 1.3467,
      "step": 340
    },
    {
      "epoch": 2.559099437148218,
      "grad_norm": 0.20280379056930542,
      "learning_rate": 0.00013170733915347451,
      "loss": 1.0259,
      "step": 341
    },
    {
      "epoch": 2.5666041275797373,
      "grad_norm": 0.2019176334142685,
      "learning_rate": 0.00013120312001786477,
      "loss": 1.0695,
      "step": 342
    },
    {
      "epoch": 2.574108818011257,
      "grad_norm": 0.21050766110420227,
      "learning_rate": 0.0001306980728484627,
      "loss": 1.2062,
      "step": 343
    },
    {
      "epoch": 2.5816135084427767,
      "grad_norm": 0.1875392496585846,
      "learning_rate": 0.00013019221149139162,
      "loss": 1.1699,
      "step": 344
    },
    {
      "epoch": 2.5891181988742966,
      "grad_norm": 0.20021779835224152,
      "learning_rate": 0.00012968554981509622,
      "loss": 1.2177,
      "step": 345
    },
    {
      "epoch": 2.596622889305816,
      "grad_norm": 0.19720782339572906,
      "learning_rate": 0.00012917810170996218,
      "loss": 1.1044,
      "step": 346
    },
    {
      "epoch": 2.604127579737336,
      "grad_norm": 0.21691443026065826,
      "learning_rate": 0.0001286698810879357,
      "loss": 1.2225,
      "step": 347
    },
    {
      "epoch": 2.6116322701688555,
      "grad_norm": 0.20864421129226685,
      "learning_rate": 0.00012816090188214182,
      "loss": 1.0991,
      "step": 348
    },
    {
      "epoch": 2.6191369606003754,
      "grad_norm": 0.2138601839542389,
      "learning_rate": 0.00012765117804650267,
      "loss": 1.2061,
      "step": 349
    },
    {
      "epoch": 2.626641651031895,
      "grad_norm": 0.22152799367904663,
      "learning_rate": 0.0001271407235553546,
      "loss": 1.2227,
      "step": 350
    },
    {
      "epoch": 2.6341463414634148,
      "grad_norm": 0.18970821797847748,
      "learning_rate": 0.00012662955240306538,
      "loss": 1.1466,
      "step": 351
    },
    {
      "epoch": 2.641651031894934,
      "grad_norm": 0.21198885142803192,
      "learning_rate": 0.00012611767860365038,
      "loss": 1.1457,
      "step": 352
    },
    {
      "epoch": 2.649155722326454,
      "grad_norm": 0.22691020369529724,
      "learning_rate": 0.00012560511619038827,
      "loss": 1.2352,
      "step": 353
    },
    {
      "epoch": 2.6566604127579736,
      "grad_norm": 0.20230384171009064,
      "learning_rate": 0.00012509187921543667,
      "loss": 1.0936,
      "step": 354
    },
    {
      "epoch": 2.6641651031894935,
      "grad_norm": 0.19077181816101074,
      "learning_rate": 0.00012457798174944645,
      "loss": 1.0474,
      "step": 355
    },
    {
      "epoch": 2.671669793621013,
      "grad_norm": 0.2141447812318802,
      "learning_rate": 0.00012406343788117625,
      "loss": 1.2607,
      "step": 356
    },
    {
      "epoch": 2.679174484052533,
      "grad_norm": 0.20831094682216644,
      "learning_rate": 0.0001235482617171061,
      "loss": 1.0861,
      "step": 357
    },
    {
      "epoch": 2.686679174484053,
      "grad_norm": 0.19394421577453613,
      "learning_rate": 0.00012303246738105082,
      "loss": 1.1731,
      "step": 358
    },
    {
      "epoch": 2.6941838649155723,
      "grad_norm": 0.19684839248657227,
      "learning_rate": 0.00012251606901377265,
      "loss": 1.1461,
      "step": 359
    },
    {
      "epoch": 2.7016885553470917,
      "grad_norm": 0.19147644937038422,
      "learning_rate": 0.00012199908077259367,
      "loss": 1.0678,
      "step": 360
    },
    {
      "epoch": 2.7091932457786116,
      "grad_norm": 0.21550385653972626,
      "learning_rate": 0.00012148151683100776,
      "loss": 1.2684,
      "step": 361
    },
    {
      "epoch": 2.7166979362101316,
      "grad_norm": 0.20470251142978668,
      "learning_rate": 0.00012096339137829174,
      "loss": 1.044,
      "step": 362
    },
    {
      "epoch": 2.724202626641651,
      "grad_norm": 0.2159148007631302,
      "learning_rate": 0.00012044471861911666,
      "loss": 1.2416,
      "step": 363
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.2075643092393875,
      "learning_rate": 0.0001199255127731582,
      "loss": 1.2043,
      "step": 364
    },
    {
      "epoch": 2.7392120075046904,
      "grad_norm": 0.21417608857154846,
      "learning_rate": 0.00011940578807470692,
      "loss": 1.2213,
      "step": 365
    },
    {
      "epoch": 2.7467166979362103,
      "grad_norm": 0.2100561261177063,
      "learning_rate": 0.00011888555877227793,
      "loss": 1.2066,
      "step": 366
    },
    {
      "epoch": 2.75422138836773,
      "grad_norm": 0.186203733086586,
      "learning_rate": 0.00011836483912822035,
      "loss": 1.0488,
      "step": 367
    },
    {
      "epoch": 2.7617260787992497,
      "grad_norm": 0.19459331035614014,
      "learning_rate": 0.00011784364341832634,
      "loss": 1.0966,
      "step": 368
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.20786772668361664,
      "learning_rate": 0.00011732198593143949,
      "loss": 1.1638,
      "step": 369
    },
    {
      "epoch": 2.776735459662289,
      "grad_norm": 0.19376277923583984,
      "learning_rate": 0.00011679988096906333,
      "loss": 1.0929,
      "step": 370
    },
    {
      "epoch": 2.7842401500938085,
      "grad_norm": 0.19773966073989868,
      "learning_rate": 0.00011627734284496917,
      "loss": 1.0705,
      "step": 371
    },
    {
      "epoch": 2.7917448405253285,
      "grad_norm": 0.19659113883972168,
      "learning_rate": 0.00011575438588480359,
      "loss": 1.0056,
      "step": 372
    },
    {
      "epoch": 2.799249530956848,
      "grad_norm": 0.20462089776992798,
      "learning_rate": 0.00011523102442569585,
      "loss": 1.051,
      "step": 373
    },
    {
      "epoch": 2.806754221388368,
      "grad_norm": 0.20397114753723145,
      "learning_rate": 0.00011470727281586475,
      "loss": 1.1343,
      "step": 374
    },
    {
      "epoch": 2.8142589118198873,
      "grad_norm": 0.21280498802661896,
      "learning_rate": 0.00011418314541422523,
      "loss": 1.1347,
      "step": 375
    },
    {
      "epoch": 2.821763602251407,
      "grad_norm": 0.22884762287139893,
      "learning_rate": 0.00011365865658999474,
      "loss": 1.0453,
      "step": 376
    },
    {
      "epoch": 2.8292682926829267,
      "grad_norm": 0.21105338633060455,
      "learning_rate": 0.00011313382072229936,
      "loss": 1.3419,
      "step": 377
    },
    {
      "epoch": 2.8367729831144466,
      "grad_norm": 0.1944011002779007,
      "learning_rate": 0.00011260865219977954,
      "loss": 1.1077,
      "step": 378
    },
    {
      "epoch": 2.8442776735459665,
      "grad_norm": 0.2119598686695099,
      "learning_rate": 0.00011208316542019556,
      "loss": 1.1795,
      "step": 379
    },
    {
      "epoch": 2.851782363977486,
      "grad_norm": 0.21490478515625,
      "learning_rate": 0.00011155737479003301,
      "loss": 1.1474,
      "step": 380
    },
    {
      "epoch": 2.8592870544090054,
      "grad_norm": 0.2159854918718338,
      "learning_rate": 0.00011103129472410755,
      "loss": 1.0877,
      "step": 381
    },
    {
      "epoch": 2.8667917448405253,
      "grad_norm": 0.20630186796188354,
      "learning_rate": 0.00011050493964516997,
      "loss": 1.0974,
      "step": 382
    },
    {
      "epoch": 2.8742964352720453,
      "grad_norm": 0.19603201746940613,
      "learning_rate": 0.00010997832398351062,
      "loss": 0.9308,
      "step": 383
    },
    {
      "epoch": 2.8818011257035647,
      "grad_norm": 0.21403640508651733,
      "learning_rate": 0.0001094514621765639,
      "loss": 1.3101,
      "step": 384
    },
    {
      "epoch": 2.889305816135084,
      "grad_norm": 0.19674979150295258,
      "learning_rate": 0.00010892436866851235,
