{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.4800708591674048,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00354295837023915,
      "grad_norm": 57.142765045166016,
      "learning_rate": 2.02e-06,
      "loss": 2.8154,
      "step": 1
    },
    {
      "epoch": 0.0070859167404783,
      "grad_norm": 53.29192352294922,
      "learning_rate": 4.04e-06,
      "loss": 2.8741,
      "step": 2
    },
    {
      "epoch": 0.010628875110717449,
      "grad_norm": 51.12057876586914,
      "learning_rate": 6.06e-06,
      "loss": 2.702,
      "step": 3
    },
    {
      "epoch": 0.0141718334809566,
      "grad_norm": 37.336307525634766,
      "learning_rate": 8.08e-06,
      "loss": 2.2855,
      "step": 4
    },
    {
      "epoch": 0.01771479185119575,
      "grad_norm": 27.943105697631836,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 1.797,
      "step": 5
    },
    {
      "epoch": 0.021257750221434897,
      "grad_norm": 30.679658889770508,
      "learning_rate": 1.212e-05,
      "loss": 1.4587,
      "step": 6
    },
    {
      "epoch": 0.024800708591674048,
      "grad_norm": 29.234771728515625,
      "learning_rate": 1.4140000000000002e-05,
      "loss": 0.9948,
      "step": 7
    },
    {
      "epoch": 0.0283436669619132,
      "grad_norm": 18.120752334594727,
      "learning_rate": 1.616e-05,
      "loss": 0.5268,
      "step": 8
    },
    {
      "epoch": 0.03188662533215235,
      "grad_norm": 14.929753303527832,
      "learning_rate": 1.818e-05,
      "loss": 0.3965,
      "step": 9
    },
    {
      "epoch": 0.0354295837023915,
      "grad_norm": 10.846837997436523,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.2508,
      "step": 10
    },
    {
      "epoch": 0.03897254207263064,
      "grad_norm": 6.162725448608398,
      "learning_rate": 2.222e-05,
      "loss": 0.1554,
      "step": 11
    },
    {
      "epoch": 0.042515500442869794,
      "grad_norm": 6.691348552703857,
      "learning_rate": 2.424e-05,
      "loss": 0.1236,
      "step": 12
    },
    {
      "epoch": 0.046058458813108945,
      "grad_norm": 6.757739067077637,
      "learning_rate": 2.6260000000000003e-05,
      "loss": 0.0947,
      "step": 13
    },
    {
      "epoch": 0.049601417183348095,
      "grad_norm": 9.472763061523438,
      "learning_rate": 2.8280000000000004e-05,
      "loss": 0.0859,
      "step": 14
    },
    {
      "epoch": 0.053144375553587246,
      "grad_norm": 8.504354476928711,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 0.1265,
      "step": 15
    },
    {
      "epoch": 0.0566873339238264,
      "grad_norm": 4.339751720428467,
      "learning_rate": 3.232e-05,
      "loss": 0.0734,
      "step": 16
    },
    {
      "epoch": 0.06023029229406555,
      "grad_norm": 4.536198139190674,
      "learning_rate": 3.434e-05,
      "loss": 0.0854,
      "step": 17
    },
    {
      "epoch": 0.0637732506643047,
      "grad_norm": 5.113945484161377,
      "learning_rate": 3.636e-05,
      "loss": 0.0837,
      "step": 18
    },
    {
      "epoch": 0.06731620903454384,
      "grad_norm": 3.7606422901153564,
      "learning_rate": 3.838e-05,
      "loss": 0.0588,
      "step": 19
    },
    {
      "epoch": 0.070859167404783,
      "grad_norm": 6.364375114440918,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 0.0546,
      "step": 20
    },
    {
      "epoch": 0.07440212577502214,
      "grad_norm": 4.742648601531982,
      "learning_rate": 4.242e-05,
      "loss": 0.0534,
      "step": 21
    },
    {
      "epoch": 0.07794508414526129,
      "grad_norm": 3.6039655208587646,
      "learning_rate": 4.444e-05,
      "loss": 0.04,
      "step": 22
    },
    {
      "epoch": 0.08148804251550044,
      "grad_norm": 3.1530134677886963,
      "learning_rate": 4.6460000000000006e-05,
      "loss": 0.0473,
      "step": 23
    },
    {
      "epoch": 0.08503100088573959,
      "grad_norm": 4.383698463439941,
      "learning_rate": 4.848e-05,
      "loss": 0.0476,
      "step": 24
    },
    {
      "epoch": 0.08857395925597875,
      "grad_norm": 3.4666121006011963,
      "learning_rate": 5.05e-05,
      "loss": 0.031,
      "step": 25
    },
    {
      "epoch": 0.09211691762621789,
      "grad_norm": 5.188110828399658,
      "learning_rate": 5.2520000000000005e-05,
      "loss": 0.0569,
      "step": 26
    },
    {
      "epoch": 0.09565987599645705,
      "grad_norm": 3.7114498615264893,
      "learning_rate": 5.454e-05,
      "loss": 0.0574,
      "step": 27
    },
    {
      "epoch": 0.09920283436669619,
      "grad_norm": 2.91119122505188,
      "learning_rate": 5.656000000000001e-05,
      "loss": 0.0322,
      "step": 28
    },
    {
      "epoch": 0.10274579273693533,
      "grad_norm": 3.792724609375,
      "learning_rate": 5.858e-05,
      "loss": 0.0409,
      "step": 29
    },
    {
      "epoch": 0.10628875110717449,
      "grad_norm": 3.0739586353302,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 0.0402,
      "step": 30
    },
    {
      "epoch": 0.10983170947741364,
      "grad_norm": 2.034921407699585,
      "learning_rate": 6.262000000000001e-05,
      "loss": 0.0403,
      "step": 31
    },
    {
      "epoch": 0.1133746678476528,
      "grad_norm": 2.346412181854248,
      "learning_rate": 6.464e-05,
      "loss": 0.0251,
      "step": 32
    },
    {
      "epoch": 0.11691762621789194,
      "grad_norm": 2.205854892730713,
      "learning_rate": 6.666e-05,
      "loss": 0.0369,
      "step": 33
    },
    {
      "epoch": 0.1204605845881311,
      "grad_norm": 3.260859727859497,
      "learning_rate": 6.868e-05,
      "loss": 0.0232,
      "step": 34
    },
    {
      "epoch": 0.12400354295837024,
      "grad_norm": 3.787526845932007,
      "learning_rate": 7.07e-05,
      "loss": 0.0379,
      "step": 35
    },
    {
      "epoch": 0.1275465013286094,
      "grad_norm": 3.597693920135498,
      "learning_rate": 7.272e-05,
      "loss": 0.0427,
      "step": 36
    },
    {
      "epoch": 0.13108945969884853,
      "grad_norm": 3.5158910751342773,
      "learning_rate": 7.474e-05,
      "loss": 0.0329,
      "step": 37
    },
    {
      "epoch": 0.13463241806908768,
      "grad_norm": 2.715430736541748,
      "learning_rate": 7.676e-05,
      "loss": 0.0368,
      "step": 38
    },
    {
      "epoch": 0.13817537643932684,
      "grad_norm": 2.582350254058838,
      "learning_rate": 7.878e-05,
      "loss": 0.037,
      "step": 39
    },
    {
      "epoch": 0.141718334809566,
      "grad_norm": 13.194321632385254,
      "learning_rate": 8.080000000000001e-05,
      "loss": 0.0468,
      "step": 40
    },
    {
      "epoch": 0.14526129317980513,
      "grad_norm": 4.773246765136719,
      "learning_rate": 8.282e-05,
      "loss": 0.037,
      "step": 41
    },
    {
      "epoch": 0.1488042515500443,
      "grad_norm": 2.5849812030792236,
      "learning_rate": 8.484e-05,
      "loss": 0.0268,
      "step": 42
    },
    {
      "epoch": 0.15234720992028344,
      "grad_norm": 3.463045597076416,
      "learning_rate": 8.686e-05,
      "loss": 0.0543,
      "step": 43
    },
    {
      "epoch": 0.15589016829052257,
      "grad_norm": 2.3256499767303467,
      "learning_rate": 8.888e-05,
      "loss": 0.0395,
      "step": 44
    },
    {
      "epoch": 0.15943312666076173,
      "grad_norm": 3.239624261856079,
      "learning_rate": 9.09e-05,
      "loss": 0.0344,
      "step": 45
    },
    {
      "epoch": 0.1629760850310009,
      "grad_norm": 2.3957650661468506,
      "learning_rate": 9.292000000000001e-05,
      "loss": 0.0405,
      "step": 46
    },
    {
      "epoch": 0.16651904340124005,
      "grad_norm": 3.645236015319824,
      "learning_rate": 9.494e-05,
      "loss": 0.0349,
      "step": 47
    },
    {
      "epoch": 0.17006200177147918,
      "grad_norm": 2.4192404747009277,
      "learning_rate": 9.696e-05,
      "loss": 0.0492,
      "step": 48
    },
    {
      "epoch": 0.17360496014171833,
      "grad_norm": 3.1436197757720947,
      "learning_rate": 9.898e-05,
      "loss": 0.0434,
      "step": 49
    },
    {
      "epoch": 0.1771479185119575,
      "grad_norm": 3.3469741344451904,
      "learning_rate": 0.000101,
      "loss": 0.0593,
      "step": 50
    },
    {
      "epoch": 0.18069087688219662,
      "grad_norm": 5.7896857261657715,
      "learning_rate": 0.00010302,
      "loss": 0.0494,
      "step": 51
    },
    {
      "epoch": 0.18423383525243578,
      "grad_norm": 1.7877720594406128,
      "learning_rate": 0.00010504000000000001,
      "loss": 0.0489,
      "step": 52
    },
    {
      "epoch": 0.18777679362267494,
      "grad_norm": 2.22179913520813,
      "learning_rate": 0.00010706000000000001,
      "loss": 0.0317,
      "step": 53
    },
    {
      "epoch": 0.1913197519929141,
      "grad_norm": 18.348230361938477,
      "learning_rate": 0.00010908,
      "loss": 0.0232,
      "step": 54
    },
    {
      "epoch": 0.19486271036315322,
      "grad_norm": 3.2421276569366455,
      "learning_rate": 0.00011110000000000002,
      "loss": 0.0315,
      "step": 55
    },
    {
      "epoch": 0.19840566873339238,
      "grad_norm": 3.7831366062164307,
      "learning_rate": 0.00011312000000000001,
      "loss": 0.0607,
      "step": 56
    },
    {
      "epoch": 0.20194862710363154,
      "grad_norm": 2.6712589263916016,
      "learning_rate": 0.00011514,
      "loss": 0.026,
      "step": 57
    },
    {
      "epoch": 0.20549158547387067,
      "grad_norm": 8.629843711853027,
      "learning_rate": 0.00011716,
      "loss": 0.0505,
      "step": 58
    },
    {
      "epoch": 0.20903454384410983,
      "grad_norm": 3.9085004329681396,
      "learning_rate": 0.00011918,
      "loss": 0.037,
      "step": 59
    },
    {
      "epoch": 0.21257750221434898,
      "grad_norm": 2.262557029724121,
      "learning_rate": 0.00012119999999999999,
      "loss": 0.0246,
      "step": 60
    },
    {
      "epoch": 0.21612046058458814,
      "grad_norm": 2.5943472385406494,
      "learning_rate": 0.00012322,
      "loss": 0.0486,
      "step": 61
    },
    {
      "epoch": 0.21966341895482727,
      "grad_norm": 2.832202196121216,
      "learning_rate": 0.00012524000000000001,
      "loss": 0.041,
      "step": 62
    },
    {
      "epoch": 0.22320637732506643,
      "grad_norm": 5.549154758453369,
      "learning_rate": 0.00012726,
      "loss": 0.0617,
      "step": 63
    },
    {
      "epoch": 0.2267493356953056,
      "grad_norm": 2.2473719120025635,
      "learning_rate": 0.00012928,
      "loss": 0.0294,
      "step": 64
    },
    {
      "epoch": 0.23029229406554472,
      "grad_norm": 4.177027225494385,
      "learning_rate": 0.00013130000000000002,
      "loss": 0.034,
      "step": 65
    },
    {
      "epoch": 0.23383525243578387,
      "grad_norm": 3.6904377937316895,
      "learning_rate": 0.00013332,
      "loss": 0.0379,
      "step": 66
    },
    {
      "epoch": 0.23737821080602303,
      "grad_norm": 9.35798168182373,
      "learning_rate": 0.00013534000000000002,
      "loss": 0.0659,
      "step": 67
    },
    {
      "epoch": 0.2409211691762622,
      "grad_norm": 5.50313663482666,
      "learning_rate": 0.00013736,
      "loss": 0.0997,
      "step": 68
    },
    {
      "epoch": 0.24446412754650132,
      "grad_norm": 4.885682106018066,
      "learning_rate": 0.00013937999999999998,
      "loss": 0.0525,
      "step": 69
    },
    {
      "epoch": 0.24800708591674048,
      "grad_norm": 2.4924678802490234,
      "learning_rate": 0.0001414,
      "loss": 0.0462,
      "step": 70
    },
    {
      "epoch": 0.25155004428697963,
      "grad_norm": 4.032700538635254,
      "learning_rate": 0.00014342,
      "loss": 0.0485,
      "step": 71
    },
    {
      "epoch": 0.2550930026572188,
      "grad_norm": 3.3813512325286865,
      "learning_rate": 0.00014544,
      "loss": 0.0397,
      "step": 72
    },
    {
      "epoch": 0.25863596102745795,
      "grad_norm": 3.404021978378296,
      "learning_rate": 0.00014746,
      "loss": 0.062,
      "step": 73
    },
    {
      "epoch": 0.26217891939769705,
      "grad_norm": 2.480051040649414,
      "learning_rate": 0.00014948,
      "loss": 0.0472,
      "step": 74
    },
    {
      "epoch": 0.2657218777679362,
      "grad_norm": 3.1103131771087646,
      "learning_rate": 0.0001515,
      "loss": 0.0457,
      "step": 75
    },
    {
      "epoch": 0.26926483613817537,
      "grad_norm": 2.0310606956481934,
      "learning_rate": 0.00015352,
      "loss": 0.0402,
      "step": 76
    },
    {
      "epoch": 0.2728077945084145,
      "grad_norm": 3.9252572059631348,
      "learning_rate": 0.00015554000000000002,
      "loss": 0.056,
      "step": 77
    },
    {
      "epoch": 0.2763507528786537,
      "grad_norm": 9.44784927368164,
      "learning_rate": 0.00015756,
      "loss": 0.0575,
      "step": 78
    },
    {
      "epoch": 0.27989371124889284,
      "grad_norm": 5.473963737487793,
      "learning_rate": 0.00015958000000000001,
      "loss": 0.0604,
      "step": 79
    },
    {
      "epoch": 0.283436669619132,
      "grad_norm": 3.968559741973877,
      "learning_rate": 0.00016160000000000002,
      "loss": 0.062,
      "step": 80
    },
    {
      "epoch": 0.2869796279893711,
      "grad_norm": 4.429202556610107,
      "learning_rate": 0.00016362,
      "loss": 0.0634,
      "step": 81
    },
    {
      "epoch": 0.29052258635961026,
      "grad_norm": 11.875776290893555,
      "learning_rate": 0.00016564,
      "loss": 0.1017,
      "step": 82
    },
    {
      "epoch": 0.2940655447298494,
      "grad_norm": 2.318000555038452,
      "learning_rate": 0.00016766,
      "loss": 0.0305,
      "step": 83
    },
    {
      "epoch": 0.2976085031000886,
      "grad_norm": 55.53030776977539,
      "learning_rate": 0.00016968,
      "loss": 0.1664,
      "step": 84
    },
    {
      "epoch": 0.30115146147032773,
      "grad_norm": 2.9864108562469482,
      "learning_rate": 0.0001717,
      "loss": 0.0356,
      "step": 85
    },
    {
      "epoch": 0.3046944198405669,