      "loss": 1.1427,
      "step": 385
    },
    {
      "epoch": 2.896810506566604,
      "grad_norm": 0.1999547779560089,
      "learning_rate": 0.0001083970579098908,
      "loss": 1.2693,
      "step": 386
    },
    {
      "epoch": 2.904315196998124,
      "grad_norm": 0.20687858760356903,
      "learning_rate": 0.00010786954435719008,
      "loss": 1.0435,
      "step": 387
    },
    {
      "epoch": 2.9118198874296435,
      "grad_norm": 0.227890744805336,
      "learning_rate": 0.00010734184247246066,
      "loss": 1.2724,
      "step": 388
    },
    {
      "epoch": 2.919324577861163,
      "grad_norm": 0.221451535820961,
      "learning_rate": 0.00010681396672291631,
      "loss": 1.0916,
      "step": 389
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.19945359230041504,
      "learning_rate": 0.00010628593158053734,
      "loss": 0.9519,
      "step": 390
    },
    {
      "epoch": 2.9343339587242028,
      "grad_norm": 0.21845029294490814,
      "learning_rate": 0.00010575775152167391,
      "loss": 1.0897,
      "step": 391
    },
    {
      "epoch": 2.9418386491557222,
      "grad_norm": 0.20638839900493622,
      "learning_rate": 0.00010522944102664915,
      "loss": 1.0732,
      "step": 392
    },
    {
      "epoch": 2.949343339587242,
      "grad_norm": 0.2033357471227646,
      "learning_rate": 0.00010470101457936219,
      "loss": 1.0864,
      "step": 393
    },
    {
      "epoch": 2.9568480300187616,
      "grad_norm": 0.21390041708946228,
      "learning_rate": 0.00010417248666689095,
      "loss": 1.142,
      "step": 394
    },
    {
      "epoch": 2.9643527204502815,
      "grad_norm": 0.20132870972156525,
      "learning_rate": 0.00010364387177909521,
      "loss": 1.0189,
      "step": 395
    },
    {
      "epoch": 2.971857410881801,
      "grad_norm": 0.23031547665596008,
      "learning_rate": 0.00010311518440821906,
      "loss": 1.2755,
      "step": 396
    },
    {
      "epoch": 2.979362101313321,
      "grad_norm": 0.21445465087890625,
      "learning_rate": 0.0001025864390484939,
      "loss": 1.0362,
      "step": 397
    },
    {
      "epoch": 2.9868667917448404,
      "grad_norm": 0.22398123145103455,
      "learning_rate": 0.00010205765019574084,
      "loss": 1.1973,
      "step": 398
    },
    {
      "epoch": 2.9943714821763603,
      "grad_norm": 0.20312932133674622,
      "learning_rate": 0.00010152883234697336,
      "loss": 1.1794,
      "step": 399
    },
    {
      "epoch": 3.0018761726078798,
      "grad_norm": 0.45374202728271484,
      "learning_rate": 0.000101,
      "loss": 2.3355,
      "step": 400
    },
    {
      "epoch": 3.0093808630393997,
      "grad_norm": 0.1999255269765854,
      "learning_rate": 0.00010047116765302661,
      "loss": 1.0259,
      "step": 401
    },
    {
      "epoch": 3.016885553470919,
      "grad_norm": 0.20511388778686523,
      "learning_rate": 9.994234980425921e-05,
      "loss": 1.1053,
      "step": 402
    },
    {
      "epoch": 3.024390243902439,
      "grad_norm": 0.20204809308052063,
      "learning_rate": 9.941356095150613e-05,
      "loss": 1.0375,
      "step": 403
    },
    {
      "epoch": 3.0318949343339585,
      "grad_norm": 0.22650550305843353,
      "learning_rate": 9.888481559178096e-05,
      "loss": 1.1406,
      "step": 404
    },
    {
      "epoch": 3.0393996247654784,
      "grad_norm": 0.20978747308254242,
      "learning_rate": 9.835612822090483e-05,
      "loss": 0.9847,
      "step": 405
    },
    {
      "epoch": 3.0469043151969983,
      "grad_norm": 0.2218317836523056,
      "learning_rate": 9.782751333310905e-05,
      "loss": 1.1565,
      "step": 406
    },
    {
      "epoch": 3.054409005628518,
      "grad_norm": 0.22472409904003143,
      "learning_rate": 9.72989854206378e-05,
      "loss": 1.179,
      "step": 407
    },
    {
      "epoch": 3.0619136960600377,
      "grad_norm": 0.1973821222782135,
      "learning_rate": 9.677055897335087e-05,
      "loss": 0.9192,
      "step": 408
    },
    {
      "epoch": 3.069418386491557,
      "grad_norm": 0.2326335310935974,
      "learning_rate": 9.62422484783261e-05,
      "loss": 1.1715,
      "step": 409
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.22134754061698914,
      "learning_rate": 9.571406841946267e-05,
      "loss": 1.073,
      "step": 410
    },
    {
      "epoch": 3.0844277673545966,
      "grad_norm": 0.21591591835021973,
      "learning_rate": 9.518603327708372e-05,
      "loss": 1.0896,
      "step": 411
    },
    {
      "epoch": 3.0919324577861165,
      "grad_norm": 0.22626814246177673,
      "learning_rate": 9.465815752753935e-05,
      "loss": 1.1149,
      "step": 412
    },
    {
      "epoch": 3.099437148217636,
      "grad_norm": 0.22526519000530243,
      "learning_rate": 9.413045564280998e-05,
      "loss": 1.2017,
      "step": 413
    },
    {
      "epoch": 3.106941838649156,
      "grad_norm": 0.21663722395896912,
      "learning_rate": 9.360294209010923e-05,
      "loss": 0.9858,
      "step": 414
    },
    {
      "epoch": 3.1144465290806753,
      "grad_norm": 0.21184810996055603,
      "learning_rate": 9.307563133148767e-05,
      "loss": 1.0184,
      "step": 415
    },
    {
      "epoch": 3.1219512195121952,
      "grad_norm": 0.22375060617923737,
      "learning_rate": 9.254853782343616e-05,
      "loss": 1.1505,
      "step": 416
    },
    {
      "epoch": 3.1294559099437147,
      "grad_norm": 0.24047525227069855,
      "learning_rate": 9.202167601648942e-05,
      "loss": 1.0803,
      "step": 417
    },
    {
      "epoch": 3.1369606003752346,
      "grad_norm": 0.2374037206172943,
      "learning_rate": 9.149506035483005e-05,
      "loss": 1.0936,
      "step": 418
    },
    {
      "epoch": 3.144465290806754,
      "grad_norm": 0.21291612088680267,
      "learning_rate": 9.096870527589248e-05,
      "loss": 1.037,
      "step": 419
    },
    {
      "epoch": 3.151969981238274,
      "grad_norm": 0.2159591168165207,
      "learning_rate": 9.044262520996702e-05,
      "loss": 1.1543,
      "step": 420
    },
    {
      "epoch": 3.1594746716697935,
      "grad_norm": 0.21883559226989746,
      "learning_rate": 8.991683457980443e-05,
      "loss": 1.0408,
      "step": 421
    },
    {
      "epoch": 3.1669793621013134,
      "grad_norm": 0.2221754789352417,
      "learning_rate": 8.93913478002205e-05,
      "loss": 0.9817,
      "step": 422
    },
    {
      "epoch": 3.174484052532833,
      "grad_norm": 0.20712389051914215,
      "learning_rate": 8.886617927770065e-05,
      "loss": 1.0438,
      "step": 423
    },
    {
      "epoch": 3.1819887429643527,
      "grad_norm": 0.2215949445962906,
      "learning_rate": 8.834134341000527e-05,
      "loss": 1.0819,
      "step": 424
    },
    {
      "epoch": 3.189493433395872,
      "grad_norm": 0.21339358389377594,
      "learning_rate": 8.781685458577481e-05,
      "loss": 1.1165,
      "step": 425
    },
    {
      "epoch": 3.196998123827392,
      "grad_norm": 0.22264236211776733,
      "learning_rate": 8.729272718413527e-05,
      "loss": 1.1062,
      "step": 426
    },
    {
      "epoch": 3.2045028142589116,
      "grad_norm": 0.23997175693511963,
      "learning_rate": 8.676897557430415e-05,
      "loss": 1.1483,
      "step": 427
    },
    {
      "epoch": 3.2120075046904315,
      "grad_norm": 0.23225075006484985,
      "learning_rate": 8.624561411519644e-05,
      "loss": 1.1092,
      "step": 428
    },
    {
      "epoch": 3.2195121951219514,
      "grad_norm": 0.21948185563087463,
      "learning_rate": 8.572265715503086e-05,
      "loss": 1.0546,
      "step": 429
    },
    {
      "epoch": 3.227016885553471,
      "grad_norm": 0.21425573527812958,
      "learning_rate": 8.520011903093666e-05,
      "loss": 0.9975,
      "step": 430
    },
    {
      "epoch": 3.234521575984991,
      "grad_norm": 0.25481465458869934,
      "learning_rate": 8.467801406856054e-05,
      "loss": 1.1169,
      "step": 431
    },
    {
      "epoch": 3.2420262664165103,
      "grad_norm": 0.22400236129760742,
      "learning_rate": 8.415635658167368e-05,
      "loss": 1.145,
      "step": 432
    },
    {
      "epoch": 3.24953095684803,
      "grad_norm": 0.24496909976005554,
      "learning_rate": 8.363516087177962e-05,
      "loss": 1.1359,
      "step": 433
    },
    {
      "epoch": 3.2570356472795496,
      "grad_norm": 0.23271729052066803,
      "learning_rate": 8.31144412277221e-05,
      "loss": 1.0242,
      "step": 434
    },
    {
      "epoch": 3.2645403377110696,
      "grad_norm": 0.22434957325458527,
      "learning_rate": 8.25942119252931e-05,
      "loss": 1.127,
      "step": 435
    },
    {
      "epoch": 3.272045028142589,
      "grad_norm": 0.27049028873443604,
      "learning_rate": 8.20744872268418e-05,
      "loss": 1.2073,
      "step": 436
    },
    {
      "epoch": 3.279549718574109,
      "grad_norm": 0.24478651583194733,
      "learning_rate": 8.155528138088337e-05,
      "loss": 1.0861,
      "step": 437
    },
    {
      "epoch": 3.2870544090056284,
      "grad_norm": 0.2177598774433136,
      "learning_rate": 8.103660862170826e-05,
      "loss": 0.9586,
      "step": 438
    },
    {
      "epoch": 3.2945590994371483,
      "grad_norm": 0.23956093192100525,
      "learning_rate": 8.051848316899227e-05,
      "loss": 1.0826,
      "step": 439
    },
    {
      "epoch": 3.3020637898686678,
      "grad_norm": 0.21806806325912476,
      "learning_rate": 8.000091922740633e-05,
      "loss": 1.0262,
      "step": 440
    },
    {
      "epoch": 3.3095684803001877,
      "grad_norm": 0.22320154309272766,
      "learning_rate": 7.948393098622737e-05,
      "loss": 1.13,
      "step": 441
    },
    {
      "epoch": 3.317073170731707,
      "grad_norm": 0.23721568286418915,
      "learning_rate": 7.896753261894923e-05,
      "loss": 1.0601,
      "step": 442
    },
    {
      "epoch": 3.324577861163227,
      "grad_norm": 0.2219947874546051,
      "learning_rate": 7.845173828289392e-05,
      "loss": 0.977,
      "step": 443
    },
    {
      "epoch": 3.3320825515947465,
      "grad_norm": 0.2508564591407776,
      "learning_rate": 7.793656211882377e-05,
      "loss": 1.0355,
      "step": 444
    },
    {
      "epoch": 3.3395872420262664,
      "grad_norm": 0.2457573562860489,
      "learning_rate": 7.74220182505536e-05,
      "loss": 1.1448,
      "step": 445
    },
    {
      "epoch": 3.3470919324577864,
      "grad_norm": 0.23329044878482819,
      "learning_rate": 7.690812078456336e-05,
      "loss": 1.1191,
      "step": 446
    },
    {
      "epoch": 3.354596622889306,
      "grad_norm": 0.24624013900756836,
      "learning_rate": 7.639488380961173e-05,
      "loss": 1.1743,
      "step": 447
    },
    {
      "epoch": 3.3621013133208253,
      "grad_norm": 0.2244034856557846,
      "learning_rate": 7.588232139634968e-05,
      "loss": 0.9891,
      "step": 448
    },
    {
      "epoch": 3.369606003752345,
      "grad_norm": 0.22842642664909363,
      "learning_rate": 7.537044759693463e-05,
      "loss": 0.9862,
      "step": 449
    },
    {
      "epoch": 3.377110694183865,
      "grad_norm": 0.22759254276752472,
      "learning_rate": 7.48592764446454e-05,
      "loss": 1.0067,
      "step": 450
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 0.23476208746433258,
      "learning_rate": 7.434882195349736e-05,
      "loss": 1.1237,
      "step": 451
    },
    {
      "epoch": 3.3921200750469045,
      "grad_norm": 0.24154874682426453,
      "learning_rate": 7.383909811785817e-05,
      "loss": 1.0117,
      "step": 452
    },
    {
      "epoch": 3.399624765478424,
      "grad_norm": 0.22935400903224945,
      "learning_rate": 7.333011891206432e-05,
      "loss": 1.0591,
      "step": 453
    },
    {
      "epoch": 3.407129455909944,
      "grad_norm": 0.23829422891139984,
      "learning_rate": 7.282189829003785e-05,
      "loss": 1.224,
      "step": 454
    },
    {
      "epoch": 3.4146341463414633,
      "grad_norm": 0.2275237888097763,
      "learning_rate": 7.231445018490381e-05,
      "loss": 1.0314,
      "step": 455
    },
    {
      "epoch": 3.4221388367729833,
      "grad_norm": 0.22103069722652435,
      "learning_rate": 7.180778850860835e-05,
      "loss": 1.1062,
      "step": 456
    },
    {
      "epoch": 3.4296435272045027,
      "grad_norm": 0.2439967691898346,
      "learning_rate": 7.130192715153731e-05,
      "loss": 1.0664,
      "step": 457
    },
    {
      "epoch": 3.4371482176360226,
      "grad_norm": 0.245634064078331,
      "learning_rate": 7.079687998213526e-05,
      "loss": 0.9741,
      "step": 458
    },
    {
      "epoch": 3.444652908067542,
      "grad_norm": 0.23626022040843964,
      "learning_rate": 7.029266084652548e-05,
      "loss": 1.0694,
      "step": 459
    },
    {
      "epoch": 3.452157598499062,
      "grad_norm": 0.2408812940120697,
      "learning_rate": 6.978928356813031e-05,
      "loss": 1.0809,
      "step": 460
    },
    {
      "epoch": 3.4596622889305815,
      "grad_norm": 0.23747210204601288,
      "learning_rate": 6.92867619472921e-05,
      "loss": 1.0981,
      "step": 461
    },
    {
      "epoch": 3.4671669793621014,
      "grad_norm": 0.2686862349510193,
      "learning_rate": 6.878510976089493e-05,
      "loss": 1.1492,
      "step": 462
    },
    {
      "epoch": 3.474671669793621,
      "grad_norm": 0.21842597424983978,
      "learning_rate": 6.828434076198693e-05,
      "loss": 0.9768,
      "step": 463
    },
    {
      "epoch": 3.4821763602251408,
      "grad_norm": 0.23978471755981445,
      "learning_rate": 6.77844686794031e-05,
      "loss": 1.1168,
      "step": 464
    },
    {
      "epoch": 3.4896810506566602,
      "grad_norm": 0.2502591609954834,
      "learning_rate": 6.728550721738915e-05,
      "loss": 1.1756,
      "step": 465
    },
    {
      "epoch": 3.49718574108818,
      "grad_norm": 0.2414221614599228,
      "learning_rate": 6.678747005522557e-05,
      "loss": 1.1609,
      "step": 466
    },
    {
      "epoch": 3.5046904315197,
      "grad_norm": 0.24773305654525757,
      "learning_rate": 6.629037084685278e-05,
      "loss": 0.9935,
      "step": 467
    },
    {
      "epoch": 3.5121951219512195,
      "grad_norm": 0.22421503067016602,
      "learning_rate": 6.579422322049668e-05,
      "loss": 0.9635,
      "step": 468
    },
    {
      "epoch": 3.519699812382739,
      "grad_norm": 0.22202187776565552,
      "learning_rate": 6.529904077829505e-05,
      "loss": 1.0198,
      "step": 469
    },
    {
      "epoch": 3.527204502814259,
      "grad_norm": 0.23719052970409393,
      "learning_rate": 6.480483709592468e-05,
      "loss": 1.0361,
      "step": 470
    },
    {
      "epoch": 3.534709193245779,
      "grad_norm": 0.25018665194511414,
      "learning_rate": 6.43116257222292e-05,
      "loss": 1.1185,
      "step": 471
    },
    {
      "epoch": 3.5422138836772983,
      "grad_norm": 0.241459459066391,
      "learning_rate": 6.381942017884753e-05,
      "loss": 1.1042,
      "step": 472
    },
    {
      "epoch": 3.5497185741088177,
      "grad_norm": 0.25100335478782654,
      "learning_rate": 6.33282339598433e-05,
      "loss": 1.1655,
      "step": 473
    },
    {
      "epoch": 3.5572232645403377,
      "grad_norm": 0.23576097190380096,
      "learning_rate": 6.283808053133484e-05,
      "loss": 1.0164,
      "step": 474
    },
    {
      "epoch": 3.5647279549718576,
      "grad_norm": 0.25716736912727356,
      "learning_rate": 6.234897333112594e-05,
      "loss": 1.1124,
      "step": 475