      "grad_norm": 11.996382713317871,
      "learning_rate": 0.00017372,
      "loss": 0.0597,
      "step": 86
    },
    {
      "epoch": 0.30823737821080605,
      "grad_norm": 85.5622787475586,
      "learning_rate": 0.00017574,
      "loss": 0.9084,
      "step": 87
    },
    {
      "epoch": 0.31178033658104515,
      "grad_norm": 15.457459449768066,
      "learning_rate": 0.00017776,
      "loss": 0.2301,
      "step": 88
    },
    {
      "epoch": 0.3153232949512843,
      "grad_norm": 60.711517333984375,
      "learning_rate": 0.00017978000000000002,
      "loss": 0.8903,
      "step": 89
    },
    {
      "epoch": 0.31886625332152346,
      "grad_norm": 10.001251220703125,
      "learning_rate": 0.0001818,
      "loss": 0.1763,
      "step": 90
    },
    {
      "epoch": 0.3224092116917626,
      "grad_norm": 5.155291557312012,
      "learning_rate": 0.00018382,
      "loss": 0.0778,
      "step": 91
    },
    {
      "epoch": 0.3259521700620018,
      "grad_norm": 6.2389326095581055,
      "learning_rate": 0.00018584000000000002,
      "loss": 0.1137,
      "step": 92
    },
    {
      "epoch": 0.32949512843224094,
      "grad_norm": 11.356403350830078,
      "learning_rate": 0.00018786,
      "loss": 0.0675,
      "step": 93
    },
    {
      "epoch": 0.3330380868024801,
      "grad_norm": 13.513825416564941,
      "learning_rate": 0.00018988,
      "loss": 0.1551,
      "step": 94
    },
    {
      "epoch": 0.3365810451727192,
      "grad_norm": 257.41607666015625,
      "learning_rate": 0.0001919,
      "loss": 1.5096,
      "step": 95
    },
    {
      "epoch": 0.34012400354295835,
      "grad_norm": 25.03782844543457,
      "learning_rate": 0.00019392,
      "loss": 0.3461,
      "step": 96
    },
    {
      "epoch": 0.3436669619131975,
      "grad_norm": 6.5227556228637695,
      "learning_rate": 0.00019594,
      "loss": 0.0875,
      "step": 97
    },
    {
      "epoch": 0.34720992028343667,
      "grad_norm": 4.013297080993652,
      "learning_rate": 0.00019796,
      "loss": 0.0362,
      "step": 98
    },
    {
      "epoch": 0.3507528786536758,
      "grad_norm": 2.942173480987549,
      "learning_rate": 0.00019998,
      "loss": 0.038,
      "step": 99
    },
    {
      "epoch": 0.354295837023915,
      "grad_norm": 7.893823623657227,
      "learning_rate": 0.000202,
      "loss": 0.0616,
      "step": 100
    },
    {
      "epoch": 0.35783879539415414,
      "grad_norm": 7.354511737823486,
      "learning_rate": 0.0002019986155169901,
      "loss": 0.0805,
      "step": 101
    },
    {
      "epoch": 0.36138175376439324,
      "grad_norm": 22.7661190032959,
      "learning_rate": 0.00020199446210591673,
      "loss": 0.2151,
      "step": 102
    },
    {
      "epoch": 0.3649247121346324,
      "grad_norm": 4.101150989532471,
      "learning_rate": 0.00020198753988064772,
      "loss": 0.0387,
      "step": 103
    },
    {
      "epoch": 0.36846767050487156,
      "grad_norm": 16.70463752746582,
      "learning_rate": 0.0002019778490309594,
      "loss": 0.1978,
      "step": 104
    },
    {
      "epoch": 0.3720106288751107,
      "grad_norm": 7.770139217376709,
      "learning_rate": 0.00020196538982253126,
      "loss": 0.094,
      "step": 105
    },
    {
      "epoch": 0.3755535872453499,
      "grad_norm": 4.9610795974731445,
      "learning_rate": 0.0002019501625969389,
      "loss": 0.1127,
      "step": 106
    },
    {
      "epoch": 0.37909654561558903,
      "grad_norm": 5.921096324920654,
      "learning_rate": 0.00020193216777164428,
      "loss": 0.0596,
      "step": 107
    },
    {
      "epoch": 0.3826395039858282,
      "grad_norm": 12.536717414855957,
      "learning_rate": 0.0002019114058399847,
      "loss": 0.0652,
      "step": 108
    },
    {
      "epoch": 0.3861824623560673,
      "grad_norm": 4.937745094299316,
      "learning_rate": 0.00020188787737115897,
      "loss": 0.0986,
      "step": 109
    },
    {
      "epoch": 0.38972542072630645,
      "grad_norm": 3.928455114364624,
      "learning_rate": 0.00020186158301021195,
      "loss": 0.0533,
      "step": 110
    },
    {
      "epoch": 0.3932683790965456,
      "grad_norm": 4.18632173538208,
      "learning_rate": 0.00020183252347801686,
      "loss": 0.0661,
      "step": 111
    },
    {
      "epoch": 0.39681133746678476,
      "grad_norm": 3.5552995204925537,
      "learning_rate": 0.00020180069957125544,
      "loss": 0.0576,
      "step": 112
    },
    {
      "epoch": 0.4003542958370239,
      "grad_norm": 5.155257225036621,
      "learning_rate": 0.00020176611216239613,
      "loss": 0.0512,
      "step": 113
    },
    {
      "epoch": 0.4038972542072631,
      "grad_norm": 4.801251411437988,
      "learning_rate": 0.00020172876219967027,
      "loss": 0.0478,
      "step": 114
    },
    {
      "epoch": 0.40744021257750224,
      "grad_norm": 4.40409517288208,
      "learning_rate": 0.00020168865070704594,
      "loss": 0.0417,
      "step": 115
    },
    {
      "epoch": 0.41098317094774134,
      "grad_norm": 3.4286465644836426,
      "learning_rate": 0.00020164577878419994,
      "loss": 0.0287,
      "step": 116
    },
    {
      "epoch": 0.4145261293179805,
      "grad_norm": 6.111563682556152,
      "learning_rate": 0.00020160014760648774,
      "loss": 0.0288,
      "step": 117
    },
    {
      "epoch": 0.41806908768821965,
      "grad_norm": 9.4945650100708,
      "learning_rate": 0.00020155175842491107,
      "loss": 0.1203,
      "step": 118
    },
    {
      "epoch": 0.4216120460584588,
      "grad_norm": 4.352670192718506,
      "learning_rate": 0.00020150061256608387,
      "loss": 0.0608,
      "step": 119
    },
    {
      "epoch": 0.42515500442869797,
      "grad_norm": 3.108896255493164,
      "learning_rate": 0.0002014467114321956,
      "loss": 0.0497,
      "step": 120
    },
    {
      "epoch": 0.4286979627989371,
      "grad_norm": 10.037585258483887,
      "learning_rate": 0.00020139005650097317,
      "loss": 0.0714,
      "step": 121
    },
    {
      "epoch": 0.4322409211691763,
      "grad_norm": 3.0859506130218506,
      "learning_rate": 0.00020133064932564002,
      "loss": 0.0597,
      "step": 122
    },
    {
      "epoch": 0.4357838795394154,
      "grad_norm": 2.885244846343994,
      "learning_rate": 0.000201268491534874,
      "loss": 0.0421,
      "step": 123
    },
    {
      "epoch": 0.43932683790965454,
      "grad_norm": 5.023353576660156,
      "learning_rate": 0.00020120358483276227,
      "loss": 0.0793,
      "step": 124
    },
    {
      "epoch": 0.4428697962798937,
      "grad_norm": 5.706242084503174,
      "learning_rate": 0.00020113593099875486,
      "loss": 0.0673,
      "step": 125
    },
    {
      "epoch": 0.44641275465013286,
      "grad_norm": 5.34003210067749,
      "learning_rate": 0.00020106553188761582,
      "loss": 0.0555,
      "step": 126
    },
    {
      "epoch": 0.449955713020372,
      "grad_norm": 3.727440118789673,
      "learning_rate": 0.0002009923894293723,
      "loss": 0.0697,
      "step": 127
    },
    {
      "epoch": 0.4534986713906112,
      "grad_norm": 5.76682186126709,
      "learning_rate": 0.00020091650562926183,
      "loss": 0.0942,
      "step": 128
    },
    {
      "epoch": 0.45704162976085033,
      "grad_norm": 4.1472249031066895,
      "learning_rate": 0.00020083788256767702,
      "loss": 0.0707,
      "step": 129
    },
    {
      "epoch": 0.46058458813108943,
      "grad_norm": 3.9839630126953125,
      "learning_rate": 0.00020075652240010892,
      "loss": 0.0804,
      "step": 130
    },
    {
      "epoch": 0.4641275465013286,
      "grad_norm": 3.167534589767456,
      "learning_rate": 0.00020067242735708754,
      "loss": 0.0524,
      "step": 131
    },
    {
      "epoch": 0.46767050487156775,
      "grad_norm": 4.119954586029053,
      "learning_rate": 0.00020058559974412102,
      "loss": 0.0565,
      "step": 132
    },
    {
      "epoch": 0.4712134632418069,
      "grad_norm": 3.3348987102508545,
      "learning_rate": 0.00020049604194163217,
      "loss": 0.0414,
      "step": 133
    },
    {
      "epoch": 0.47475642161204606,
      "grad_norm": 6.277345657348633,
      "learning_rate": 0.00020040375640489343,
      "loss": 0.0467,
      "step": 134
    },
    {
      "epoch": 0.4782993799822852,
      "grad_norm": 3.517561912536621,
      "learning_rate": 0.00020030874566395943,
      "loss": 0.0535,
      "step": 135
    },
    {
      "epoch": 0.4818423383525244,
      "grad_norm": 5.174115180969238,
      "learning_rate": 0.00020021101232359757,
      "loss": 0.0528,
      "step": 136
    },
    {
      "epoch": 0.4853852967227635,
      "grad_norm": 10.583273887634277,
      "learning_rate": 0.00020011055906321676,
      "loss": 0.0571,
      "step": 137
    },
    {
      "epoch": 0.48892825509300264,
      "grad_norm": 2.8985111713409424,
      "learning_rate": 0.0002000073886367939,
      "loss": 0.04,
      "step": 138
    },
    {
      "epoch": 0.4924712134632418,
      "grad_norm": 3.7959556579589844,
      "learning_rate": 0.00019990150387279835,
      "loss": 0.0459,
      "step": 139
    },
    {
      "epoch": 0.49601417183348095,
      "grad_norm": 5.849306583404541,
      "learning_rate": 0.00019979290767411438,
      "loss": 0.0497,
      "step": 140
    },
    {
      "epoch": 0.4995571302037201,
      "grad_norm": 16.320491790771484,
      "learning_rate": 0.00019968160301796163,
      "loss": 0.1,
      "step": 141
    },
    {
      "epoch": 0.5031000885739593,
      "grad_norm": 3.156020164489746,
      "learning_rate": 0.0001995675929558135,
      "loss": 0.0478,
      "step": 142
    },
    {
      "epoch": 0.5066430469441984,
      "grad_norm": 3.743115186691284,
      "learning_rate": 0.0001994508806133134,
      "loss": 0.0462,
      "step": 143
    },
    {
      "epoch": 0.5101860053144376,
      "grad_norm": 4.2121262550354,
      "learning_rate": 0.0001993314691901892,
      "loss": 0.0413,
      "step": 144
    },
    {
      "epoch": 0.5137289636846767,
      "grad_norm": 3.4036576747894287,
      "learning_rate": 0.00019920936196016534,
      "loss": 0.0487,
      "step": 145
    },
    {
      "epoch": 0.5172719220549159,
      "grad_norm": 2.6088032722473145,
      "learning_rate": 0.00019908456227087326,
      "loss": 0.0528,
      "step": 146
    },
    {
      "epoch": 0.520814880425155,
      "grad_norm": 1.3771120309829712,
      "learning_rate": 0.00019895707354375945,
      "loss": 0.0248,
      "step": 147
    },
    {
      "epoch": 0.5243578387953941,
      "grad_norm": 2.9524734020233154,
      "learning_rate": 0.00019882689927399174,
      "loss": 0.031,
      "step": 148
    },
    {
      "epoch": 0.5279007971656333,
      "grad_norm": 15.377911567687988,
      "learning_rate": 0.00019869404303036355,
      "loss": 0.0842,
      "step": 149
    },
    {
      "epoch": 0.5314437555358724,
      "grad_norm": 7.4169135093688965,
      "learning_rate": 0.00019855850845519588,
      "loss": 0.0923,
      "step": 150
    },
    {
      "epoch": 0.5349867139061116,
      "grad_norm": 7.898824214935303,
      "learning_rate": 0.00019842029926423762,
      "loss": 0.0391,
      "step": 151
    },
    {
      "epoch": 0.5385296722763507,
      "grad_norm": 9.734644889831543,
      "learning_rate": 0.00019827941924656348,
      "loss": 0.1149,
      "step": 152
    },
    {
      "epoch": 0.54207263064659,
      "grad_norm": 4.397625923156738,
      "learning_rate": 0.00019813587226447034,
      "loss": 0.0584,
      "step": 153
    },
    {
      "epoch": 0.545615589016829,
      "grad_norm": 4.095290660858154,
      "learning_rate": 0.00019798966225337126,
      "loss": 0.0464,
      "step": 154
    },
    {
      "epoch": 0.5491585473870682,
      "grad_norm": 5.90880012512207,
      "learning_rate": 0.00019784079322168752,
      "loss": 0.0445,
      "step": 155
    },
    {
      "epoch": 0.5527015057573074,
      "grad_norm": 4.075512886047363,
      "learning_rate": 0.00019768926925073878,
      "loss": 0.0474,
      "step": 156
    },
    {
      "epoch": 0.5562444641275465,
      "grad_norm": 2.5753815174102783,
      "learning_rate": 0.00019753509449463134,
      "loss": 0.0519,
      "step": 157
    },
    {
      "epoch": 0.5597874224977857,
      "grad_norm": 1.7522939443588257,
      "learning_rate": 0.00019737827318014396,
      "loss": 0.0348,
      "step": 158
    },
    {
      "epoch": 0.5633303808680248,
      "grad_norm": 4.002205848693848,
      "learning_rate": 0.00019721880960661223,
      "loss": 0.0609,
      "step": 159
    },
    {
      "epoch": 0.566873339238264,
      "grad_norm": 7.690733909606934,
      "learning_rate": 0.00019705670814581052,
      "loss": 0.0842,
      "step": 160
    },
    {
      "epoch": 0.5704162976085031,
      "grad_norm": 5.834394454956055,
      "learning_rate": 0.0001968919732418323,
      "loss": 0.0731,
      "step": 161
    },
    {
      "epoch": 0.5739592559787422,
      "grad_norm": 3.365072727203369,
      "learning_rate": 0.00019672460941096818,
      "loss": 0.0945,
      "step": 162
    },
    {
      "epoch": 0.5775022143489814,
      "grad_norm": 3.1405742168426514,
      "learning_rate": 0.0001965546212415821,
      "loss": 0.0547,
      "step": 163
    },
    {
      "epoch": 0.5810451727192205,
      "grad_norm": 4.006103038787842,
      "learning_rate": 0.0001963820133939856,
      "loss": 0.0766,
      "step": 164
    },
    {
      "epoch": 0.5845881310894597,
      "grad_norm": 3.036452293395996,
      "learning_rate": 0.00019620679060031003,
      "loss": 0.0423,
      "step": 165
    },
    {
      "epoch": 0.5881310894596988,
      "grad_norm": 6.233029842376709,
      "learning_rate": 0.00019602895766437678,
      "loss": 0.061,
      "step": 166
    },
    {
      "epoch": 0.591674047829938,
      "grad_norm": 2.5433619022369385,
      "learning_rate": 0.0001958485194615656,
      "loss": 0.0604,
      "step": 167
    },
    {
      "epoch": 0.5952170062001771,
      "grad_norm": 2.707562208175659,
      "learning_rate": 0.00019566548093868106,
      "loss": 0.0324,
      "step": 168
    },
    {
      "epoch": 0.5987599645704162,
      "grad_norm": 2.866919755935669,
      "learning_rate": 0.00019547984711381662,
      "loss": 0.0555,
      "step": 169
    },
    {
      "epoch": 0.6023029229406555,
      "grad_norm": 2.392132043838501,
      "learning_rate": 0.00019529162307621738,
      "loss": 0.0296,
      "step": 170
    },
    {
      "epoch": 0.6058458813108946,
      "grad_norm": 8.598843574523926,
      "learning_rate": 0.00019510081398614045,
      "loss": 0.0763,
      "step": 171
    },
    {
      "epoch": 0.6093888396811338,
      "grad_norm": 4.732206344604492,
      "learning_rate": 0.00019490742507471338,
      "loss": 0.0343,
      "step": 172
    },
    {
      "epoch": 0.6129317980513729,
      "grad_norm": 9.818289756774902,
      "learning_rate": 0.00019471146164379093,
      "loss": 0.0724,
      "step": 173
    },
    {
      "epoch": 0.6164747564216121,
      "grad_norm": 8.120203018188477,
      "learning_rate": 0.00019451292906580948,
      "loss": 0.0685,
      "step": 174
    },
    {
      "epoch": 0.6200177147918512,
      "grad_norm": 4.892403602600098,
      "learning_rate": 0.00019431183278363997,
      "loss": 0.1228,
      "step": 175
    },
    {
      "epoch": 0.6235606731620903,
      "grad_norm": 2.2905120849609375,
      "learning_rate": 0.00019410817831043856,
      "loss": 0.0494,
      "step": 176
    },
    {
      "epoch": 0.6271036315323295,
      "grad_norm": 1.9224469661712646,
      "learning_rate": 0.00019390197122949552,
      "loss": 0.0359,
      "step": 177
    },
    {
      "epoch": 0.6306465899025686,
      "grad_norm": 3.4474852085113525,
      "learning_rate": 0.0001936932171940821,
      "loss": 0.0428,
      "step": 178
    },
    {
      "epoch": 0.6341895482728078,
      "grad_norm": 5.11275577545166,
      "learning_rate": 0.0001934819219272957,
      "loss": 0.0769,
      "step": 179
    },
    {
      "epoch": 0.6377325066430469,
      "grad_norm": 3.1210732460021973,
      "learning_rate": 0.0001932680912219027,
      "loss": 0.053,
      "step": 180
    },
    {
      "epoch": 0.6412754650132861,
      "grad_norm": 6.006719589233398,
      "learning_rate": 0.00019305173094017996,
      "loss": 0.0728,
      "step": 181
    },
    {
      "epoch": 0.6448184233835252,
      "grad_norm": 3.2702836990356445,
      "learning_rate": 0.00019283284701375393,
      "loss": 0.0798,
      "step": 182
    },
    {
      "epoch": 0.6483613817537643,
      "grad_norm": 2.9264016151428223,
      "learning_rate": 0.00019261144544343794,
      "loss": 0.0588,
      "step": 183
    },
    {
      "epoch": 0.6519043401240036,
      "grad_norm": 8.378257751464844,
      "learning_rate": 0.00019238753229906797,
      "loss": 0.081,
      "step": 184
    },
    {
      "epoch": 0.6554472984942427,
      "grad_norm": 2.7256081104278564,
      "learning_rate": 0.00019216111371933594,
      "loss": 0.0334,
      "step": 185
    },
    {
      "epoch": 0.6589902568644819,
      "grad_norm": 2.700265645980835,
      "learning_rate": 0.00019193219591162155,
      "loss": 0.0967,
      "step": 186
    },
    {
      "epoch": 0.662533215234721,
      "grad_norm": 5.8346452713012695,
      "learning_rate": 0.00019170078515182216,
      "loss": 0.138,
      "step": 187
    },
    {
      "epoch": 0.6660761736049602,
      "grad_norm": 36.33310317993164,
      "learning_rate": 0.0001914668877841807,
      "loss": 0.1906,
      "step": 188
    },
    {
      "epoch": 0.6696191319751993,
      "grad_norm": 2.1639604568481445,
      "learning_rate": 0.0001912305102211116,
      "loss": 0.0524,
      "step": 189
    },
    {
      "epoch": 0.6731620903454384,
      "grad_norm": 5.425674915313721,
      "learning_rate": 0.00019099165894302515,
      "loss": 0.0425,
      "step": 190
    },
    {
      "epoch": 0.6767050487156776,
      "grad_norm": 3.9154422283172607,
      "learning_rate": 0.00019075034049814983,
      "loss": 0.0525,
      "step": 191
    },
    {
      "epoch": 0.6802480070859167,
      "grad_norm": 3.229419469833374,
      "learning_rate": 0.00019050656150235268,
      "loss": 0.0378,
      "step": 192
    },
    {
      "epoch": 0.6837909654561559,
      "grad_norm": 2.457200527191162,
      "learning_rate": 0.00019026032863895805,
      "loss": 0.0446,
      "step": 193
    },
    {
      "epoch": 0.687333923826395,
      "grad_norm": 4.7356038093566895,
      "learning_rate": 0.0001900116486585642,
      "loss": 0.1273,
      "step": 194
    },
    {
      "epoch": 0.6908768821966342,
      "grad_norm": 2.678158760070801,
      "learning_rate": 0.0001897605283788585,
      "loss": 0.0343,
      "step": 195
    },
    {
      "epoch": 0.6944198405668733,
      "grad_norm": 2.674647092819214,
      "learning_rate": 0.0001895069746844302,
      "loss": 0.04,
      "step": 196
    },
    {
      "epoch": 0.6979627989371124,
      "grad_norm": 5.925049304962158,
      "learning_rate": 0.000189250994526582,
      "loss": 0.0583,
      "step": 197
    },
    {
      "epoch": 0.7015057573073517,
      "grad_norm": 2.67167067527771,
      "learning_rate": 0.00018899259492313915,
      "loss": 0.0392,
      "step": 198
    },
    {
      "epoch": 0.7050487156775908,
      "grad_norm": 2.6344971656799316,
      "learning_rate": 0.00018873178295825732,
      "loss": 0.0326,
      "step": 199
    },
    {
      "epoch": 0.70859167404783,
      "grad_norm": 1.7553050518035889,
      "learning_rate": 0.00018846856578222832,
      "loss": 0.0337,
      "step": 200
    },
    {
      "epoch": 0.7121346324180691,
      "grad_norm": 2.535490036010742,
      "learning_rate": 0.00018820295061128394,
      "loss": 0.0504,
      "step": 201
    },
    {
      "epoch": 0.7156775907883083,
      "grad_norm": 2.3876333236694336,
      "learning_rate": 0.00018793494472739831,
      "loss": 0.023,
      "step": 202
    },
    {
      "epoch": 0.7192205491585474,
      "grad_norm": 1.6958738565444946,
      "learning_rate": 0.00018766455547808813,
      "loss": 0.034,
      "step": 203
    },
    {
      "epoch": 0.7227635075287865,
      "grad_norm": 5.384578227996826,
      "learning_rate": 0.0001873917902762112,
      "loss": 0.027,
      "step": 204
    },
    {
      "epoch": 0.7263064658990257,
      "grad_norm": 2.5751683712005615,
      "learning_rate": 0.0001871166565997633,
      "loss": 0.0331,
      "step": 205
    },
    {
      "epoch": 0.7298494242692648,
      "grad_norm": 2.7341415882110596,
      "learning_rate": 0.00018683916199167325,
      "loss": 0.0317,
      "step": 206
    },
    {
      "epoch": 0.733392382639504,
      "grad_norm": 3.3571903705596924,
      "learning_rate": 0.00018655931405959586,
      "loss": 0.0534,
      "step": 207
    },
    {
      "epoch": 0.7369353410097431,
      "grad_norm": 3.4114151000976562,
      "learning_rate": 0.00018627712047570352,
      "loss": 0.0557,
      "step": 208
    },
    {
      "epoch": 0.7404782993799823,
      "grad_norm": 2.152780294418335,
      "learning_rate": 0.00018599258897647594,
      "loss": 0.0268,
      "step": 209
    },
    {
      "epoch": 0.7440212577502214,
      "grad_norm": 6.7897162437438965,
      "learning_rate": 0.00018570572736248782,
      "loss": 0.0647,
      "step": 210
    },
    {
      "epoch": 0.7475642161204605,
      "grad_norm": 2.8548331260681152,
      "learning_rate": 0.0001854165434981953,
      "loss": 0.0568,
      "step": 211
    },
    {
      "epoch": 0.7511071744906997,
      "grad_norm": 4.032414436340332,
      "learning_rate": 0.00018512504531172005,
      "loss": 0.0373,
      "step": 212
    },
    {
      "epoch": 0.7546501328609388,
      "grad_norm": 4.85910177230835,
      "learning_rate": 0.0001848312407946321,
      "loss": 0.0552,
      "step": 213
    },
    {
      "epoch": 0.7581930912311781,
      "grad_norm": 3.7247204780578613,
      "learning_rate": 0.00018453513800173072,
      "loss": 0.0589,
      "step": 214
    },
    {
      "epoch": 0.7617360496014172,
      "grad_norm": 2.951002597808838,
      "learning_rate": 0.00018423674505082356,
      "loss": 0.0356,
      "step": 215
    },
    {
      "epoch": 0.7652790079716564,
      "grad_norm": 2.9198479652404785,
      "learning_rate": 0.0001839360701225041,
      "loss": 0.0386,
      "step": 216
    },
    {
      "epoch": 0.7688219663418955,
      "grad_norm": 1.3766493797302246,
      "learning_rate": 0.00018363312145992737,
      "loss": 0.0157,
      "step": 217
    },
    {
      "epoch": 0.7723649247121346,
      "grad_norm": 3.254570960998535,
      "learning_rate": 0.00018332790736858397,
      "loss": 0.0551,
      "step": 218
    },
    {
      "epoch": 0.7759078830823738,
      "grad_norm": 7.05130672454834,
      "learning_rate": 0.00018302043621607245,
      "loss": 0.0856,
      "step": 219
    },
    {
      "epoch": 0.7794508414526129,
      "grad_norm": 3.749147415161133,
      "learning_rate": 0.00018271071643186968,
      "loss": 0.0723,
      "step": 220
    },
    {
      "epoch": 0.7829937998228521,
      "grad_norm": 1.6400352716445923,
      "learning_rate": 0.0001823987565071001,
      "loss": 0.0262,
      "step": 221
    },
    {
      "epoch": 0.7865367581930912,
      "grad_norm": 2.1209402084350586,
      "learning_rate": 0.00018208456499430256,
      "loss": 0.0434,
      "step": 222
    },
    {
      "epoch": 0.7900797165633304,
      "grad_norm": 3.1746721267700195,
      "learning_rate": 0.00018176815050719615,
      "loss": 0.0394,
      "step": 223
    },
    {
      "epoch": 0.7936226749335695,
      "grad_norm": 1.7733818292617798,
      "learning_rate": 0.00018144952172044381,
      "loss": 0.0359,
      "step": 224
    },
    {
      "epoch": 0.7971656333038086,
      "grad_norm": 2.6539013385772705,
      "learning_rate": 0.00018112868736941477,
      "loss": 0.0263,
      "step": 225
    },
    {
      "epoch": 0.8007085916740478,
      "grad_norm": 2.6746320724487305,
      "learning_rate": 0.00018080565624994474,
      "loss": 0.0294,
      "step": 226
    },
    {
      "epoch": 0.804251550044287,
      "grad_norm": 3.646777391433716,
      "learning_rate": 0.00018048043721809507,
      "loss": 0.0387,
      "step": 227
    },
    {
      "epoch": 0.8077945084145262,
      "grad_norm": 4.014429092407227,
      "learning_rate": 0.00018015303918990982,
      "loss": 0.0423,
      "step": 228
    },
    {
      "epoch": 0.8113374667847653,
      "grad_norm": 6.775087833404541,
      "learning_rate": 0.0001798234711411713,
      "loss": 0.0516,
      "step": 229
    },
    {
      "epoch": 0.8148804251550045,
      "grad_norm": 1.8717288970947266,
      "learning_rate": 0.00017949174210715407,
      "loss": 0.0351,
      "step": 230
    },
    {
      "epoch": 0.8184233835252436,
      "grad_norm": 6.993075847625732,
      "learning_rate": 0.00017915786118237714,
      "loss": 0.0918,
      "step": 231
    },
    {
      "epoch": 0.8219663418954827,
      "grad_norm": 2.1425578594207764,
      "learning_rate": 0.0001788218375203547,
      "loss": 0.0594,
      "step": 232
    },
    {
      "epoch": 0.8255093002657219,
      "grad_norm": 3.5647621154785156,
      "learning_rate": 0.00017848368033334528,
      "loss": 0.0528,
      "step": 233
    },
    {
      "epoch": 0.829052258635961,
      "grad_norm": 3.7833359241485596,
      "learning_rate": 0.00017814339889209887,
      "loss": 0.0497,
      "step": 234
    },
    {
      "epoch": 0.8325952170062002,
      "grad_norm": 1.3636075258255005,
      "learning_rate": 0.00017780100252560313,
      "loss": 0.0181,
      "step": 235
    },
    {
      "epoch": 0.8361381753764393,
      "grad_norm": 1.8320082426071167,
      "learning_rate": 0.0001774565006208274,
      "loss": 0.0183,
      "step": 236
    },
    {
      "epoch": 0.8396811337466785,
      "grad_norm": 2.3706352710723877,
      "learning_rate": 0.00017710990262246543,
      "loss": 0.048,
      "step": 237
    },
    {
      "epoch": 0.8432240921169176,
      "grad_norm": 2.8141376972198486,
      "learning_rate": 0.0001767612180326764,
      "loss": 0.0347,
      "step": 238
    },
    {
      "epoch": 0.8467670504871567,
      "grad_norm": 1.5940569639205933,
      "learning_rate": 0.00017641045641082453,
      "loss": 0.0342,
      "step": 239
    },
    {
      "epoch": 0.8503100088573959,
      "grad_norm": 1.785904884338379,
      "learning_rate": 0.00017605762737321683,
      "loss": 0.0298,
      "step": 240
    },
    {
      "epoch": 0.853852967227635,
      "grad_norm": 2.5388872623443604,
      "learning_rate": 0.0001757027405928396,
      "loss": 0.0205,
      "step": 241
    },
    {
      "epoch": 0.8573959255978743,
      "grad_norm": 3.2817773818969727,
      "learning_rate": 0.0001753458057990932,
      "loss": 0.0336,
      "step": 242
    },
    {
      "epoch": 0.8609388839681134,
      "grad_norm": 1.6606643199920654,
      "learning_rate": 0.00017498683277752527,
      "loss": 0.0216,
      "step": 243
    },
    {
      "epoch": 0.8644818423383526,
      "grad_norm": 4.023453712463379,
      "learning_rate": 0.00017462583136956258,
      "loss": 0.0397,
      "step": 244
    },
    {
      "epoch": 0.8680248007085917,
      "grad_norm": 0.8544020056724548,
      "learning_rate": 0.00017426281147224105,
      "loss": 0.0144,
      "step": 245
    },
    {
      "epoch": 0.8715677590788308,
      "grad_norm": 2.7009518146514893,
      "learning_rate": 0.00017389778303793457,
      "loss": 0.0243,
      "step": 246
    },
    {
      "epoch": 0.87511071744907,
      "grad_norm": 2.2449357509613037,
      "learning_rate": 0.00017353075607408209,
      "loss": 0.0181,
      "step": 247
    },
    {
      "epoch": 0.8786536758193091,
      "grad_norm": 3.0773513317108154,
      "learning_rate": 0.00017316174064291315,
      "loss": 0.0357,
      "step": 248
    },
    {