    },
    {
      "epoch": 3.572232645403377,
      "grad_norm": 0.25846216082572937,
      "learning_rate": 6.186092576833761e-05,
      "loss": 1.0799,
      "step": 476
    },
    {
      "epoch": 3.579737335834897,
      "grad_norm": 0.24888475239276886,
      "learning_rate": 6.137395122304033e-05,
      "loss": 1.0889,
      "step": 477
    },
    {
      "epoch": 3.5872420262664164,
      "grad_norm": 0.2608623206615448,
      "learning_rate": 6.088806304588717e-05,
      "loss": 1.1481,
      "step": 478
    },
    {
      "epoch": 3.5947467166979363,
      "grad_norm": 0.24895603954792023,
      "learning_rate": 6.0403274557748035e-05,
      "loss": 1.0254,
      "step": 479
    },
    {
      "epoch": 3.602251407129456,
      "grad_norm": 0.23485563695430756,
      "learning_rate": 5.9919599049344194e-05,
      "loss": 1.0436,
      "step": 480
    },
    {
      "epoch": 3.6097560975609757,
      "grad_norm": 0.2389983832836151,
      "learning_rate": 5.943704978088402e-05,
      "loss": 0.9625,
      "step": 481
    },
    {
      "epoch": 3.617260787992495,
      "grad_norm": 0.27799031138420105,
      "learning_rate": 5.89556399816995e-05,
      "loss": 1.1257,
      "step": 482
    },
    {
      "epoch": 3.624765478424015,
      "grad_norm": 0.24474041163921356,
      "learning_rate": 5.847538284988341e-05,
      "loss": 1.0092,
      "step": 483
    },
    {
      "epoch": 3.6322701688555346,
      "grad_norm": 0.2799815833568573,
      "learning_rate": 5.7996291551927666e-05,
      "loss": 1.1885,
      "step": 484
    },
    {
      "epoch": 3.6397748592870545,
      "grad_norm": 0.26766714453697205,
      "learning_rate": 5.751837922236217e-05,
      "loss": 1.0662,
      "step": 485
    },
    {
      "epoch": 3.647279549718574,
      "grad_norm": 0.24683302640914917,
      "learning_rate": 5.704165896339494e-05,
      "loss": 1.1489,
      "step": 486
    },
    {
      "epoch": 3.654784240150094,
      "grad_norm": 0.2518823444843292,
      "learning_rate": 5.656614384455257e-05,
      "loss": 1.1066,
      "step": 487
    },
    {
      "epoch": 3.6622889305816138,
      "grad_norm": 0.2337695062160492,
      "learning_rate": 5.609184690232235e-05,
      "loss": 1.0152,
      "step": 488
    },
    {
      "epoch": 3.6697936210131332,
      "grad_norm": 0.23957422375679016,
      "learning_rate": 5.5618781139794465e-05,
      "loss": 0.9883,
      "step": 489
    },
    {
      "epoch": 3.6772983114446527,
      "grad_norm": 0.2578599154949188,
      "learning_rate": 5.514695952630578e-05,
      "loss": 1.0424,
      "step": 490
    },
    {
      "epoch": 3.6848030018761726,
      "grad_norm": 0.27467238903045654,
      "learning_rate": 5.467639499708423e-05,
      "loss": 1.1804,
      "step": 491
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.25902947783470154,
      "learning_rate": 5.420710045289399e-05,
      "loss": 1.1015,
      "step": 492
    },
    {
      "epoch": 3.699812382739212,
      "grad_norm": 0.2409105747938156,
      "learning_rate": 5.373908875968211e-05,
      "loss": 0.9903,
      "step": 493
    },
    {
      "epoch": 3.7073170731707314,
      "grad_norm": 0.2591743767261505,
      "learning_rate": 5.3272372748225556e-05,
      "loss": 1.0643,
      "step": 494
    },
    {
      "epoch": 3.7148217636022514,
      "grad_norm": 0.2520532011985779,
      "learning_rate": 5.2806965213779544e-05,
      "loss": 1.0095,
      "step": 495
    },
    {
      "epoch": 3.7223264540337713,
      "grad_norm": 0.2366853654384613,
      "learning_rate": 5.234287891572674e-05,
      "loss": 1.079,
      "step": 496
    },
    {
      "epoch": 3.7298311444652907,
      "grad_norm": 0.24774064123630524,
      "learning_rate": 5.1880126577227464e-05,
      "loss": 1.1007,
      "step": 497
    },
    {
      "epoch": 3.7373358348968106,
      "grad_norm": 0.2352713793516159,
      "learning_rate": 5.141872088487078e-05,
      "loss": 1.0291,
      "step": 498
    },
    {
      "epoch": 3.74484052532833,
      "grad_norm": 0.22620277106761932,
      "learning_rate": 5.095867448832683e-05,
      "loss": 0.9753,
      "step": 499
    },
    {
      "epoch": 3.75234521575985,
      "grad_norm": 0.24645531177520752,
      "learning_rate": 5.050000000000002e-05,
      "loss": 1.1673,
      "step": 500
    },
    {
      "epoch": 3.7598499061913695,
      "grad_norm": 0.27812206745147705,
      "learning_rate": 5.004270999468307e-05,
      "loss": 1.1147,
      "step": 501
    },
    {
      "epoch": 3.7673545966228894,
      "grad_norm": 0.24802781641483307,
      "learning_rate": 4.95868170092125e-05,
      "loss": 1.1097,
      "step": 502
    },
    {
      "epoch": 3.774859287054409,
      "grad_norm": 0.24411192536354065,
      "learning_rate": 4.913233354212485e-05,
      "loss": 0.8716,
      "step": 503
    },
    {
      "epoch": 3.782363977485929,
      "grad_norm": 0.24518345296382904,
      "learning_rate": 4.867927205331386e-05,
      "loss": 0.9773,
      "step": 504
    },
    {
      "epoch": 3.7898686679174483,
      "grad_norm": 0.2709593176841736,
      "learning_rate": 4.822764496368917e-05,
      "loss": 1.0717,
      "step": 505
    },
    {
      "epoch": 3.797373358348968,
      "grad_norm": 0.2512838840484619,
      "learning_rate": 4.7777464654835564e-05,
      "loss": 1.0045,
      "step": 506
    },
    {
      "epoch": 3.8048780487804876,
      "grad_norm": 0.24504464864730835,
      "learning_rate": 4.732874346867362e-05,
      "loss": 1.1807,
      "step": 507
    },
    {
      "epoch": 3.8123827392120075,
      "grad_norm": 0.233612060546875,
      "learning_rate": 4.6881493707121315e-05,
      "loss": 1.015,
      "step": 508
    },
    {
      "epoch": 3.8198874296435275,
      "grad_norm": 0.2554488778114319,
      "learning_rate": 4.643572763175684e-05,
      "loss": 1.0611,
      "step": 509
    },
    {
      "epoch": 3.827392120075047,
      "grad_norm": 0.2590954005718231,
      "learning_rate": 4.5991457463482264e-05,
      "loss": 1.0556,
      "step": 510
    },
    {
      "epoch": 3.8348968105065664,
      "grad_norm": 0.2538789212703705,
      "learning_rate": 4.554869538218868e-05,
      "loss": 1.0193,
      "step": 511
    },
    {
      "epoch": 3.8424015009380863,
      "grad_norm": 0.24614295363426208,
      "learning_rate": 4.5107453526422255e-05,
      "loss": 1.0238,
      "step": 512
    },
    {
      "epoch": 3.849906191369606,
      "grad_norm": 0.24469436705112457,
      "learning_rate": 4.46677439930513e-05,
      "loss": 1.0548,
      "step": 513
    },
    {
      "epoch": 3.8574108818011257,
      "grad_norm": 0.25107303261756897,
      "learning_rate": 4.422957883693483e-05,
      "loss": 0.9978,
      "step": 514
    },
    {
      "epoch": 3.864915572232645,
      "grad_norm": 0.2393045276403427,
      "learning_rate": 4.3792970070591906e-05,
      "loss": 0.949,
      "step": 515
    },
    {
      "epoch": 3.872420262664165,
      "grad_norm": 0.26291531324386597,
      "learning_rate": 4.3357929663872406e-05,
      "loss": 1.4184,
      "step": 516
    },
    {
      "epoch": 3.879924953095685,
      "grad_norm": 0.2205982357263565,
      "learning_rate": 4.29244695436289e-05,
      "loss": 1.058,
      "step": 517
    },
    {
      "epoch": 3.8874296435272044,
      "grad_norm": 0.21487586200237274,
      "learning_rate": 4.249260159338946e-05,
      "loss": 0.9027,
      "step": 518
    },
    {
      "epoch": 3.8949343339587243,
      "grad_norm": 0.2405632734298706,
      "learning_rate": 4.2062337653032146e-05,
      "loss": 1.0681,
      "step": 519
    },
    {
      "epoch": 3.902439024390244,
      "grad_norm": 0.2456521987915039,
      "learning_rate": 4.1633689518460225e-05,
      "loss": 1.1246,
      "step": 520
    },