      "epoch": 0.8821966341895483,
      "grad_norm": 2.021902322769165,
      "learning_rate": 0.00017279074686117225,
      "loss": 0.0229,
      "step": 249
    },
    {
      "epoch": 0.8857395925597874,
      "grad_norm": 1.1855192184448242,
      "learning_rate": 0.0001724177848998413,
      "loss": 0.0159,
      "step": 250
    },
    {
      "epoch": 0.8892825509300266,
      "grad_norm": 2.7824289798736572,
      "learning_rate": 0.0001720428649838609,
      "loss": 0.0414,
      "step": 251
    },
    {
      "epoch": 0.8928255093002657,
      "grad_norm": 3.3806657791137695,
      "learning_rate": 0.0001716659973918499,
      "loss": 0.0401,
      "step": 252
    },
    {
      "epoch": 0.8963684676705048,
      "grad_norm": 2.7893075942993164,
      "learning_rate": 0.00017128719245582374,
      "loss": 0.0421,
      "step": 253
    },
    {
      "epoch": 0.899911426040744,
      "grad_norm": 3.2988412380218506,
      "learning_rate": 0.0001709064605609111,
      "loss": 0.0376,
      "step": 254
    },
    {
      "epoch": 0.9034543844109831,
      "grad_norm": 2.2144927978515625,
      "learning_rate": 0.00017052381214506914,
      "loss": 0.0209,
      "step": 255
    },
    {
      "epoch": 0.9069973427812223,
      "grad_norm": 4.647397518157959,
      "learning_rate": 0.00017013925769879755,
      "loss": 0.0367,
      "step": 256
    },
    {
      "epoch": 0.9105403011514615,
      "grad_norm": 3.4333746433258057,
      "learning_rate": 0.0001697528077648507,
      "loss": 0.0255,
      "step": 257
    },
    {
      "epoch": 0.9140832595217007,
      "grad_norm": 1.8429880142211914,
      "learning_rate": 0.00016936447293794873,
      "loss": 0.0184,
      "step": 258
    },
    {
      "epoch": 0.9176262178919398,
      "grad_norm": 1.512812614440918,
      "learning_rate": 0.0001689742638644871,
      "loss": 0.0156,
      "step": 259
    },
    {
      "epoch": 0.9211691762621789,
      "grad_norm": 2.512474298477173,
      "learning_rate": 0.0001685821912422447,
      "loss": 0.0435,
      "step": 260
    },
    {
      "epoch": 0.9247121346324181,
      "grad_norm": 3.278831720352173,
      "learning_rate": 0.00016818826582009044,
      "loss": 0.0274,
      "step": 261
    },
    {
      "epoch": 0.9282550930026572,
      "grad_norm": 4.005672454833984,
      "learning_rate": 0.00016779249839768884,
      "loss": 0.0355,
      "step": 262
    },
    {
      "epoch": 0.9317980513728964,
      "grad_norm": 4.274734973907471,
      "learning_rate": 0.00016739489982520368,
      "loss": 0.0316,
      "step": 263
    },
    {
      "epoch": 0.9353410097431355,
      "grad_norm": 3.6488261222839355,
      "learning_rate": 0.00016699548100300066,
      "loss": 0.0385,
      "step": 264
    },
    {
      "epoch": 0.9388839681133747,
      "grad_norm": 1.8470478057861328,
      "learning_rate": 0.00016659425288134854,
      "loss": 0.0255,
      "step": 265
    },
    {
      "epoch": 0.9424269264836138,
      "grad_norm": 1.6004862785339355,
      "learning_rate": 0.00016619122646011902,
      "loss": 0.0129,
      "step": 266
    },
    {
      "epoch": 0.9459698848538529,
      "grad_norm": 1.422613501548767,
      "learning_rate": 0.00016578641278848497,
      "loss": 0.0129,
      "step": 267
    },
    {
      "epoch": 0.9495128432240921,
      "grad_norm": 2.5075550079345703,
      "learning_rate": 0.00016537982296461768,
      "loss": 0.0424,
      "step": 268
    },
    {
      "epoch": 0.9530558015943312,
      "grad_norm": 1.9118682146072388,
      "learning_rate": 0.00016497146813538257,
      "loss": 0.018,
      "step": 269
    },
    {
      "epoch": 0.9565987599645704,
      "grad_norm": 1.4825668334960938,
      "learning_rate": 0.00016456135949603358,
      "loss": 0.014,
      "step": 270
    },
    {
      "epoch": 0.9601417183348095,
      "grad_norm": 1.6457098722457886,
      "learning_rate": 0.00016414950828990625,
      "loss": 0.0169,
      "step": 271
    },
    {
      "epoch": 0.9636846767050488,
      "grad_norm": 1.626107096672058,
      "learning_rate": 0.00016373592580810935,
      "loss": 0.0212,
      "step": 272
    },
    {
      "epoch": 0.9672276350752879,
      "grad_norm": 1.8600038290023804,
      "learning_rate": 0.00016332062338921563,
      "loss": 0.0201,
      "step": 273
    },
    {
      "epoch": 0.970770593445527,
      "grad_norm": 1.0375657081604004,
      "learning_rate": 0.00016290361241895064,
      "loss": 0.0177,
      "step": 274
    },
    {
      "epoch": 0.9743135518157662,
      "grad_norm": 2.0309953689575195,
      "learning_rate": 0.0001624849043298808,
      "loss": 0.0316,
      "step": 275
    },
    {
      "epoch": 0.9778565101860053,
      "grad_norm": 2.301088333129883,
      "learning_rate": 0.00016206451060109988,
      "loss": 0.0175,
      "step": 276
    },
    {
      "epoch": 0.9813994685562445,
      "grad_norm": 1.9179376363754272,
      "learning_rate": 0.0001616424427579143,
      "loss": 0.0289,
      "step": 277
    },
    {
      "epoch": 0.9849424269264836,
      "grad_norm": 4.722525596618652,
      "learning_rate": 0.0001612187123715272,
      "loss": 0.0436,
      "step": 278
    },
    {
      "epoch": 0.9884853852967228,
      "grad_norm": 2.8682198524475098,
      "learning_rate": 0.00016079333105872118,
      "loss": 0.0357,
      "step": 279
    },
    {
      "epoch": 0.9920283436669619,
      "grad_norm": 3.7090978622436523,
      "learning_rate": 0.00016036631048153979,
      "loss": 0.0424,
      "step": 280
    },
    {
      "epoch": 0.995571302037201,
      "grad_norm": 2.536424398422241,
      "learning_rate": 0.00015993766234696785,
      "loss": 0.0226,
      "step": 281
    },
    {
      "epoch": 0.9991142604074402,
      "grad_norm": 3.139700174331665,
      "learning_rate": 0.00015950739840661055,
      "loss": 0.0257,
      "step": 282
    },
    {
      "epoch": 1.0026572187776794,
      "grad_norm": 10.850149154663086,
      "learning_rate": 0.00015907553045637116,
      "loss": 0.0598,
      "step": 283
    },
    {
      "epoch": 1.0062001771479185,
      "grad_norm": 3.1698343753814697,
      "learning_rate": 0.00015864207033612762,
      "loss": 0.0637,
      "step": 284
    },
    {
      "epoch": 1.0097431355181576,
      "grad_norm": 1.1451352834701538,
      "learning_rate": 0.00015820702992940813,
      "loss": 0.0111,
      "step": 285
    },
    {
      "epoch": 1.0132860938883967,
      "grad_norm": 1.3373949527740479,
      "learning_rate": 0.0001577704211630652,
      "loss": 0.0118,
      "step": 286
    },
    {
      "epoch": 1.016829052258636,
      "grad_norm": 3.494053602218628,
      "learning_rate": 0.00015733225600694873,
      "loss": 0.0251,
      "step": 287
    },
    {
      "epoch": 1.0203720106288752,
      "grad_norm": 2.2158610820770264,
      "learning_rate": 0.00015689254647357776,
      "loss": 0.0186,
      "step": 288
    },
    {
      "epoch": 1.0239149689991143,
      "grad_norm": 1.5546177625656128,
      "learning_rate": 0.0001564513046178113,
      "loss": 0.0152,
      "step": 289
    },
    {
      "epoch": 1.0274579273693534,
      "grad_norm": 2.4147849082946777,
      "learning_rate": 0.00015600854253651776,
      "loss": 0.0166,
      "step": 290
    },
    {
      "epoch": 1.0310008857395925,
      "grad_norm": 0.9152072668075562,
      "learning_rate": 0.00015556427236824318,
      "loss": 0.0111,
      "step": 291
    },
    {
      "epoch": 1.0345438441098318,
      "grad_norm": 4.969722270965576,
      "learning_rate": 0.00015511850629287865,
      "loss": 0.0324,
      "step": 292
    },
    {
      "epoch": 1.038086802480071,
      "grad_norm": 3.5214040279388428,
      "learning_rate": 0.00015467125653132637,
      "loss": 0.0211,
      "step": 293
    },
    {
      "epoch": 1.04162976085031,
      "grad_norm": 2.154360771179199,
      "learning_rate": 0.00015422253534516444,
      "loss": 0.0364,
      "step": 294
    },
    {
      "epoch": 1.045172719220549,
      "grad_norm": 2.339195728302002,
      "learning_rate": 0.00015377235503631083,
      "loss": 0.0181,
      "step": 295
    },
    {
      "epoch": 1.0487156775907882,
      "grad_norm": 1.7135610580444336,
      "learning_rate": 0.00015332072794668617,
      "loss": 0.0114,
      "step": 296
    },
    {
      "epoch": 1.0522586359610275,
      "grad_norm": 3.1167871952056885,
      "learning_rate": 0.0001528676664578752,
      "loss": 0.0438,
      "step": 297
    },
    {
      "epoch": 1.0558015943312666,
      "grad_norm": 4.3309006690979,
      "learning_rate": 0.00015241318299078751,
      "loss": 0.0261,
      "step": 298
    },
    {
      "epoch": 1.0593445527015057,
      "grad_norm": 3.1596126556396484,
      "learning_rate": 0.00015195729000531694,
      "loss": 0.0212,
      "step": 299
    },
    {
      "epoch": 1.0628875110717448,
      "grad_norm": 1.7270561456680298,
      "learning_rate": 0.0001515,
      "loss": 0.0243,
      "step": 300
    },
    {
      "epoch": 1.066430469441984,
      "grad_norm": 3.2787768840789795,
      "learning_rate": 0.00015104132551167318,
      "loss": 0.0196,
      "step": 301
    },
    {
      "epoch": 1.0699734278122233,
      "grad_norm": 9.35844898223877,
      "learning_rate": 0.00015058127911512923,
      "loss": 0.1387,
      "step": 302
    },
    {
      "epoch": 1.0735163861824624,
      "grad_norm": 2.0956687927246094,
      "learning_rate": 0.00015011987342277255,
      "loss": 0.0242,
      "step": 303
    },
    {
      "epoch": 1.0770593445527015,
      "grad_norm": 2.920234203338623,
      "learning_rate": 0.00014965712108427323,
      "loss": 0.0267,
      "step": 304
    },
    {
      "epoch": 1.0806023029229406,
      "grad_norm": 1.2659822702407837,
      "learning_rate": 0.00014919303478622045,
      "loss": 0.0183,
      "step": 305
    },
    {
      "epoch": 1.08414526129318,
      "grad_norm": 2.3984029293060303,
      "learning_rate": 0.00014872762725177447,
      "loss": 0.0215,
      "step": 306
    },
    {
      "epoch": 1.087688219663419,
      "grad_norm": 2.2857351303100586,
      "learning_rate": 0.00014826091124031792,
      "loss": 0.023,
      "step": 307
    },
    {
      "epoch": 1.091231178033658,
      "grad_norm": 2.947275400161743,
      "learning_rate": 0.00014779289954710604,
      "loss": 0.0152,
      "step": 308
    },
    {
      "epoch": 1.0947741364038972,
      "grad_norm": 3.4228570461273193,
      "learning_rate": 0.00014732360500291583,
      "loss": 0.0268,
      "step": 309
    },
    {
      "epoch": 1.0983170947741363,
      "grad_norm": 1.8594281673431396,
      "learning_rate": 0.00014685304047369423,
      "loss": 0.0147,
      "step": 310
    },
    {
      "epoch": 1.1018600531443756,
      "grad_norm": 2.166658639907837,
      "learning_rate": 0.00014638121886020555,
      "loss": 0.0189,
      "step": 311
    },
    {
      "epoch": 1.1054030115146147,
      "grad_norm": 2.4011685848236084,
      "learning_rate": 0.00014590815309767767,
      "loss": 0.0148,
      "step": 312
    },
    {
      "epoch": 1.1089459698848538,
      "grad_norm": 2.7356932163238525,
      "learning_rate": 0.00014543385615544744,
      "loss": 0.0199,
      "step": 313
    },
    {
      "epoch": 1.112488928255093,
      "grad_norm": 2.269199848175049,
      "learning_rate": 0.0001449583410366051,
      "loss": 0.0155,
      "step": 314
    },
    {
      "epoch": 1.1160318866253323,
      "grad_norm": 2.468417167663574,
      "learning_rate": 0.00014448162077763783,
      "loss": 0.0127,
      "step": 315
    },
    {
      "epoch": 1.1195748449955714,
      "grad_norm": 1.4242230653762817,
      "learning_rate": 0.00014400370844807234,
      "loss": 0.0105,
      "step": 316
    },
    {
      "epoch": 1.1231178033658105,
      "grad_norm": 1.694300889968872,
      "learning_rate": 0.0001435246171501166,
      "loss": 0.014,
      "step": 317
    },
    {
      "epoch": 1.1266607617360496,
      "grad_norm": 1.0504374504089355,
      "learning_rate": 0.00014304436001830054,
      "loss": 0.0049,
      "step": 318
    },
    {
      "epoch": 1.1302037201062887,
      "grad_norm": 2.216432571411133,
      "learning_rate": 0.000142562950219116,
      "loss": 0.0306,
      "step": 319
    },
    {
      "epoch": 1.133746678476528,
      "grad_norm": 2.4010708332061768,
      "learning_rate": 0.00014208040095065584,
      "loss": 0.0481,
      "step": 320
    },
    {
      "epoch": 1.137289636846767,
      "grad_norm": 1.2632100582122803,
      "learning_rate": 0.000141596725442252,
      "loss": 0.017,
      "step": 321
    },
    {
      "epoch": 1.1408325952170062,
      "grad_norm": 2.802880048751831,
      "learning_rate": 0.00014111193695411285,
      "loss": 0.0138,
      "step": 322
    },
    {
      "epoch": 1.1443755535872453,
      "grad_norm": 1.7126070261001587,
      "learning_rate": 0.00014062604877695972,
      "loss": 0.0137,
      "step": 323
    },
    {
      "epoch": 1.1479185119574846,
      "grad_norm": 2.4996962547302246,
      "learning_rate": 0.0001401390742316624,
      "loss": 0.0295,
      "step": 324
    },
    {
      "epoch": 1.1514614703277237,
      "grad_norm": 1.2408533096313477,
      "learning_rate": 0.00013965102666887408,
      "loss": 0.0102,
      "step": 325
    },
    {
      "epoch": 1.1550044286979628,
      "grad_norm": 0.49025270342826843,
      "learning_rate": 0.0001391619194686652,
      "loss": 0.0068,
      "step": 326
    },
    {
      "epoch": 1.158547387068202,
      "grad_norm": 2.0421504974365234,
      "learning_rate": 0.00013867176604015672,
      "loss": 0.009,
      "step": 327
    },
    {
      "epoch": 1.162090345438441,
      "grad_norm": 1.5926761627197266,
      "learning_rate": 0.0001381805798211525,
      "loss": 0.0191,
      "step": 328
    },
    {
      "epoch": 1.1656333038086801,
      "grad_norm": 2.0150179862976074,
      "learning_rate": 0.00013768837427777082,
      "loss": 0.007,
      "step": 329
    },
    {
      "epoch": 1.1691762621789195,