    {
      "epoch": 3.9099437148217637,
      "grad_norm": 0.24548792839050293,
      "learning_rate": 4.1206668941278826e-05,
      "loss": 1.1283,
      "step": 521
    },
    {
      "epoch": 3.917448405253283,
      "grad_norm": 0.24099896848201752,
      "learning_rate": 4.078128762847279e-05,
      "loss": 1.1552,
      "step": 522
    },
    {
      "epoch": 3.924953095684803,
      "grad_norm": 0.24163898825645447,
      "learning_rate": 4.035755724208573e-05,
      "loss": 1.2333,
      "step": 523
    },
    {
      "epoch": 3.9324577861163226,
      "grad_norm": 0.22314009070396423,
      "learning_rate": 3.9935489398900145e-05,
      "loss": 0.9311,
      "step": 524
    },
    {
      "epoch": 3.9399624765478425,
      "grad_norm": 0.25609979033470154,
      "learning_rate": 3.951509567011922e-05,
      "loss": 1.0701,
      "step": 525
    },
    {
      "epoch": 3.947467166979362,
      "grad_norm": 0.2479907125234604,
      "learning_rate": 3.90963875810494e-05,
      "loss": 1.0864,
      "step": 526
    },
    {
      "epoch": 3.954971857410882,
      "grad_norm": 0.27016589045524597,
      "learning_rate": 3.86793766107844e-05,
      "loss": 1.084,
      "step": 527
    },
    {
      "epoch": 3.9624765478424013,
      "grad_norm": 0.2595582604408264,
      "learning_rate": 3.826407419189066e-05,
      "loss": 0.9904,
      "step": 528
    },
    {
      "epoch": 3.9699812382739212,
      "grad_norm": 0.2425074577331543,
      "learning_rate": 3.785049171009381e-05,
      "loss": 1.0076,
      "step": 529
    },
    {
      "epoch": 3.977485928705441,
      "grad_norm": 0.24700041115283966,
      "learning_rate": 3.743864050396644e-05,
      "loss": 1.0935,
      "step": 530
    },
    {
      "epoch": 3.9849906191369606,
      "grad_norm": 0.2440115362405777,
      "learning_rate": 3.7028531864617444e-05,
      "loss": 1.0667,
      "step": 531
    },
    {
      "epoch": 3.99249530956848,
      "grad_norm": 0.263894647359848,
      "learning_rate": 3.662017703538234e-05,
      "loss": 1.2119,
      "step": 532
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5553321242332458,
      "learning_rate": 3.621358721151505e-05,
      "loss": 1.8768,
      "step": 533
    },
    {
      "epoch": 4.00750469043152,
      "grad_norm": 0.23720650374889374,
      "learning_rate": 3.5808773539880973e-05,
      "loss": 1.0723,
      "step": 534
    },
    {
      "epoch": 4.01500938086304,
      "grad_norm": 0.24943062663078308,
      "learning_rate": 3.540574711865146e-05,
      "loss": 0.9952,
      "step": 535
    },
    {
      "epoch": 4.022514071294559,
      "grad_norm": 0.25107982754707336,
      "learning_rate": 3.500451899699935e-05,
      "loss": 1.0325,
      "step": 536
    },
    {
      "epoch": 4.030018761726079,
      "grad_norm": 0.2480771690607071,
      "learning_rate": 3.460510017479631e-05,
      "loss": 1.0892,
      "step": 537
    },
    {
      "epoch": 4.037523452157599,
      "grad_norm": 0.2446199208498001,
      "learning_rate": 3.420750160231118e-05,
      "loss": 1.1082,
      "step": 538
    },
    {
      "epoch": 4.045028142589119,
      "grad_norm": 0.2848975956439972,
      "learning_rate": 3.381173417990957e-05,
      "loss": 1.0322,
      "step": 539
    },
    {
      "epoch": 4.052532833020638,
      "grad_norm": 0.2551771104335785,
      "learning_rate": 3.3417808757755355e-05,
      "loss": 1.029,
      "step": 540
    },
    {
      "epoch": 4.0600375234521575,
      "grad_norm": 0.2457280158996582,
      "learning_rate": 3.302573613551292e-05,
      "loss": 1.0687,
      "step": 541
    },
    {
      "epoch": 4.067542213883677,
      "grad_norm": 0.23483915627002716,
      "learning_rate": 3.263552706205128e-05,
      "loss": 1.0712,
      "step": 542
    },
    {
      "epoch": 4.075046904315197,
      "grad_norm": 0.2625795602798462,
      "learning_rate": 3.22471922351493e-05,
      "loss": 1.0424,
      "step": 543
    },
    {
      "epoch": 4.082551594746716,
      "grad_norm": 0.25196078419685364,
      "learning_rate": 3.186074230120244e-05,
      "loss": 0.9866,
      "step": 544
    },
    {
      "epoch": 4.090056285178236,
      "grad_norm": 0.26508569717407227,
      "learning_rate": 3.147618785493083e-05,
      "loss": 1.0696,
      "step": 545
    },
    {
      "epoch": 4.097560975609756,
      "grad_norm": 0.26318731904029846,
      "learning_rate": 3.109353943908893e-05,
      "loss": 1.1073,
      "step": 546
    },
    {
      "epoch": 4.105065666041276,
      "grad_norm": 0.2586771249771118,
      "learning_rate": 3.071280754417626e-05,
      "loss": 1.0968,
      "step": 547
    },
    {
      "epoch": 4.112570356472795,
      "grad_norm": 0.26417478919029236,
      "learning_rate": 3.033400260815008e-05,
      "loss": 1.0892,
      "step": 548
    },
    {
      "epoch": 4.120075046904315,
      "grad_norm": 0.25766685605049133,
      "learning_rate": 2.9957135016139122e-05,
      "loss": 0.9827,
      "step": 549
    },
    {
      "epoch": 4.127579737335835,
      "grad_norm": 0.2724906802177429,
      "learning_rate": 2.9582215100158706e-05,
      "loss": 1.1577,
      "step": 550
    },
    {
      "epoch": 4.135084427767355,
      "grad_norm": 0.24692752957344055,
      "learning_rate": 2.920925313882776e-05,
      "loss": 1.1081,
      "step": 551
    },
    {
      "epoch": 4.142589118198874,
      "grad_norm": 0.27086886763572693,
      "learning_rate": 2.8838259357086884e-05,
      "loss": 0.9835,
      "step": 552
    },
    {
      "epoch": 4.150093808630394,
      "grad_norm": 0.25520050525665283,
      "learning_rate": 2.846924392591794e-05,
      "loss": 1.0608,
      "step": 553
    },
    {
      "epoch": 4.157598499061914,
      "grad_norm": 0.2637060582637787,
      "learning_rate": 2.8102216962065423e-05,
      "loss": 0.9214,
      "step": 554
    },
    {
      "epoch": 4.165103189493434,
      "grad_norm": 0.2559458613395691,
      "learning_rate": 2.7737188527758972e-05,
      "loss": 0.9785,
      "step": 555
    },
    {
      "epoch": 4.1726078799249535,
      "grad_norm": 0.2692560851573944,
      "learning_rate": 2.7374168630437456e-05,
      "loss": 1.1126,
      "step": 556
    },
    {
      "epoch": 4.1801125703564725,
      "grad_norm": 0.26712581515312195,
      "learning_rate": 2.7013167222474756e-05,
      "loss": 0.9964,
      "step": 557
    },
    {
      "epoch": 4.1876172607879925,
      "grad_norm": 0.269401878118515,
      "learning_rate": 2.6654194200906833e-05,
      "loss": 1.0647,
      "step": 558
    },
    {
      "epoch": 4.195121951219512,
      "grad_norm": 0.2693142294883728,
      "learning_rate": 2.629725940716041e-05,
      "loss": 1.0686,
      "step": 559
    },
    {
      "epoch": 4.202626641651032,
      "grad_norm": 0.23360580205917358,
      "learning_rate": 2.5942372626783172e-05,
      "loss": 0.9899,
      "step": 560
    },
    {
      "epoch": 4.210131332082551,
      "grad_norm": 0.27039438486099243,
      "learning_rate": 2.5589543589175485e-05,
      "loss": 1.1598,
      "step": 561
    },
    {
      "epoch": 4.217636022514071,
      "grad_norm": 0.2767198979854584,
      "learning_rate": 2.523878196732358e-05,
      "loss": 0.9577,
      "step": 562
    },
    {
      "epoch": 4.225140712945591,
      "grad_norm": 0.27451032400131226,
      "learning_rate": 2.489009737753459e-05,
      "loss": 1.2496,
      "step": 563
    },
    {
      "epoch": 4.232645403377111,
      "grad_norm": 0.24222443997859955,
      "learning_rate": 2.4543499379172615e-05,
      "loss": 1.018,
      "step": 564
    },
    {
      "epoch": 4.24015009380863,
      "grad_norm": 0.26674020290374756,
      "learning_rate": 2.4198997474396877e-05,
      "loss": 1.0977,
      "step": 565
    },
    {
      "epoch": 4.24765478424015,