      "grad_norm": 2.1326754093170166,
      "learning_rate": 0.0001371951629040753,
      "loss": 0.0196,
      "step": 330
    },
    {
      "epoch": 1.1727192205491586,
      "grad_norm": 0.9948328137397766,
      "learning_rate": 0.00013670095922170498,
      "loss": 0.0091,
      "step": 331
    },
    {
      "epoch": 1.1762621789193977,
      "grad_norm": 5.809532642364502,
      "learning_rate": 0.00013620577677950335,
      "loss": 0.0114,
      "step": 332
    },
    {
      "epoch": 1.1798051372896368,
      "grad_norm": 2.064669609069824,
      "learning_rate": 0.00013570962915314725,
      "loss": 0.0245,
      "step": 333
    },
    {
      "epoch": 1.183348095659876,
      "grad_norm": 2.766402244567871,
      "learning_rate": 0.00013521252994477446,
      "loss": 0.0273,
      "step": 334
    },
    {
      "epoch": 1.1868910540301152,
      "grad_norm": 2.8589165210723877,
      "learning_rate": 0.00013471449278261086,
      "loss": 0.0252,
      "step": 335
    },
    {
      "epoch": 1.1904340124003543,
      "grad_norm": 2.930178642272949,
      "learning_rate": 0.0001342155313205969,
      "loss": 0.0105,
      "step": 336
    },
    {
      "epoch": 1.1939769707705934,
      "grad_norm": 2.9018197059631348,
      "learning_rate": 0.0001337156592380131,
      "loss": 0.0213,
      "step": 337
    },
    {
      "epoch": 1.1975199291408325,
      "grad_norm": 1.4895365238189697,
      "learning_rate": 0.00013321489023910508,
      "loss": 0.0149,
      "step": 338
    },
    {
      "epoch": 1.2010628875110718,
      "grad_norm": 2.0806243419647217,
      "learning_rate": 0.0001327132380527079,
      "loss": 0.0118,
      "step": 339
    },
    {
      "epoch": 1.204605845881311,
      "grad_norm": 2.0162320137023926,
      "learning_rate": 0.0001322107164318697,
      "loss": 0.0109,
      "step": 340
    },
    {
      "epoch": 1.20814880425155,
      "grad_norm": 1.5813847780227661,
      "learning_rate": 0.00013170733915347451,
      "loss": 0.0262,
      "step": 341
    },
    {
      "epoch": 1.2116917626217891,
      "grad_norm": 1.8747210502624512,
      "learning_rate": 0.00013120312001786477,
      "loss": 0.0181,
      "step": 342
    },
    {
      "epoch": 1.2152347209920284,
      "grad_norm": 1.9213546514511108,
      "learning_rate": 0.0001306980728484627,
      "loss": 0.0195,
      "step": 343
    },
    {
      "epoch": 1.2187776793622676,
      "grad_norm": 3.8879480361938477,
      "learning_rate": 0.00013019221149139162,
      "loss": 0.0162,
      "step": 344
    },
    {
      "epoch": 1.2223206377325067,
      "grad_norm": 1.8992359638214111,
      "learning_rate": 0.00012968554981509622,
      "loss": 0.017,
      "step": 345
    },
    {
      "epoch": 1.2258635961027458,
      "grad_norm": 1.4245083332061768,
      "learning_rate": 0.00012917810170996218,
      "loss": 0.01,
      "step": 346
    },
    {
      "epoch": 1.2294065544729849,
      "grad_norm": 1.8198248147964478,
      "learning_rate": 0.0001286698810879357,
      "loss": 0.01,
      "step": 347
    },
    {
      "epoch": 1.2329495128432242,
      "grad_norm": 2.0853192806243896,
      "learning_rate": 0.00012816090188214182,
      "loss": 0.0117,
      "step": 348
    },
    {
      "epoch": 1.2364924712134633,
      "grad_norm": 2.3420705795288086,
      "learning_rate": 0.00012765117804650267,
      "loss": 0.014,
      "step": 349
    },
    {
      "epoch": 1.2400354295837024,
      "grad_norm": 1.28701651096344,
      "learning_rate": 0.0001271407235553546,
      "loss": 0.0157,
      "step": 350
    },
    {
      "epoch": 1.2435783879539415,
      "grad_norm": 1.0581600666046143,
      "learning_rate": 0.00012662955240306538,
      "loss": 0.0143,
      "step": 351
    },
    {
      "epoch": 1.2471213463241808,
      "grad_norm": 1.606265664100647,
      "learning_rate": 0.00012611767860365038,
      "loss": 0.0123,
      "step": 352
    },
    {
      "epoch": 1.25066430469442,
      "grad_norm": 1.4095802307128906,
      "learning_rate": 0.00012560511619038827,
      "loss": 0.022,
      "step": 353
    },
    {
      "epoch": 1.254207263064659,
      "grad_norm": 0.7781283855438232,
      "learning_rate": 0.00012509187921543667,
      "loss": 0.0069,
      "step": 354
    },
    {
      "epoch": 1.2577502214348981,
      "grad_norm": 1.0837255716323853,
      "learning_rate": 0.00012457798174944645,
      "loss": 0.0101,
      "step": 355
    },
    {
      "epoch": 1.2612931798051372,
      "grad_norm": 2.2564620971679688,
      "learning_rate": 0.00012406343788117625,
      "loss": 0.0143,
      "step": 356
    },
    {
      "epoch": 1.2648361381753763,
      "grad_norm": 0.644507646560669,
      "learning_rate": 0.0001235482617171061,
      "loss": 0.0086,
      "step": 357
    },
    {
      "epoch": 1.2683790965456156,
      "grad_norm": 0.8635385036468506,
      "learning_rate": 0.00012303246738105082,
      "loss": 0.0087,
      "step": 358
    },
    {
      "epoch": 1.2719220549158547,
      "grad_norm": 1.4721509218215942,
      "learning_rate": 0.00012251606901377265,
      "loss": 0.0117,
      "step": 359
    },
    {
      "epoch": 1.2754650132860939,
      "grad_norm": 0.686865508556366,
      "learning_rate": 0.00012199908077259367,
      "loss": 0.0138,
      "step": 360
    },
    {
      "epoch": 1.2790079716563332,
      "grad_norm": 0.518863320350647,
      "learning_rate": 0.00012148151683100776,
      "loss": 0.004,
      "step": 361
    },
    {
      "epoch": 1.2825509300265723,
      "grad_norm": 1.1533836126327515,
      "learning_rate": 0.00012096339137829174,
      "loss": 0.0065,
      "step": 362
    },
    {
      "epoch": 1.2860938883968114,
      "grad_norm": 1.2232677936553955,
      "learning_rate": 0.00012044471861911666,
      "loss": 0.0176,
      "step": 363
    },
    {
      "epoch": 1.2896368467670505,
      "grad_norm": 1.2730021476745605,
      "learning_rate": 0.0001199255127731582,
      "loss": 0.0066,
      "step": 364
    },
    {
      "epoch": 1.2931798051372896,
      "grad_norm": 1.2388250827789307,
      "learning_rate": 0.00011940578807470692,
      "loss": 0.0135,
      "step": 365
    },
    {
      "epoch": 1.2967227635075287,
      "grad_norm": 1.5637211799621582,
      "learning_rate": 0.00011888555877227793,
      "loss": 0.009,
      "step": 366
    },
    {
      "epoch": 1.300265721877768,
      "grad_norm": 1.2626923322677612,
      "learning_rate": 0.00011836483912822035,
      "loss": 0.0144,
      "step": 367
    },
    {
      "epoch": 1.3038086802480071,
      "grad_norm": 3.295642137527466,
      "learning_rate": 0.00011784364341832634,
      "loss": 0.0206,
      "step": 368
    },
    {
      "epoch": 1.3073516386182462,
      "grad_norm": 1.3101221323013306,
      "learning_rate": 0.00011732198593143949,
      "loss": 0.0116,
      "step": 369
    },
    {
      "epoch": 1.3108945969884853,
      "grad_norm": 1.9212620258331299,
      "learning_rate": 0.00011679988096906333,
      "loss": 0.0204,
      "step": 370
    },
    {
      "epoch": 1.3144375553587246,
      "grad_norm": 2.631624221801758,
      "learning_rate": 0.00011627734284496917,
      "loss": 0.0204,
      "step": 371
    },
    {
      "epoch": 1.3179805137289637,
      "grad_norm": 1.0080429315567017,
      "learning_rate": 0.00011575438588480359,
      "loss": 0.0102,
      "step": 372
    },
    {
      "epoch": 1.3215234720992028,
      "grad_norm": 3.3895442485809326,
      "learning_rate": 0.00011523102442569585,
      "loss": 0.0465,
      "step": 373
    },
    {
      "epoch": 1.325066430469442,
      "grad_norm": 1.0477094650268555,
      "learning_rate": 0.00011470727281586475,
      "loss": 0.007,
      "step": 374
    },
    {
      "epoch": 1.328609388839681,
      "grad_norm": 2.4912171363830566,
      "learning_rate": 0.00011418314541422523,
      "loss": 0.0076,
      "step": 375
    },
    {
      "epoch": 1.3321523472099202,
      "grad_norm": 1.2812281847000122,
      "learning_rate": 0.00011365865658999474,
      "loss": 0.0121,
      "step": 376
    },
    {
      "epoch": 1.3356953055801595,
      "grad_norm": 1.0239572525024414,
      "learning_rate": 0.00011313382072229936,
      "loss": 0.0084,
      "step": 377
    },
    {
      "epoch": 1.3392382639503986,
      "grad_norm": 5.346238136291504,
      "learning_rate": 0.00011260865219977954,
      "loss": 0.0435,
      "step": 378
    },
    {
      "epoch": 1.3427812223206377,
      "grad_norm": 5.027090072631836,
      "learning_rate": 0.00011208316542019556,
      "loss": 0.0328,
      "step": 379
    },
    {
      "epoch": 1.346324180690877,
      "grad_norm": 3.219966411590576,
      "learning_rate": 0.00011155737479003301,
      "loss": 0.0447,
      "step": 380
    },
    {
      "epoch": 1.349867139061116,
      "grad_norm": 1.3265341520309448,
      "learning_rate": 0.00011103129472410755,
      "loss": 0.0106,
      "step": 381
    },
    {
      "epoch": 1.3534100974313552,
      "grad_norm": 3.4696638584136963,
      "learning_rate": 0.00011050493964516997,
      "loss": 0.0057,
      "step": 382
    },
    {
      "epoch": 1.3569530558015943,
      "grad_norm": 1.6994476318359375,
      "learning_rate": 0.00010997832398351062,
      "loss": 0.0063,
      "step": 383
    },
    {
      "epoch": 1.3604960141718334,
      "grad_norm": 2.916574716567993,
      "learning_rate": 0.0001094514621765639,
      "loss": 0.0146,
      "step": 384
    },
    {
      "epoch": 1.3640389725420725,
      "grad_norm": 2.093273162841797,
      "learning_rate": 0.00010892436866851235,
      "loss": 0.0124,
      "step": 385
    },
    {
      "epoch": 1.3675819309123118,
      "grad_norm": 1.3318369388580322,
      "learning_rate": 0.0001083970579098908,
      "loss": 0.006,
      "step": 386
    },
    {
      "epoch": 1.371124889282551,
      "grad_norm": 0.8935598134994507,
      "learning_rate": 0.00010786954435719008,
      "loss": 0.0106,
      "step": 387
    },
    {
      "epoch": 1.37466784765279,
      "grad_norm": 1.9593559503555298,
      "learning_rate": 0.00010734184247246066,
      "loss": 0.0088,
      "step": 388
    },
    {
      "epoch": 1.3782108060230294,
      "grad_norm": 4.444485187530518,
      "learning_rate": 0.00010681396672291631,
      "loss": 0.0241,
      "step": 389
    },
    {
      "epoch": 1.3817537643932685,
      "grad_norm": 5.653156757354736,
      "learning_rate": 0.00010628593158053734,
      "loss": 0.0108,
      "step": 390
    },
    {
      "epoch": 1.3852967227635076,
      "grad_norm": 1.5325078964233398,
      "learning_rate": 0.00010575775152167391,
      "loss": 0.0083,
      "step": 391
    },
    {
      "epoch": 1.3888396811337467,
      "grad_norm": 3.7924163341522217,
      "learning_rate": 0.00010522944102664915,
      "loss": 0.0226,
      "step": 392
    },
    {
      "epoch": 1.3923826395039858,
      "grad_norm": 1.4520882368087769,
      "learning_rate": 0.00010470101457936219,
      "loss": 0.0167,
      "step": 393
    },
    {
      "epoch": 1.3959255978742249,
      "grad_norm": 1.0412254333496094,
      "learning_rate": 0.00010417248666689095,
      "loss": 0.0074,
      "step": 394
    },
    {
      "epoch": 1.3994685562444642,
      "grad_norm": 2.2040746212005615,
      "learning_rate": 0.00010364387177909521,
      "loss": 0.0217,
      "step": 395
    },
    {
      "epoch": 1.4030115146147033,
      "grad_norm": 0.7997134327888489,
      "learning_rate": 0.00010311518440821906,
      "loss": 0.0093,
      "step": 396
    },
    {
      "epoch": 1.4065544729849424,
      "grad_norm": 0.548248827457428,
      "learning_rate": 0.0001025864390484939,
      "loss": 0.0022,
      "step": 397
    },
    {
      "epoch": 1.4100974313551815,
      "grad_norm": 1.4946162700653076,
      "learning_rate": 0.00010205765019574084,
      "loss": 0.0131,
      "step": 398
    },
    {
      "epoch": 1.4136403897254208,
      "grad_norm": 2.3386268615722656,
      "learning_rate": 0.00010152883234697336,
      "loss": 0.0213,
      "step": 399
    },
    {
      "epoch": 1.41718334809566,
      "grad_norm": 1.4776427745819092,
      "learning_rate": 0.000101,
      "loss": 0.0091,
      "step": 400
    },
    {
      "epoch": 1.420726306465899,
      "grad_norm": 1.1784685850143433,
      "learning_rate": 0.00010047116765302661,
      "loss": 0.0085,
      "step": 401
    },
    {
      "epoch": 1.4242692648361381,
      "grad_norm": 1.204289436340332,
      "learning_rate": 9.994234980425921e-05,
      "loss": 0.0081,
      "step": 402
    },
    {
      "epoch": 1.4278122232063772,
      "grad_norm": 0.621965765953064,
      "learning_rate": 9.941356095150613e-05,
      "loss": 0.0058,
      "step": 403
    },
    {
      "epoch": 1.4313551815766163,
      "grad_norm": 4.946881294250488,
      "learning_rate": 9.888481559178096e-05,
      "loss": 0.0193,
      "step": 404
    },
    {
      "epoch": 1.4348981399468557,
      "grad_norm": 0.6689812541007996,
      "learning_rate": 9.835612822090483e-05,
      "loss": 0.0061,
      "step": 405
    },
    {
      "epoch": 1.4384410983170948,
      "grad_norm": 0.9967235922813416,
      "learning_rate": 9.782751333310905e-05,
      "loss": 0.0139,
      "step": 406
    },
    {
      "epoch": 1.4419840566873339,
      "grad_norm": 1.5874682664871216,
      "learning_rate": 9.72989854206378e-05,
      "loss": 0.0125,
      "step": 407
    },
    {
      "epoch": 1.4455270150575732,
      "grad_norm": 0.9432677626609802,
      "learning_rate": 9.677055897335087e-05,
      "loss": 0.0027,
      "step": 408
    },
    {
      "epoch": 1.4490699734278123,
      "grad_norm": 0.9786640405654907,
      "learning_rate": 9.62422484783261e-05,
      "loss": 0.012,
      "step": 409
    },
    {
      "epoch": 1.4526129317980514,
      "grad_norm": 1.9975026845932007,
      "learning_rate": 9.571406841946267e-05,
      "loss": 0.0043,
      "step": 410
    },
    {
      "epoch": 1.4561558901682905,
      "grad_norm": 1.5318927764892578,
      "learning_rate": 9.518603327708372e-05,
      "loss": 0.0092,
      "step": 411
    },
    {
      "epoch": 1.4596988485385296,
      "grad_norm": 1.9841229915618896,
      "learning_rate": 9.465815752753935e-05,
      "loss": 0.0217,
      "step": 412
    },
    {
      "epoch": 1.4632418069087687,
      "grad_norm": 0.8720324039459229,
      "learning_rate": 9.413045564280998e-05,
      "loss": 0.0154,
      "step": 413
    },
    {
      "epoch": 1.466784765279008,
      "grad_norm": 1.511605978012085,
      "learning_rate": 9.360294209010923e-05,
      "loss": 0.0091,
      "step": 414
    },
    {
      "epoch": 1.4703277236492471,
      "grad_norm": 1.48939847946167,
      "learning_rate": 9.307563133148767e-05,
      "loss": 0.0159,
      "step": 415
    },
    {
      "epoch": 1.4738706820194862,
      "grad_norm": 1.8348933458328247,
      "learning_rate": 9.254853782343616e-05,
      "loss": 0.0071,
      "step": 416
    },
    {
      "epoch": 1.4774136403897256,
      "grad_norm": 1.1766817569732666,
      "learning_rate": 9.202167601648942e-05,
      "loss": 0.0064,
      "step": 417
    },
    {
      "epoch": 1.4809565987599647,
      "grad_norm": 0.40942609310150146,
      "learning_rate": 9.149506035483005e-05,
      "loss": 0.0029,
      "step": 418
    },
    {
      "epoch": 1.4844995571302038,
      "grad_norm": 1.4401791095733643,
      "learning_rate": 9.096870527589248e-05,
      "loss": 0.0217,
      "step": 419
    },
    {
      "epoch": 1.4880425155004429,
      "grad_norm": 0.8925746083259583,
      "learning_rate": 9.044262520996702e-05,
      "loss": 0.0093,
      "step": 420
    },
    {
      "epoch": 1.491585473870682,
      "grad_norm": 1.014664888381958,
      "learning_rate": 8.991683457980443e-05,
      "loss": 0.0039,
      "step": 421
    },
    {
      "epoch": 1.495128432240921,
      "grad_norm": 0.7078022956848145,
      "learning_rate": 8.93913478002205e-05,
      "loss": 0.0078,
      "step": 422
    },
    {
      "epoch": 1.4986713906111604,
      "grad_norm": 1.650102138519287,
      "learning_rate": 8.886617927770065e-05,
      "loss": 0.0125,
      "step": 423
    },
    {
      "epoch": 1.5022143489813995,
      "grad_norm": 1.4028496742248535,
      "learning_rate": 8.834134341000527e-05,
      "loss": 0.0102,
      "step": 424
    },
    {
      "epoch": 1.5057573073516386,
      "grad_norm": 1.5076647996902466,
      "learning_rate": 8.781685458577481e-05,
      "loss": 0.0107,
      "step": 425
    },
    {
      "epoch": 1.509300265721878,
      "grad_norm": 1.2729740142822266,
      "learning_rate": 8.729272718413527e-05,
      "loss": 0.0037,
      "step": 426
    },
    {
      "epoch": 1.512843224092117,
      "grad_norm": 1.9747451543807983,
      "learning_rate": 8.676897557430415e-05,
      "loss": 0.0233,
      "step": 427
    },
    {
      "epoch": 1.5163861824623561,
      "grad_norm": 1.1384601593017578,
      "learning_rate": 8.624561411519644e-05,
      "loss": 0.0119,
      "step": 428
    },
    {
      "epoch": 1.5199291408325952,
      "grad_norm": 0.7009539008140564,
      "learning_rate": 8.572265715503086e-05,
      "loss": 0.0022,
      "step": 429
    },
    {
      "epoch": 1.5234720992028343,
      "grad_norm": 0.71835857629776,
      "learning_rate": 8.520011903093666e-05,
      "loss": 0.0088,
      "step": 430
    },
    {
      "epoch": 1.5270150575730734,
      "grad_norm": 1.6711697578430176,
      "learning_rate": 8.467801406856054e-05,
      "loss": 0.0056,
      "step": 431
    },
    {
      "epoch": 1.5305580159433125,
      "grad_norm": 0.5530605912208557,
      "learning_rate": 8.415635658167368e-05,
      "loss": 0.0125,
      "step": 432
    },
    {
      "epoch": 1.5341009743135519,
      "grad_norm": 2.0657296180725098,
      "learning_rate": 8.363516087177962e-05,
      "loss": 0.0268,
      "step": 433
    },
    {
      "epoch": 1.537643932683791,
      "grad_norm": 0.4960460960865021,
      "learning_rate": 8.31144412277221e-05,
      "loss": 0.0037,
      "step": 434
    },
    {
      "epoch": 1.54118689105403,
      "grad_norm": 0.5358054041862488,
      "learning_rate": 8.25942119252931e-05,
      "loss": 0.008,
      "step": 435
    },
    {
      "epoch": 1.5447298494242694,
      "grad_norm": 0.6574037075042725,
      "learning_rate": 8.20744872268418e-05,
      "loss": 0.0088,
      "step": 436
    },
    {
      "epoch": 1.5482728077945085,
      "grad_norm": 1.1722267866134644,
      "learning_rate": 8.155528138088337e-05,
      "loss": 0.0053,
      "step": 437
    },
    {
      "epoch": 1.5518157661647476,
      "grad_norm": 0.7049641013145447,
      "learning_rate": 8.103660862170826e-05,
      "loss": 0.0046,
      "step": 438
    },
    {
      "epoch": 1.5553587245349867,
      "grad_norm": 1.2453325986862183,
      "learning_rate": 8.051848316899227e-05,
      "loss": 0.0196,
      "step": 439
    },
    {
      "epoch": 1.5589016829052258,
      "grad_norm": 0.6382216215133667,
      "learning_rate": 8.000091922740633e-05,
      "loss": 0.0032,
      "step": 440
    },
    {
      "epoch": 1.562444641275465,
      "grad_norm": 0.8369922637939453,
      "learning_rate": 7.948393098622737e-05,
      "loss": 0.0092,
      "step": 441
    },
    {
      "epoch": 1.565987599645704,
      "grad_norm": 1.2144800424575806,
      "learning_rate": 7.896753261894923e-05,
      "loss": 0.009,
      "step": 442
    },
    {
      "epoch": 1.5695305580159433,
      "grad_norm": 0.6583603620529175,
      "learning_rate": 7.845173828289392e-05,
      "loss": 0.0037,
      "step": 443
    },
    {
      "epoch": 1.5730735163861824,
      "grad_norm": 2.239748001098633,
      "learning_rate": 7.793656211882377e-05,
      "loss": 0.0091,
      "step": 444
    },
    {
      "epoch": 1.5766164747564217,
      "grad_norm": 1.1221357583999634,
      "learning_rate": 7.74220182505536e-05,
      "loss": 0.0053,
      "step": 445
    },
    {
      "epoch": 1.5801594331266609,
      "grad_norm": 2.2642343044281006,
      "learning_rate": 7.690812078456336e-05,
      "loss": 0.0318,
      "step": 446
    },
    {
      "epoch": 1.5837023914969,
      "grad_norm": 1.1405282020568848,
      "learning_rate": 7.639488380961173e-05,
      "loss": 0.008,
      "step": 447
    },
    {
      "epoch": 1.587245349867139,
      "grad_norm": 1.2091599702835083,
      "learning_rate": 7.588232139634968e-05,
      "loss": 0.0042,
      "step": 448
    },
    {
      "epoch": 1.5907883082373782,
      "grad_norm": 0.8716428875923157,
      "learning_rate": 7.537044759693463e-05,
      "loss": 0.0076,
      "step": 449
    },
    {
      "epoch": 1.5943312666076173,
      "grad_norm": 1.059916377067566,
      "learning_rate": 7.48592764446454e-05,
      "loss": 0.0135,
      "step": 450
    },
    {
      "epoch": 1.5978742249778564,
      "grad_norm": 3.421222686767578,
      "learning_rate": 7.434882195349736e-05,
      "loss": 0.0065,
      "step": 451
    },
    {
      "epoch": 1.6014171833480957,
      "grad_norm": 1.4951756000518799,
      "learning_rate": 7.383909811785817e-05,
      "loss": 0.0148,
      "step": 452
    },
    {
      "epoch": 1.6049601417183348,
      "grad_norm": 0.8243270516395569,
      "learning_rate": 7.333011891206432e-05,
      "loss": 0.0043,
      "step": 453
    },
    {
      "epoch": 1.6085031000885741,
      "grad_norm": 1.3193808794021606,
      "learning_rate": 7.282189829003785e-05,
      "loss": 0.0073,
      "step": 454
    },
    {
      "epoch": 1.6120460584588132,
      "grad_norm": 0.263898640871048,
      "learning_rate": 7.231445018490381e-05,
      "loss": 0.0017,
      "step": 455
    },
    {
      "epoch": 1.6155890168290523,
      "grad_norm": 1.1934754848480225,
      "learning_rate": 7.180778850860835e-05,
      "loss": 0.0075,
      "step": 456
    },
    {
      "epoch": 1.6191319751992914,
      "grad_norm": 0.7904897332191467,
      "learning_rate": 7.130192715153731e-05,
      "loss": 0.0078,
      "step": 457
    },
    {
      "epoch": 1.6226749335695305,
      "grad_norm": 1.2521803379058838,
      "learning_rate": 7.079687998213526e-05,
      "loss": 0.0093,
      "step": 458
    },
    {
      "epoch": 1.6262178919397696,
      "grad_norm": 0.7100082635879517,
      "learning_rate": 7.029266084652548e-05,
      "loss": 0.0087,
      "step": 459
    },
    {
      "epoch": 1.6297608503100087,
      "grad_norm": 1.4606846570968628,
      "learning_rate": 6.978928356813031e-05,
      "loss": 0.01,
      "step": 460
    },
    {
      "epoch": 1.633303808680248,
      "grad_norm": 1.6624317169189453,
      "learning_rate": 6.92867619472921e-05,
      "loss": 0.0182,
      "step": 461
    },
    {
      "epoch": 1.6368467670504872,
      "grad_norm": 1.014849305152893,
      "learning_rate": 6.878510976089493e-05,
      "loss": 0.0048,
      "step": 462
    },
    {
      "epoch": 1.6403897254207263,
      "grad_norm": 0.7187803387641907,
      "learning_rate": 6.828434076198693e-05,
      "loss": 0.011,
      "step": 463
    },
    {
      "epoch": 1.6439326837909656,
      "grad_norm": 4.439116954803467,
      "learning_rate": 6.77844686794031e-05,
      "loss": 0.0175,
      "step": 464
    },
    {
      "epoch": 1.6474756421612047,
      "grad_norm": 0.4163685142993927,
      "learning_rate": 6.728550721738915e-05,
      "loss": 0.0031,
      "step": 465
    },
    {
      "epoch": 1.6510186005314438,
      "grad_norm": 0.4196591079235077,
      "learning_rate": 6.678747005522557e-05,
      "loss": 0.0052,
      "step": 466
    },
    {
      "epoch": 1.6545615589016829,
      "grad_norm": 0.576309084892273,
      "learning_rate": 6.629037084685278e-05,
      "loss": 0.0075,
      "step": 467
    },
    {
      "epoch": 1.658104517271922,
      "grad_norm": 0.6565279364585876,
      "learning_rate": 6.579422322049668e-05,
      "loss": 0.0025,
      "step": 468
    },
    {
      "epoch": 1.661647475642161,
      "grad_norm": 0.4174266457557678,
      "learning_rate": 6.529904077829505e-05,
      "loss": 0.0044,
      "step": 469
    },
    {
      "epoch": 1.6651904340124002,
      "grad_norm": 0.5562516450881958,
      "learning_rate": 6.480483709592468e-05,
      "loss": 0.0046,
      "step": 470
    },
    {
      "epoch": 1.6687333923826395,
      "grad_norm": 1.159236192703247,
      "learning_rate": 6.43116257222292e-05,
      "loss": 0.0071,
      "step": 471
    },
    {
      "epoch": 1.6722763507528786,
      "grad_norm": 0.38918453454971313,
      "learning_rate": 6.381942017884753e-05,
      "loss": 0.0033,
      "step": 472
    },
    {
      "epoch": 1.675819309123118,
      "grad_norm": 0.512407660484314,
      "learning_rate": 6.33282339598433e-05,
      "loss": 0.0045,
      "step": 473
    },
    {
      "epoch": 1.679362267493357,
      "grad_norm": 0.4515012502670288,
      "learning_rate": 6.283808053133484e-05,
      "loss": 0.0029,
      "step": 474
    },
    {
      "epoch": 1.6829052258635961,
      "grad_norm": 0.7073500752449036,
      "learning_rate": 6.234897333112594e-05,
      "loss": 0.009,
      "step": 475
    },
    {
      "epoch": 1.6864481842338352,
      "grad_norm": 0.5313925743103027,
      "learning_rate": 6.186092576833761e-05,
      "loss": 0.0045,
      "step": 476
    },
    {
      "epoch": 1.6899911426040743,
      "grad_norm": 0.6446258425712585,
      "learning_rate": 6.137395122304033e-05,
      "loss": 0.0022,
      "step": 477
    },
    {
      "epoch": 1.6935341009743134,
      "grad_norm": 0.4338213801383972,
      "learning_rate": 6.088806304588717e-05,
      "loss": 0.0051,
      "step": 478
    },
    {
      "epoch": 1.6970770593445526,
      "grad_norm": 0.6530065536499023,
      "learning_rate": 6.0403274557748035e-05,
      "loss": 0.0042,
      "step": 479
    },
    {
      "epoch": 1.7006200177147919,
      "grad_norm": 0.29270273447036743,
      "learning_rate": 5.9919599049344194e-05,
      "loss": 0.0024,
      "step": 480
    },
    {
      "epoch": 1.704162976085031,
      "grad_norm": 0.4943452775478363,
      "learning_rate": 5.943704978088402e-05,
      "loss": 0.004,
      "step": 481
    },
    {
      "epoch": 1.7077059344552703,
      "grad_norm": 0.6132350564002991,
      "learning_rate": 5.89556399816995e-05,
      "loss": 0.0028,
      "step": 482
    },
    {
      "epoch": 1.7112488928255094,
      "grad_norm": 1.1808321475982666,
      "learning_rate": 5.847538284988341e-05,
      "loss": 0.004,
      "step": 483
    },
    {
      "epoch": 1.7147918511957485,
      "grad_norm": 1.4576940536499023,
      "learning_rate": 5.7996291551927666e-05,
      "loss": 0.0066,
      "step": 484
    },
    {
      "epoch": 1.7183348095659876,
      "grad_norm": 5.167119979858398,
      "learning_rate": 5.751837922236217e-05,
      "loss": 0.0019,
      "step": 485
    },
    {
      "epoch": 1.7218777679362267,
      "grad_norm": 1.4188152551651,
      "learning_rate": 5.704165896339494e-05,
      "loss": 0.009,
      "step": 486
    },
    {
      "epoch": 1.7254207263064658,
      "grad_norm": 1.1834161281585693,
      "learning_rate": 5.656614384455257e-05,
      "loss": 0.007,
      "step": 487
    },
    {
      "epoch": 1.728963684676705,
      "grad_norm": 1.286049723625183,
      "learning_rate": 5.609184690232235e-05,
      "loss": 0.0095,
      "step": 488
    },
    {
      "epoch": 1.7325066430469442,
      "grad_norm": 0.2622765004634857,
      "learning_rate": 5.5618781139794465e-05,
      "loss": 0.0013,
      "step": 489
    },
    {
      "epoch": 1.7360496014171833,
      "grad_norm": 0.5204204320907593,
      "learning_rate": 5.514695952630578e-05,
      "loss": 0.001,
      "step": 490
    },
    {
      "epoch": 1.7395925597874224,
      "grad_norm": 0.2435167282819748,
      "learning_rate": 5.467639499708423e-05,
      "loss": 0.0011,
      "step": 491
    },
    {
      "epoch": 1.7431355181576618,
      "grad_norm": 0.17584297060966492,