      "grad_norm": 0.2519540786743164,
      "learning_rate": 2.3856601107901166e-05,
      "loss": 1.0182,
      "step": 566
    },
    {
      "epoch": 4.25515947467167,
      "grad_norm": 0.2805282771587372,
      "learning_rate": 2.351631966665476e-05,
      "loss": 0.9412,
      "step": 567
    },
    {
      "epoch": 4.26266416510319,
      "grad_norm": 0.25769346952438354,
      "learning_rate": 2.31781624796453e-05,
      "loss": 1.0087,
      "step": 568
    },
    {
      "epoch": 4.270168855534709,
      "grad_norm": 0.25440049171447754,
      "learning_rate": 2.2842138817622883e-05,
      "loss": 1.0862,
      "step": 569
    },
    {
      "epoch": 4.277673545966229,
      "grad_norm": 0.25586551427841187,
      "learning_rate": 2.250825789284594e-05,
      "loss": 1.0173,
      "step": 570
    },
    {
      "epoch": 4.285178236397749,
      "grad_norm": 0.2734562158584595,
      "learning_rate": 2.217652885882869e-05,
      "loss": 1.0565,
      "step": 571
    },
    {
      "epoch": 4.2926829268292686,
      "grad_norm": 0.2658538818359375,
      "learning_rate": 2.1846960810090188e-05,
      "loss": 1.0235,
      "step": 572
    },
    {
      "epoch": 4.300187617260788,
      "grad_norm": 0.2710430920124054,
      "learning_rate": 2.151956278190494e-05,
      "loss": 1.0054,
      "step": 573
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 0.2484230250120163,
      "learning_rate": 2.119434375005527e-05,
      "loss": 0.9814,
      "step": 574
    },
    {
      "epoch": 4.315196998123827,
      "grad_norm": 0.27289608120918274,
      "learning_rate": 2.087131263058526e-05,
      "loss": 1.0434,
      "step": 575
    },
    {
      "epoch": 4.322701688555347,
      "grad_norm": 0.25142380595207214,
      "learning_rate": 2.055047827955618e-05,
      "loss": 0.996,
      "step": 576
    },
    {
      "epoch": 4.330206378986867,
      "grad_norm": 0.281368225812912,
      "learning_rate": 2.0231849492803852e-05,
      "loss": 1.0575,
      "step": 577
    },
    {
      "epoch": 4.337711069418386,
      "grad_norm": 0.28785380721092224,
      "learning_rate": 1.991543500569745e-05,
      "loss": 1.0859,
      "step": 578
    },
    {
      "epoch": 4.345215759849906,
      "grad_norm": 0.28092002868652344,
      "learning_rate": 1.960124349289992e-05,
      "loss": 1.1349,
      "step": 579
    },
    {
      "epoch": 4.352720450281426,
      "grad_norm": 0.2787632644176483,
      "learning_rate": 1.928928356813032e-05,
      "loss": 1.0366,
      "step": 580
    },
    {
      "epoch": 4.360225140712946,
      "grad_norm": 0.2454954981803894,
      "learning_rate": 1.8979563783927565e-05,
      "loss": 0.9675,
      "step": 581
    },
    {
      "epoch": 4.367729831144465,
      "grad_norm": 0.2587539851665497,
      "learning_rate": 1.8672092631416013e-05,
      "loss": 1.0678,
      "step": 582
    },
    {
      "epoch": 4.375234521575985,
      "grad_norm": 0.2558777928352356,
      "learning_rate": 1.8366878540072614e-05,
      "loss": 0.9883,
      "step": 583
    },
    {
      "epoch": 4.382739212007505,
      "grad_norm": 0.2781766355037689,
      "learning_rate": 1.8063929877495892e-05,
      "loss": 0.9614,
      "step": 584
    },
    {
      "epoch": 4.390243902439025,
      "grad_norm": 0.3783913254737854,
      "learning_rate": 1.7763254949176414e-05,
      "loss": 1.0883,
      "step": 585
    },
    {
      "epoch": 4.397748592870544,
      "grad_norm": 0.25674161314964294,
      "learning_rate": 1.7464861998269243e-05,
      "loss": 1.03,
      "step": 586
    },
    {
      "epoch": 4.405253283302064,
      "grad_norm": 0.2544950842857361,
      "learning_rate": 1.7168759205367893e-05,
      "loss": 1.0044,
      "step": 587
    },
    {
      "epoch": 4.412757973733584,
      "grad_norm": 0.2635541558265686,
      "learning_rate": 1.6874954688279956e-05,
      "loss": 0.9767,
      "step": 588
    },
    {
      "epoch": 4.4202626641651035,
      "grad_norm": 0.27090808749198914,
      "learning_rate": 1.6583456501804725e-05,
      "loss": 0.9826,
      "step": 589
    },
    {
      "epoch": 4.4277673545966225,
      "grad_norm": 0.2597501277923584,
      "learning_rate": 1.6294272637512183e-05,
      "loss": 1.0187,
      "step": 590
    },
    {
      "epoch": 4.435272045028142,
      "grad_norm": 0.28622573614120483,
      "learning_rate": 1.600741102352409e-05,
      "loss": 1.0623,
      "step": 591
    },
    {
      "epoch": 4.442776735459662,
      "grad_norm": 0.25954049825668335,
      "learning_rate": 1.57228795242965e-05,
      "loss": 1.0268,
      "step": 592
    },
    {
      "epoch": 4.450281425891182,
      "grad_norm": 0.2589193284511566,
      "learning_rate": 1.544068594040417e-05,
      "loss": 0.9672,
      "step": 593
    },
    {
      "epoch": 4.457786116322701,
      "grad_norm": 0.2728925943374634,
      "learning_rate": 1.516083800832676e-05,
      "loss": 1.0571,
      "step": 594
    },
    {
      "epoch": 4.465290806754221,
      "grad_norm": 0.2503071129322052,
      "learning_rate": 1.488334340023669e-05,
      "loss": 1.0172,
      "step": 595
    },
    {
      "epoch": 4.472795497185741,
      "grad_norm": 0.2626873552799225,
      "learning_rate": 1.4608209723788835e-05,
      "loss": 1.036,
      "step": 596
    },
    {
      "epoch": 4.480300187617261,
      "grad_norm": 0.3006345331668854,
      "learning_rate": 1.4335444521911899e-05,
      "loss": 1.1155,
      "step": 597
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 0.2929830849170685,
      "learning_rate": 1.4065055272601703e-05,
      "loss": 1.0768,
      "step": 598
    },
    {
      "epoch": 4.4953095684803,
      "grad_norm": 0.25911423563957214,
      "learning_rate": 1.3797049388716065e-05,
      "loss": 1.0368,
      "step": 599
    },
    {
      "epoch": 4.50281425891182,
      "grad_norm": 0.2785639464855194,
      "learning_rate": 1.3531434217771692e-05,
      "loss": 1.0747,
      "step": 600
    },
    {
      "epoch": 4.51031894934334,
      "grad_norm": 0.27841538190841675,
      "learning_rate": 1.3268217041742701e-05,
      "loss": 1.0116,
      "step": 601
    },
    {
      "epoch": 4.517823639774859,
      "grad_norm": 0.27080854773521423,
      "learning_rate": 1.3007405076860875e-05,
      "loss": 0.9667,
      "step": 602
    },
    {
      "epoch": 4.525328330206379,
      "grad_norm": 0.25426366925239563,
      "learning_rate": 1.2749005473418015e-05,
      "loss": 0.9367,
      "step": 603
    },
    {
      "epoch": 4.532833020637899,
      "grad_norm": 0.24661357700824738,
      "learning_rate": 1.2493025315569801e-05,
      "loss": 0.9841,
      "step": 604
    },
    {
      "epoch": 4.5403377110694185,
      "grad_norm": 0.2595573365688324,
      "learning_rate": 1.2239471621141508e-05,
      "loss": 0.9494,
      "step": 605
    },
    {
      "epoch": 4.547842401500938,
      "grad_norm": 0.27605584263801575,
      "learning_rate": 1.1988351341435792e-05,
      "loss": 1.0487,
      "step": 606
    },
    {
      "epoch": 4.5553470919324575,
      "grad_norm": 0.2632475197315216,
      "learning_rate": 1.173967136104196e-05,
      "loss": 1.0948,
      "step": 607
    },
    {
      "epoch": 4.562851782363977,
      "grad_norm": 0.2705388367176056,
      "learning_rate": 1.1493438497647313e-05,
      "loss": 1.0667,
      "step": 608
    },
    {
      "epoch": 4.570356472795497,
      "grad_norm": 0.25751256942749023,
      "learning_rate": 1.1249659501850155e-05,
      "loss": 1.1783,
      "step": 609
    },
    {
      "epoch": 4.577861163227017,
      "grad_norm": 0.2475949227809906,
      "learning_rate": 1.1008341056974854e-05,
      "loss": 1.0667,
      "step": 610
    },
    {
      "epoch": 4.585365853658536,