      "learning_rate": 5.420710045289399e-05,
      "loss": 0.0015,
      "step": 492
    },
    {
      "epoch": 1.7466784765279009,
      "grad_norm": 0.4891042709350586,
      "learning_rate": 5.373908875968211e-05,
      "loss": 0.0063,
      "step": 493
    },
    {
      "epoch": 1.75022143489814,
      "grad_norm": 0.2887985110282898,
      "learning_rate": 5.3272372748225556e-05,
      "loss": 0.0016,
      "step": 494
    },
    {
      "epoch": 1.753764393268379,
      "grad_norm": 0.8054269552230835,
      "learning_rate": 5.2806965213779544e-05,
      "loss": 0.0015,
      "step": 495
    },
    {
      "epoch": 1.7573073516386182,
      "grad_norm": 0.4256502091884613,
      "learning_rate": 5.234287891572674e-05,
      "loss": 0.0027,
      "step": 496
    },
    {
      "epoch": 1.7608503100088573,
      "grad_norm": 1.075775146484375,
      "learning_rate": 5.1880126577227464e-05,
      "loss": 0.008,
      "step": 497
    },
    {
      "epoch": 1.7643932683790964,
      "grad_norm": 0.8572506308555603,
      "learning_rate": 5.141872088487078e-05,
      "loss": 0.0068,
      "step": 498
    },
    {
      "epoch": 1.7679362267493357,
      "grad_norm": 2.1471755504608154,
      "learning_rate": 5.095867448832683e-05,
      "loss": 0.0086,
      "step": 499
    },
    {
      "epoch": 1.7714791851195748,
      "grad_norm": 0.7083767056465149,
      "learning_rate": 5.050000000000002e-05,
      "loss": 0.0028,
      "step": 500
    },
    {
      "epoch": 1.7750221434898141,
      "grad_norm": 1.5434383153915405,
      "learning_rate": 5.004270999468307e-05,
      "loss": 0.0131,
      "step": 501
    },
    {
      "epoch": 1.7785651018600532,
      "grad_norm": 0.9444709420204163,
      "learning_rate": 4.95868170092125e-05,
      "loss": 0.0045,
      "step": 502
    },
    {
      "epoch": 1.7821080602302923,
      "grad_norm": 0.9909008145332336,
      "learning_rate": 4.913233354212485e-05,
      "loss": 0.0052,
      "step": 503
    },
    {
      "epoch": 1.7856510186005314,
      "grad_norm": 0.7938915491104126,
      "learning_rate": 4.867927205331386e-05,
      "loss": 0.0089,
      "step": 504
    },
    {
      "epoch": 1.7891939769707705,
      "grad_norm": 0.22993983328342438,
      "learning_rate": 4.822764496368917e-05,
      "loss": 0.0051,
      "step": 505
    },
    {
      "epoch": 1.7927369353410096,
      "grad_norm": 1.152287483215332,
      "learning_rate": 4.7777464654835564e-05,
      "loss": 0.0058,
      "step": 506
    },
    {
      "epoch": 1.7962798937112487,
      "grad_norm": 1.2265174388885498,
      "learning_rate": 4.732874346867362e-05,
      "loss": 0.0062,
      "step": 507
    },
    {
      "epoch": 1.799822852081488,
      "grad_norm": 0.5400823354721069,
      "learning_rate": 4.6881493707121315e-05,
      "loss": 0.0017,
      "step": 508
    },
    {
      "epoch": 1.8033658104517272,
      "grad_norm": 0.5171305537223816,
      "learning_rate": 4.643572763175684e-05,
      "loss": 0.0021,
      "step": 509
    },
    {
      "epoch": 1.8069087688219665,
      "grad_norm": 1.48573899269104,
      "learning_rate": 4.5991457463482264e-05,
      "loss": 0.0068,
      "step": 510
    },
    {
      "epoch": 1.8104517271922056,
      "grad_norm": 0.4449637532234192,
      "learning_rate": 4.554869538218868e-05,
      "loss": 0.0039,
      "step": 511
    },
    {
      "epoch": 1.8139946855624447,
      "grad_norm": 0.5684170126914978,
      "learning_rate": 4.5107453526422255e-05,
      "loss": 0.0047,
      "step": 512
    },
    {
      "epoch": 1.8175376439326838,
      "grad_norm": 0.3927152156829834,
      "learning_rate": 4.46677439930513e-05,
      "loss": 0.0023,
      "step": 513
    },
    {
      "epoch": 1.821080602302923,
      "grad_norm": 0.2674574851989746,
      "learning_rate": 4.422957883693483e-05,
      "loss": 0.0013,
      "step": 514
    },
    {
      "epoch": 1.824623560673162,
      "grad_norm": 0.3931225836277008,
      "learning_rate": 4.3792970070591906e-05,
      "loss": 0.0016,
      "step": 515
    },
    {
      "epoch": 1.828166519043401,
      "grad_norm": 0.32513314485549927,
      "learning_rate": 4.3357929663872406e-05,
      "loss": 0.0031,
      "step": 516
    },
    {
      "epoch": 1.8317094774136404,
      "grad_norm": 0.332425981760025,
      "learning_rate": 4.29244695436289e-05,
      "loss": 0.0024,
      "step": 517
    },
    {
      "epoch": 1.8352524357838795,
      "grad_norm": 0.22357267141342163,
      "learning_rate": 4.249260159338946e-05,
      "loss": 0.0006,
      "step": 518
    },
    {
      "epoch": 1.8387953941541186,
      "grad_norm": 0.42479628324508667,
      "learning_rate": 4.2062337653032146e-05,
      "loss": 0.004,
      "step": 519
    },
    {
      "epoch": 1.842338352524358,
      "grad_norm": 1.168453335762024,
      "learning_rate": 4.1633689518460225e-05,
      "loss": 0.0047,
      "step": 520
    },
    {
      "epoch": 1.845881310894597,
      "grad_norm": 0.12201856076717377,
      "learning_rate": 4.1206668941278826e-05,
      "loss": 0.0004,
      "step": 521
    },
    {
      "epoch": 1.8494242692648362,
      "grad_norm": 0.8595163822174072,
      "learning_rate": 4.078128762847279e-05,
      "loss": 0.0055,
      "step": 522
    },
    {
      "epoch": 1.8529672276350753,
      "grad_norm": 0.3519109785556793,
      "learning_rate": 4.035755724208573e-05,
      "loss": 0.0012,
      "step": 523
    },
    {
      "epoch": 1.8565101860053144,
      "grad_norm": 0.5580080151557922,
      "learning_rate": 3.9935489398900145e-05,
      "loss": 0.008,
      "step": 524
    },
    {
      "epoch": 1.8600531443755535,
      "grad_norm": 0.6756648421287537,
      "learning_rate": 3.951509567011922e-05,
      "loss": 0.004,
      "step": 525
    },
    {
      "epoch": 1.8635961027457926,
      "grad_norm": 0.7555375695228577,
      "learning_rate": 3.90963875810494e-05,
      "loss": 0.002,
      "step": 526
    },
    {
      "epoch": 1.867139061116032,
      "grad_norm": 0.052115295082330704,
      "learning_rate": 3.86793766107844e-05,
      "loss": 0.0003,
      "step": 527
    },
    {
      "epoch": 1.870682019486271,
      "grad_norm": 0.7042465806007385,
      "learning_rate": 3.826407419189066e-05,
      "loss": 0.0022,
      "step": 528
    },
    {
      "epoch": 1.8742249778565103,
      "grad_norm": 0.4683879017829895,
      "learning_rate": 3.785049171009381e-05,
      "loss": 0.0024,
      "step": 529
    },
    {
      "epoch": 1.8777679362267494,
      "grad_norm": 0.382080614566803,
      "learning_rate": 3.743864050396644e-05,
      "loss": 0.0013,
      "step": 530
    },
    {
      "epoch": 1.8813108945969885,
      "grad_norm": 0.8347665667533875,
      "learning_rate": 3.7028531864617444e-05,
      "loss": 0.0027,
      "step": 531
    },
    {
      "epoch": 1.8848538529672276,
      "grad_norm": 0.1325908601284027,
      "learning_rate": 3.662017703538234e-05,
      "loss": 0.0007,
      "step": 532
    },
    {
      "epoch": 1.8883968113374667,
      "grad_norm": 0.4601688086986542,
      "learning_rate": 3.621358721151505e-05,
      "loss": 0.0058,
      "step": 533
    },
    {
      "epoch": 1.8919397697077058,
      "grad_norm": 0.00779580045491457,
      "learning_rate": 3.5808773539880973e-05,
      "loss": 0.0001,
      "step": 534
    },
    {
      "epoch": 1.895482728077945,
      "grad_norm": 1.2685234546661377,
      "learning_rate": 3.540574711865146e-05,
      "loss": 0.0066,
      "step": 535
    },
    {
      "epoch": 1.8990256864481843,
      "grad_norm": 0.3840484619140625,
      "learning_rate": 3.500451899699935e-05,
      "loss": 0.0011,
      "step": 536
    },
    {
      "epoch": 1.9025686448184234,
      "grad_norm": 0.6103342175483704,
      "learning_rate": 3.460510017479631e-05,
      "loss": 0.0059,
      "step": 537
    },
    {
      "epoch": 1.9061116031886627,
      "grad_norm": 0.6590524315834045,
      "learning_rate": 3.420750160231118e-05,
      "loss": 0.013,
      "step": 538
    },
    {
      "epoch": 1.9096545615589018,
      "grad_norm": 0.23092423379421234,
      "learning_rate": 3.381173417990957e-05,
      "loss": 0.0015,
      "step": 539
    },
    {
      "epoch": 1.9131975199291409,
      "grad_norm": 0.4215949773788452,
      "learning_rate": 3.3417808757755355e-05,
      "loss": 0.0014,
      "step": 540
    },
    {
      "epoch": 1.91674047829938,
      "grad_norm": 0.04598653316497803,
      "learning_rate": 3.302573613551292e-05,
      "loss": 0.0002,
      "step": 541
    },
    {
      "epoch": 1.920283436669619,
      "grad_norm": 0.2101033478975296,
      "learning_rate": 3.263552706205128e-05,
      "loss": 0.0014,
      "step": 542
    },
    {
      "epoch": 1.9238263950398582,
      "grad_norm": 0.6649401783943176,
      "learning_rate": 3.22471922351493e-05,
      "loss": 0.001,
      "step": 543
    },
    {
      "epoch": 1.9273693534100973,
      "grad_norm": 0.18337276577949524,
      "learning_rate": 3.186074230120244e-05,
      "loss": 0.0008,
      "step": 544
    },
    {
      "epoch": 1.9309123117803366,
      "grad_norm": 0.1808135062456131,
      "learning_rate": 3.147618785493083e-05,
      "loss": 0.0013,
      "step": 545
    },
    {
      "epoch": 1.9344552701505757,
      "grad_norm": 0.5936362743377686,
      "learning_rate": 3.109353943908893e-05,
      "loss": 0.003,
      "step": 546
    },
    {
      "epoch": 1.9379982285208148,
      "grad_norm": 0.5775758624076843,
      "learning_rate": 3.071280754417626e-05,
      "loss": 0.0091,
      "step": 547
    },
    {
      "epoch": 1.9415411868910541,
      "grad_norm": 0.32141023874282837,
      "learning_rate": 3.033400260815008e-05,
      "loss": 0.0007,
      "step": 548
    },
    {
      "epoch": 1.9450841452612933,
      "grad_norm": 0.08689257502555847,
      "learning_rate": 2.9957135016139122e-05,
      "loss": 0.0002,
      "step": 549
    },
    {
      "epoch": 1.9486271036315324,
      "grad_norm": 0.5258713960647583,
      "learning_rate": 2.9582215100158706e-05,
      "loss": 0.0019,
      "step": 550
    },
    {
      "epoch": 1.9521700620017715,
      "grad_norm": 0.3807646632194519,
      "learning_rate": 2.920925313882776e-05,
      "loss": 0.0028,
      "step": 551
    },
    {
      "epoch": 1.9557130203720106,
      "grad_norm": 0.649465799331665,
      "learning_rate": 2.8838259357086884e-05,
      "loss": 0.0034,
      "step": 552
    },
    {
      "epoch": 1.9592559787422497,
      "grad_norm": 0.1853775680065155,
      "learning_rate": 2.846924392591794e-05,
      "loss": 0.0005,
      "step": 553
    },
    {
      "epoch": 1.9627989371124888,
      "grad_norm": 0.03137415647506714,
      "learning_rate": 2.8102216962065423e-05,
      "loss": 0.0001,
      "step": 554
    },
    {
      "epoch": 1.966341895482728,
      "grad_norm": 0.19166480004787445,
      "learning_rate": 2.7737188527758972e-05,
      "loss": 0.0005,
      "step": 555
    },
    {
      "epoch": 1.9698848538529672,
      "grad_norm": 0.9887485504150391,
      "learning_rate": 2.7374168630437456e-05,
      "loss": 0.0018,
      "step": 556
    },
    {
      "epoch": 1.9734278122232065,
      "grad_norm": 0.3358120024204254,
      "learning_rate": 2.7013167222474756e-05,
      "loss": 0.0036,
      "step": 557
    },
    {
      "epoch": 1.9769707705934456,
      "grad_norm": 0.2517526149749756,
      "learning_rate": 2.6654194200906833e-05,
      "loss": 0.0009,
      "step": 558
    },
    {
      "epoch": 1.9805137289636847,
      "grad_norm": 1.207703709602356,
      "learning_rate": 2.629725940716041e-05,
      "loss": 0.0107,
      "step": 559
    },
    {
      "epoch": 1.9840566873339238,
      "grad_norm": 0.2873339354991913,
      "learning_rate": 2.5942372626783172e-05,
      "loss": 0.0023,
      "step": 560
    },
    {
      "epoch": 1.987599645704163,
      "grad_norm": 0.10823265463113785,
      "learning_rate": 2.5589543589175485e-05,
      "loss": 0.0003,
      "step": 561
    },
    {
      "epoch": 1.991142604074402,
      "grad_norm": 0.716543436050415,
      "learning_rate": 2.523878196732358e-05,
      "loss": 0.0042,
      "step": 562
    },
    {
      "epoch": 1.9946855624446411,
      "grad_norm": 0.3027890920639038,
      "learning_rate": 2.489009737753459e-05,
      "loss": 0.0015,
      "step": 563
    },
    {
      "epoch": 1.9982285208148804,
      "grad_norm": 0.32159334421157837,
      "learning_rate": 2.4543499379172615e-05,
      "loss": 0.0016,
      "step": 564
    },
    {
      "epoch": 2.0017714791851198,
      "grad_norm": 0.08274471014738083,
      "learning_rate": 2.4198997474396877e-05,
      "loss": 0.0003,
      "step": 565
    },
    {
      "epoch": 2.005314437555359,
      "grad_norm": 0.060062225908041,
      "learning_rate": 2.3856601107901166e-05,
      "loss": 0.0003,
      "step": 566
    },
    {
      "epoch": 2.008857395925598,
      "grad_norm": 0.1853446364402771,
      "learning_rate": 2.351631966665476e-05,
      "loss": 0.0015,
      "step": 567
    },
    {
      "epoch": 2.012400354295837,
      "grad_norm": 0.0592714287340641,
      "learning_rate": 2.31781624796453e-05,
      "loss": 0.0001,
      "step": 568
    },
    {
      "epoch": 2.015943312666076,
      "grad_norm": 0.13232074677944183,
      "learning_rate": 2.2842138817622883e-05,
      "loss": 0.0005,
      "step": 569
    },
    {
      "epoch": 2.0194862710363153,
      "grad_norm": 0.24643929302692413,
      "learning_rate": 2.250825789284594e-05,
      "loss": 0.0012,
      "step": 570
    },
    {
      "epoch": 2.0230292294065544,
      "grad_norm": 0.17991432547569275,
      "learning_rate": 2.217652885882869e-05,
      "loss": 0.0004,
      "step": 571
    },
    {
      "epoch": 2.0265721877767935,
      "grad_norm": 0.144062802195549,
      "learning_rate": 2.1846960810090188e-05,
      "loss": 0.0005,
      "step": 572
    },
    {
      "epoch": 2.0301151461470326,
      "grad_norm": 0.5604081153869629,