      "grad_norm": 0.26045867800712585,
      "learning_rate": 1.0769489778888405e-05,
      "loss": 0.9714,
      "step": 611
    },
    {
      "epoch": 4.592870544090056,
      "grad_norm": 0.2527018189430237,
      "learning_rate": 1.0533112215819298e-05,
      "loss": 1.0692,
      "step": 612
    },
    {
      "epoch": 4.600375234521576,
      "grad_norm": 0.25061407685279846,
      "learning_rate": 1.029921484817783e-05,
      "loss": 0.9363,
      "step": 613
    },
    {
      "epoch": 4.607879924953096,
      "grad_norm": 0.24939140677452087,
      "learning_rate": 1.0067804088378455e-05,
      "loss": 1.038,
      "step": 614
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.2591976523399353,
      "learning_rate": 9.8388862806641e-06,
      "loss": 0.9335,
      "step": 615
    },
    {
      "epoch": 4.622889305816135,
      "grad_norm": 0.26032814383506775,
      "learning_rate": 9.612467700932045e-06,
      "loss": 1.0562,
      "step": 616
    },
    {
      "epoch": 4.630393996247655,
      "grad_norm": 0.2716333270072937,
      "learning_rate": 9.388554556562049e-06,
      "loss": 0.9728,
      "step": 617
    },
    {
      "epoch": 4.637898686679175,
      "grad_norm": 0.26602107286453247,
      "learning_rate": 9.167152986246078e-06,
      "loss": 0.9198,
      "step": 618
    },
    {
      "epoch": 4.645403377110695,
      "grad_norm": 0.27277711033821106,
      "learning_rate": 8.948269059820025e-06,
      "loss": 1.0677,
      "step": 619
    },
    {
      "epoch": 4.652908067542214,
      "grad_norm": 0.2933201491832733,
      "learning_rate": 8.731908778097302e-06,
      "loss": 1.0856,
      "step": 620
    },
    {
      "epoch": 4.6604127579737336,
      "grad_norm": 0.26761552691459656,
      "learning_rate": 8.518078072704338e-06,
      "loss": 0.9855,
      "step": 621
    },
    {
      "epoch": 4.6679174484052535,
      "grad_norm": 0.26589658856391907,
      "learning_rate": 8.306782805917904e-06,
      "loss": 1.1391,
      "step": 622
    },
    {
      "epoch": 4.6754221388367725,
      "grad_norm": 0.245664581656456,
      "learning_rate": 8.098028770504494e-06,
      "loss": 1.0441,
      "step": 623
    },
    {
      "epoch": 4.682926829268292,
      "grad_norm": 0.25962963700294495,
      "learning_rate": 7.891821689561459e-06,
      "loss": 0.9938,
      "step": 624
    },
    {
      "epoch": 4.690431519699812,
      "grad_norm": 0.2487301230430603,
      "learning_rate": 7.68816721636004e-06,
      "loss": 1.0108,
      "step": 625
    },
    {
      "epoch": 4.697936210131332,
      "grad_norm": 0.2539227604866028,
      "learning_rate": 7.487070934190532e-06,
      "loss": 1.0537,
      "step": 626
    },
    {
      "epoch": 4.705440900562852,
      "grad_norm": 0.2824779152870178,
      "learning_rate": 7.288538356209092e-06,
      "loss": 1.0141,
      "step": 627
    },
    {
      "epoch": 4.712945590994371,
      "grad_norm": 0.24700817465782166,
      "learning_rate": 7.092574925286614e-06,
      "loss": 0.8957,
      "step": 628
    },
    {
      "epoch": 4.720450281425891,
      "grad_norm": 0.23364083468914032,
      "learning_rate": 6.899186013859561e-06,
      "loss": 1.0811,
      "step": 629
    },
    {
      "epoch": 4.727954971857411,
      "grad_norm": 0.25310617685317993,
      "learning_rate": 6.708376923782635e-06,
      "loss": 1.0577,
      "step": 630
    },
    {
      "epoch": 4.735459662288931,
      "grad_norm": 0.2497323900461197,
      "learning_rate": 6.520152886183406e-06,
      "loss": 0.9827,
      "step": 631
    },
    {
      "epoch": 4.74296435272045,
      "grad_norm": 0.27253395318984985,
      "learning_rate": 6.3345190613189635e-06,
      "loss": 0.963,
      "step": 632
    },
    {
      "epoch": 4.75046904315197,
      "grad_norm": 0.25061896443367004,
      "learning_rate": 6.151480538434382e-06,
      "loss": 1.1357,
      "step": 633
    },
    {
      "epoch": 4.75797373358349,
      "grad_norm": 0.25422045588493347,
      "learning_rate": 5.971042335623229e-06,
      "loss": 0.9789,
      "step": 634
    },
    {
      "epoch": 4.76547842401501,
      "grad_norm": 0.28153881430625916,
      "learning_rate": 5.793209399689978e-06,
      "loss": 1.0069,
      "step": 635
    },
    {
      "epoch": 4.772983114446529,
      "grad_norm": 0.26730191707611084,
      "learning_rate": 5.617986606014419e-06,
      "loss": 0.9413,
      "step": 636
    },
    {
      "epoch": 4.780487804878049,
      "grad_norm": 0.25754016637802124,
      "learning_rate": 5.445378758417925e-06,
      "loss": 0.967,
      "step": 637
    },
    {
      "epoch": 4.7879924953095685,
      "grad_norm": 0.26099568605422974,
      "learning_rate": 5.275390589031859e-06,
      "loss": 1.0104,
      "step": 638
    },
    {
      "epoch": 4.795497185741088,
      "grad_norm": 0.27538684010505676,
      "learning_rate": 5.108026758167719e-06,
      "loss": 1.0233,
      "step": 639
    },
    {
      "epoch": 4.803001876172608,
      "grad_norm": 0.26298317313194275,
      "learning_rate": 4.943291854189493e-06,
      "loss": 0.9756,
      "step": 640
    },
    {
      "epoch": 4.810506566604127,
      "grad_norm": 0.2920237183570862,
      "learning_rate": 4.781190393387796e-06,
      "loss": 0.985,
      "step": 641
    },
    {
      "epoch": 4.818011257035647,
      "grad_norm": 0.2614547908306122,
      "learning_rate": 4.6217268198560404e-06,
      "loss": 1.0421,
      "step": 642
    },
    {
      "epoch": 4.825515947467167,
      "grad_norm": 0.2709522843360901,
      "learning_rate": 4.464905505368658e-06,
      "loss": 1.0479,
      "step": 643
    },
    {
      "epoch": 4.833020637898686,
      "grad_norm": 0.2588852047920227,
      "learning_rate": 4.3107307492612086e-06,
      "loss": 0.8586,
      "step": 644
    },
    {
      "epoch": 4.840525328330206,
      "grad_norm": 0.2764037549495697,
      "learning_rate": 4.1592067783125015e-06,
      "loss": 1.0139,
      "step": 645
    },
    {
      "epoch": 4.848030018761726,
      "grad_norm": 0.2646092474460602,
      "learning_rate": 4.010337746628751e-06,
      "loss": 0.8966,
      "step": 646
    },
    {
      "epoch": 4.855534709193246,
      "grad_norm": 0.26426824927330017,
      "learning_rate": 3.864127735529656e-06,
      "loss": 1.0248,
      "step": 647
    },
    {
      "epoch": 4.863039399624766,
      "grad_norm": 0.2680165767669678,
      "learning_rate": 3.7205807534365315e-06,
      "loss": 1.002,
      "step": 648
    },
    {
      "epoch": 4.870544090056285,
      "grad_norm": 0.27381497621536255,
      "learning_rate": 3.5797007357623945e-06,
      "loss": 1.005,
      "step": 649
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 0.24401965737342834,
      "learning_rate": 3.441491544804112e-06,
      "loss": 0.9504,
      "step": 650
    },
    {
      "epoch": 4.885553470919325,
      "grad_norm": 0.28096142411231995,
      "learning_rate": 3.3059569696364502e-06,
      "loss": 1.0629,
      "step": 651
    },
    {
      "epoch": 4.893058161350845,
      "grad_norm": 0.25856783986091614,
      "learning_rate": 3.1731007260082616e-06,
      "loss": 1.0072,
      "step": 652
    },
    {
      "epoch": 4.900562851782364,
      "grad_norm": 0.25341498851776123,
      "learning_rate": 3.0429264562405776e-06,
      "loss": 1.0113,
      "step": 653
    },
    {
      "epoch": 4.9080675422138835,
      "grad_norm": 0.2537418007850647,
      "learning_rate": 2.9154377291267674e-06,
      "loss": 1.0907,
      "step": 654
    },
    {
      "epoch": 4.915572232645403,
      "grad_norm": 0.252174973487854,
      "learning_rate": 2.790638039834668e-06,
      "loss": 0.9687,
      "step": 655
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 0.26291623711586,
      "learning_rate": 2.6685308098108106e-06,
      "loss": 1.0686,
      "step": 656
    },
    {
      "epoch": 4.930581613508442,
      "grad_norm": 0.2872166037559509,
      "learning_rate": 2.5491193866866025e-06,
      "loss": 0.9811,
      "step": 657
    },
    {
      "epoch": 4.938086303939962,
      "grad_norm": 0.28239676356315613,
      "learning_rate": 2.432407044186509e-06,
      "loss": 1.0442,
      "step": 658
    },
    {
      "epoch": 4.945590994371482,
      "grad_norm": 0.2630953788757324,
      "learning_rate": 2.3183969820383735e-06,
      "loss": 0.9368,
      "step": 659
    },
    {
      "epoch": 4.953095684803002,
      "grad_norm": 0.2792162597179413,
      "learning_rate": 2.2070923258856255e-06,
      "loss": 1.1048,
      "step": 660
    },
    {
      "epoch": 4.960600375234522,
      "grad_norm": 0.2646627724170685,
      "learning_rate": 2.098496127201648e-06,
      "loss": 0.9354,
      "step": 661
    },
    {
      "epoch": 4.968105065666041,
      "grad_norm": 0.2521073818206787,
      "learning_rate": 1.992611363206103e-06,
      "loss": 1.0318,
      "step": 662
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 0.27099478244781494,
      "learning_rate": 1.889440936783242e-06,
      "loss": 1.0485,
      "step": 663
    },
    {
      "epoch": 4.983114446529081,
      "grad_norm": 0.2716807425022125,
      "learning_rate": 1.7889876764024505e-06,
      "loss": 0.9949,
      "step": 664
    },
    {
      "epoch": 4.9906191369606,
      "grad_norm": 0.2702957093715668,
      "learning_rate": 1.691254336040595e-06,
      "loss": 1.11,
      "step": 665
    },
    {
      "epoch": 4.99812382739212,
      "grad_norm": 0.2564277648925781,
      "learning_rate": 1.59624359510657e-06,
      "loss": 0.8788,
      "step": 666
    },
    {
      "epoch": 5.00562851782364,
      "grad_norm": 0.5438546538352966,
      "learning_rate": 1.5039580583678393e-06,
      "loss": 1.7329,
      "step": 667
    },
    {
      "epoch": 5.01313320825516,
      "grad_norm": 0.2500527501106262,
      "learning_rate": 1.414400255879008e-06,
      "loss": 1.012,
      "step": 668
    },
    {
      "epoch": 5.0206378986866795,
      "grad_norm": 0.2738587558269501,
      "learning_rate": 1.327572642912468e-06,
      "loss": 1.1039,
      "step": 669
    },
    {
      "epoch": 5.028142589118199,
      "grad_norm": 0.2560110092163086,
      "learning_rate": 1.2434775998910964e-06,
      "loss": 1.027,
      "step": 670
    },
    {
      "epoch": 5.0356472795497185,
      "grad_norm": 0.2611444890499115,
      "learning_rate": 1.1621174323229612e-06,
      "loss": 1.0072,
      "step": 671
    },
    {
      "epoch": 5.043151969981238,
      "grad_norm": 0.24755235016345978,
      "learning_rate": 1.0834943707381784e-06,
      "loss": 0.9297,
      "step": 672
    },
    {
      "epoch": 5.050656660412758,
      "grad_norm": 0.2472180277109146,
      "learning_rate": 1.0076105706276888e-06,
      "loss": 0.8869,
      "step": 673
    },
    {
      "epoch": 5.058161350844277,
      "grad_norm": 0.2653760313987732,
      "learning_rate": 9.344681123841967e-07,
      "loss": 0.8558,
      "step": 674
    },
    {
      "epoch": 5.065666041275797,
      "grad_norm": 0.2679085433483124,
      "learning_rate": 8.640690012451515e-07,
      "loss": 1.1758,
      "step": 675
    },
    {
      "epoch": 5.073170731707317,
      "grad_norm": 0.23908266425132751,
      "learning_rate": 7.964151672377458e-07,
      "loss": 0.8884,
      "step": 676
    },
    {
      "epoch": 5.080675422138837,
      "grad_norm": 0.2759111225605011,
      "learning_rate": 7.315084651260009e-07,
      "loss": 1.1085,
      "step": 677
    },
    {
      "epoch": 5.088180112570356,
      "grad_norm": 0.2899288833141327,
      "learning_rate": 6.69350674359959e-07,
      "loss": 1.1352,
      "step": 678
    },
    {
      "epoch": 5.095684803001876,
      "grad_norm": 0.27089551091194153,
      "learning_rate": 6.099434990268609e-07,
      "loss": 0.9914,
      "step": 679
    },
    {
      "epoch": 5.103189493433396,
      "grad_norm": 0.29111695289611816,
      "learning_rate": 5.532885678043977e-07,
      "loss": 0.971,
      "step": 680
    },
    {
      "epoch": 5.110694183864916,
      "grad_norm": 0.2626079022884369,
      "learning_rate": 4.9938743391615e-07,
      "loss": 0.9628,
      "step": 681
    },
    {
      "epoch": 5.118198874296436,
      "grad_norm": 0.27037861943244934,
      "learning_rate": 4.482415750889204e-07,
      "loss": 1.0561,
      "step": 682
    },
    {
      "epoch": 5.125703564727955,
      "grad_norm": 0.2747006118297577,
      "learning_rate": 3.998523935122772e-07,
      "loss": 1.0828,
      "step": 683
    },
    {
      "epoch": 5.133208255159475,
      "grad_norm": 0.2585286498069763,
      "learning_rate": 3.5422121580005864e-07,
      "loss": 0.9272,
      "step": 684
    },
    {
      "epoch": 5.140712945590995,
      "grad_norm": 0.2688470184803009,
      "learning_rate": 3.1134929295407564e-07,
      "loss": 1.0374,
      "step": 685
    },
    {
      "epoch": 5.1482176360225145,
      "grad_norm": 0.25770020484924316,
      "learning_rate": 2.7123780032973235e-07,
      "loss": 0.9986,
      "step": 686
    },
    {
      "epoch": 5.1557223264540335,
      "grad_norm": 0.25769320130348206,
      "learning_rate": 2.3388783760386601e-07,
      "loss": 1.0169,
      "step": 687
    },
    {
      "epoch": 5.163227016885553,
      "grad_norm": 0.29741230607032776,
      "learning_rate": 1.9930042874457254e-07,
      "loss": 1.0703,
      "step": 688
    },
    {
      "epoch": 5.170731707317073,
      "grad_norm": 0.2753787636756897,
      "learning_rate": 1.6747652198313957e-07,
      "loss": 1.0371,
      "step": 689
    },
    {
      "epoch": 5.178236397748593,
      "grad_norm": 0.2647511959075928,
      "learning_rate": 1.3841698978804285e-07,
      "loss": 1.0168,
      "step": 690
    },
    {
      "epoch": 5.185741088180112,
      "grad_norm": 0.27614572644233704,
      "learning_rate": 1.1212262884103974e-07,
      "loss": 1.0725,
      "step": 691
    },
    {
      "epoch": 5.193245778611632,
      "grad_norm": 0.2685404121875763,
      "learning_rate": 8.85941600153033e-08,
      "loss": 1.0614,
      "step": 692
    },
    {
      "epoch": 5.200750469043152,
      "grad_norm": 0.2600098252296448,
      "learning_rate": 6.783222835572055e-08,
      "loss": 1.0501,
      "step": 693
    },
    {
      "epoch": 5.208255159474672,
      "grad_norm": 0.25625452399253845,
      "learning_rate": 4.98374030611084e-08,
      "loss": 0.9857,
      "step": 694
    },
    {
      "epoch": 5.215759849906191,
      "grad_norm": 0.26039209961891174,
      "learning_rate": 3.461017746871675e-08,
      "loss": 1.017,
      "step": 695
    },
    {
      "epoch": 5.223264540337711,
      "grad_norm": 0.2616319954395294,
      "learning_rate": 2.215096904060454e-08,
      "loss": 1.0131,
      "step": 696
    },
    {
      "epoch": 5.230769230769231,
      "grad_norm": 0.264681875705719,
      "learning_rate": 1.246011935228064e-08,
      "loss": 1.0703,
      "step": 697
    },
    {
      "epoch": 5.238273921200751,
      "grad_norm": 0.26076602935791016,
      "learning_rate": 5.537894083273543e-09,
      "loss": 0.9102,
      "step": 698
    },
    {
      "epoch": 5.24577861163227,
      "grad_norm": 0.25187480449676514,
      "learning_rate": 1.384483009898796e-09,
      "loss": 0.9244,
      "step": 699
    },
    {
      "epoch": 5.25328330206379,
      "grad_norm": 0.2918677031993866,
      "learning_rate": 0.0,
      "loss": 1.1823,
      "step": 700
    }
  ],
  "logging_steps": 1,
  "max_steps": 700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5363404536186143e+18,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}