      "learning_rate": 2.151956278190494e-05,
      "loss": 0.0017,
      "step": 573
    },
    {
      "epoch": 2.033658104517272,
      "grad_norm": 0.4855101406574249,
      "learning_rate": 2.119434375005527e-05,
      "loss": 0.0008,
      "step": 574
    },
    {
      "epoch": 2.0372010628875112,
      "grad_norm": 0.9981550574302673,
      "learning_rate": 2.087131263058526e-05,
      "loss": 0.004,
      "step": 575
    },
    {
      "epoch": 2.0407440212577503,
      "grad_norm": 0.12432882189750671,
      "learning_rate": 2.055047827955618e-05,
      "loss": 0.0003,
      "step": 576
    },
    {
      "epoch": 2.0442869796279894,
      "grad_norm": 0.009903721511363983,
      "learning_rate": 2.0231849492803852e-05,
      "loss": 0.0001,
      "step": 577
    },
    {
      "epoch": 2.0478299379982285,
      "grad_norm": 0.044433970004320145,
      "learning_rate": 1.991543500569745e-05,
      "loss": 0.0002,
      "step": 578
    },
    {
      "epoch": 2.0513728963684676,
      "grad_norm": 0.2867472171783447,
      "learning_rate": 1.960124349289992e-05,
      "loss": 0.0019,
      "step": 579
    },
    {
      "epoch": 2.0549158547387067,
      "grad_norm": 0.371579110622406,
      "learning_rate": 1.928928356813032e-05,
      "loss": 0.0012,
      "step": 580
    },
    {
      "epoch": 2.058458813108946,
      "grad_norm": 0.07519207894802094,
      "learning_rate": 1.8979563783927565e-05,
      "loss": 0.0002,
      "step": 581
    },
    {
      "epoch": 2.062001771479185,
      "grad_norm": 0.20066216588020325,
      "learning_rate": 1.8672092631416013e-05,
      "loss": 0.0006,
      "step": 582
    },
    {
      "epoch": 2.065544729849424,
      "grad_norm": 0.19873085618019104,
      "learning_rate": 1.8366878540072614e-05,
      "loss": 0.0008,
      "step": 583
    },
    {
      "epoch": 2.0690876882196636,
      "grad_norm": 0.014256162568926811,
      "learning_rate": 1.8063929877495892e-05,
      "loss": 0.0001,
      "step": 584
    },
    {
      "epoch": 2.0726306465899027,
      "grad_norm": 0.03786277398467064,
      "learning_rate": 1.7763254949176414e-05,
      "loss": 0.0001,
      "step": 585
    },
    {
      "epoch": 2.076173604960142,
      "grad_norm": 0.15015755593776703,
      "learning_rate": 1.7464861998269243e-05,
      "loss": 0.0003,
      "step": 586
    },
    {
      "epoch": 2.079716563330381,
      "grad_norm": 0.3068810999393463,
      "learning_rate": 1.7168759205367893e-05,
      "loss": 0.0003,
      "step": 587
    },
    {
      "epoch": 2.08325952170062,
      "grad_norm": 0.32651999592781067,
      "learning_rate": 1.6874954688279956e-05,
      "loss": 0.0007,
      "step": 588
    },
    {
      "epoch": 2.086802480070859,
      "grad_norm": 0.5359931588172913,
      "learning_rate": 1.6583456501804725e-05,
      "loss": 0.0018,
      "step": 589
    },
    {
      "epoch": 2.090345438441098,
      "grad_norm": 0.08258194476366043,
      "learning_rate": 1.6294272637512183e-05,
      "loss": 0.0003,
      "step": 590
    },
    {
      "epoch": 2.0938883968113373,
      "grad_norm": 0.09279931336641312,
      "learning_rate": 1.600741102352409e-05,
      "loss": 0.0009,
      "step": 591
    },
    {
      "epoch": 2.0974313551815764,
      "grad_norm": 0.7681455016136169,
      "learning_rate": 1.57228795242965e-05,
      "loss": 0.0008,
      "step": 592
    },
    {
      "epoch": 2.100974313551816,
      "grad_norm": 0.5113418698310852,
      "learning_rate": 1.544068594040417e-05,
      "loss": 0.0006,
      "step": 593
    },
    {
      "epoch": 2.104517271922055,
      "grad_norm": 0.11657733470201492,
      "learning_rate": 1.516083800832676e-05,
      "loss": 0.0002,
      "step": 594
    },
    {
      "epoch": 2.108060230292294,
      "grad_norm": 0.06766359508037567,
      "learning_rate": 1.488334340023669e-05,
      "loss": 0.0002,
      "step": 595
    },
    {
      "epoch": 2.1116031886625333,
      "grad_norm": 0.01443282887339592,
      "learning_rate": 1.4608209723788835e-05,
      "loss": 0.0001,
      "step": 596
    },
    {
      "epoch": 2.1151461470327724,
      "grad_norm": 0.005601163953542709,
      "learning_rate": 1.4335444521911899e-05,
      "loss": 0.0,
      "step": 597
    },
    {
      "epoch": 2.1186891054030115,
      "grad_norm": 0.032016199082136154,
      "learning_rate": 1.4065055272601703e-05,
      "loss": 0.0001,
      "step": 598
    },
    {
      "epoch": 2.1222320637732506,
      "grad_norm": 0.04691418632864952,
      "learning_rate": 1.3797049388716065e-05,
      "loss": 0.0001,
      "step": 599
    },
    {
      "epoch": 2.1257750221434897,
      "grad_norm": 0.037374719977378845,
      "learning_rate": 1.3531434217771692e-05,
      "loss": 0.0002,
      "step": 600
    },
    {
      "epoch": 2.129317980513729,
      "grad_norm": 0.0651448667049408,
      "learning_rate": 1.3268217041742701e-05,
      "loss": 0.0001,
      "step": 601
    },
    {
      "epoch": 2.132860938883968,
      "grad_norm": 0.053037695586681366,
      "learning_rate": 1.3007405076860875e-05,
      "loss": 0.0001,
      "step": 602
    },
    {
      "epoch": 2.1364038972542074,
      "grad_norm": 0.056240785866975784,
      "learning_rate": 1.2749005473418015e-05,
      "loss": 0.0002,
      "step": 603
    },
    {
      "epoch": 2.1399468556244465,
      "grad_norm": 0.21008822321891785,
      "learning_rate": 1.2493025315569801e-05,
      "loss": 0.001,
      "step": 604
    },
    {
      "epoch": 2.1434898139946856,
      "grad_norm": 0.1465017944574356,
      "learning_rate": 1.2239471621141508e-05,
      "loss": 0.0005,
      "step": 605
    },
    {
      "epoch": 2.1470327723649247,
      "grad_norm": 0.017339682206511497,
      "learning_rate": 1.1988351341435792e-05,
      "loss": 0.0001,
      "step": 606
    },
    {
      "epoch": 2.150575730735164,
      "grad_norm": 0.5210103392601013,
      "learning_rate": 1.173967136104196e-05,
      "loss": 0.0013,
      "step": 607
    },
    {
      "epoch": 2.154118689105403,
      "grad_norm": 0.004422559402883053,
      "learning_rate": 1.1493438497647313e-05,
      "loss": 0.0,
      "step": 608
    },
    {
      "epoch": 2.157661647475642,
      "grad_norm": 0.03284211456775665,
      "learning_rate": 1.1249659501850155e-05,
      "loss": 0.0002,
      "step": 609
    },
    {
      "epoch": 2.161204605845881,
      "grad_norm": 0.03831435367465019,
      "learning_rate": 1.1008341056974854e-05,
      "loss": 0.0001,
      "step": 610
    },
    {
      "epoch": 2.1647475642161202,
      "grad_norm": 0.15469998121261597,
      "learning_rate": 1.0769489778888405e-05,
      "loss": 0.0003,
      "step": 611
    },
    {
      "epoch": 2.16829052258636,
      "grad_norm": 0.01614706963300705,
      "learning_rate": 1.0533112215819298e-05,
      "loss": 0.0,
      "step": 612
    },
    {
      "epoch": 2.171833480956599,
      "grad_norm": 0.10824459046125412,
      "learning_rate": 1.029921484817783e-05,
      "loss": 0.0004,
      "step": 613
    },
    {
      "epoch": 2.175376439326838,
      "grad_norm": 0.2749079465866089,
      "learning_rate": 1.0067804088378455e-05,
      "loss": 0.0013,
      "step": 614
    },
    {
      "epoch": 2.178919397697077,
      "grad_norm": 0.00494280643761158,
      "learning_rate": 9.8388862806641e-06,
      "loss": 0.0,
      "step": 615
    },
    {
      "epoch": 2.182462356067316,
      "grad_norm": 0.008706189692020416,
      "learning_rate": 9.612467700932045e-06,
      "loss": 0.0001,
      "step": 616
    },
    {
      "epoch": 2.1860053144375553,
      "grad_norm": 0.021209556609392166,
      "learning_rate": 9.388554556562049e-06,
      "loss": 0.0001,
      "step": 617
    },
    {
      "epoch": 2.1895482728077944,
      "grad_norm": 0.6736172437667847,
      "learning_rate": 9.167152986246078e-06,
      "loss": 0.001,
      "step": 618
    },
    {
      "epoch": 2.1930912311780335,
      "grad_norm": 0.09763076156377792,
      "learning_rate": 8.948269059820025e-06,
      "loss": 0.0007,
      "step": 619
    },
    {
      "epoch": 2.1966341895482726,
      "grad_norm": 0.38925012946128845,
      "learning_rate": 8.731908778097302e-06,
      "loss": 0.0025,
      "step": 620
    },
    {
      "epoch": 2.200177147918512,
      "grad_norm": 0.11619476974010468,
      "learning_rate": 8.518078072704338e-06,
      "loss": 0.0005,
      "step": 621
    },
    {
      "epoch": 2.2037201062887513,
      "grad_norm": 0.24150855839252472,
      "learning_rate": 8.306782805917904e-06,
      "loss": 0.0025,
      "step": 622
    },
    {
      "epoch": 2.2072630646589904,
      "grad_norm": 0.04961521923542023,
      "learning_rate": 8.098028770504494e-06,
      "loss": 0.0001,
      "step": 623
    },
    {
      "epoch": 2.2108060230292295,
      "grad_norm": 0.0838765874505043,
      "learning_rate": 7.891821689561459e-06,
      "loss": 0.0006,
      "step": 624
    },
    {
      "epoch": 2.2143489813994686,
      "grad_norm": 0.9362369179725647,
      "learning_rate": 7.68816721636004e-06,
      "loss": 0.0019,
      "step": 625
    },
    {
      "epoch": 2.2178919397697077,
      "grad_norm": 0.03214077651500702,
      "learning_rate": 7.487070934190532e-06,
      "loss": 0.0002,
      "step": 626
    },
    {
      "epoch": 2.2214348981399468,
      "grad_norm": 0.1560969352722168,
      "learning_rate": 7.288538356209092e-06,
      "loss": 0.0003,
      "step": 627
    },
    {
      "epoch": 2.224977856510186,
      "grad_norm": 0.05944027379155159,
      "learning_rate": 7.092574925286614e-06,
      "loss": 0.0002,
      "step": 628
    },
    {
      "epoch": 2.228520814880425,
      "grad_norm": 0.269961416721344,
      "learning_rate": 6.899186013859561e-06,
      "loss": 0.001,
      "step": 629
    },
    {
      "epoch": 2.2320637732506645,
      "grad_norm": 0.005772658158093691,
      "learning_rate": 6.708376923782635e-06,
      "loss": 0.0,
      "step": 630
    },
    {
      "epoch": 2.2356067316209036,
      "grad_norm": 0.0038862484507262707,
      "learning_rate": 6.520152886183406e-06,
      "loss": 0.0,
      "step": 631
    },
    {
      "epoch": 2.2391496899911427,
      "grad_norm": 0.055060192942619324,
      "learning_rate": 6.3345190613189635e-06,
      "loss": 0.0001,
      "step": 632
    },
    {
      "epoch": 2.242692648361382,
      "grad_norm": 0.020991992205381393,
      "learning_rate": 6.151480538434382e-06,
      "loss": 0.0001,
      "step": 633
    },
    {
      "epoch": 2.246235606731621,
      "grad_norm": 0.008830989710986614,
      "learning_rate": 5.971042335623229e-06,
      "loss": 0.0,
      "step": 634
    },
    {
      "epoch": 2.24977856510186,
      "grad_norm": 0.2199835330247879,
      "learning_rate": 5.793209399689978e-06,
      "loss": 0.0004,
      "step": 635
    },
    {
      "epoch": 2.253321523472099,
      "grad_norm": 0.08529902994632721,
      "learning_rate": 5.617986606014419e-06,
      "loss": 0.0004,
      "step": 636
    },
    {
      "epoch": 2.2568644818423382,
      "grad_norm": 0.19139614701271057,
      "learning_rate": 5.445378758417925e-06,
      "loss": 0.0016,
      "step": 637
    },
    {
      "epoch": 2.2604074402125773,
      "grad_norm": 0.009827593341469765,
      "learning_rate": 5.275390589031859e-06,
      "loss": 0.0001,
      "step": 638
    },
    {
      "epoch": 2.263950398582817,
      "grad_norm": 0.1290542483329773,
      "learning_rate": 5.108026758167719e-06,
      "loss": 0.0002,
      "step": 639
    },
    {
      "epoch": 2.267493356953056,
      "grad_norm": 0.17350342869758606,
      "learning_rate": 4.943291854189493e-06,
      "loss": 0.0005,
      "step": 640
    },
    {
      "epoch": 2.271036315323295,
      "grad_norm": 0.3507162928581238,
      "learning_rate": 4.781190393387796e-06,
      "loss": 0.0003,
      "step": 641
    },
    {
      "epoch": 2.274579273693534,
      "grad_norm": 0.02020534686744213,
      "learning_rate": 4.6217268198560404e-06,
      "loss": 0.0001,
      "step": 642
    },
    {
      "epoch": 2.2781222320637733,
      "grad_norm": 0.025796467438340187,
      "learning_rate": 4.464905505368658e-06,
      "loss": 0.0001,
      "step": 643
    },
    {
      "epoch": 2.2816651904340124,
      "grad_norm": 0.07381541281938553,
      "learning_rate": 4.3107307492612086e-06,
      "loss": 0.0003,
      "step": 644
    },
    {
      "epoch": 2.2852081488042515,
      "grad_norm": 0.03729734942317009,
      "learning_rate": 4.1592067783125015e-06,
      "loss": 0.0001,
      "step": 645
    },
    {
      "epoch": 2.2887511071744906,
      "grad_norm": 0.009395123459398746,
      "learning_rate": 4.010337746628751e-06,
      "loss": 0.0,
      "step": 646
    },
    {
      "epoch": 2.2922940655447297,
      "grad_norm": 0.1595831662416458,
      "learning_rate": 3.864127735529656e-06,
      "loss": 0.0004,
      "step": 647
    },
    {
      "epoch": 2.2958370239149692,
      "grad_norm": 0.008252732455730438,
      "learning_rate": 3.7205807534365315e-06,
      "loss": 0.0,
      "step": 648
    },
    {
      "epoch": 2.299379982285208,
      "grad_norm": 0.517535924911499,
      "learning_rate": 3.5797007357623945e-06,
      "loss": 0.0005,
      "step": 649
    },
    {
      "epoch": 2.3029229406554474,
      "grad_norm": 0.012122491374611855,
      "learning_rate": 3.441491544804112e-06,
      "loss": 0.0001,
      "step": 650
    },
    {
      "epoch": 2.3064658990256866,
      "grad_norm": 0.09782838821411133,
      "learning_rate": 3.3059569696364502e-06,
      "loss": 0.0004,
      "step": 651
    },
    {
      "epoch": 2.3100088573959257,
      "grad_norm": 0.19513198733329773,
      "learning_rate": 3.1731007260082616e-06,
      "loss": 0.0012,
      "step": 652
    },
    {
      "epoch": 2.3135518157661648,
      "grad_norm": 0.017000757157802582,
      "learning_rate": 3.0429264562405776e-06,
      "loss": 0.0001,
      "step": 653
    },
    {
      "epoch": 2.317094774136404,
      "grad_norm": 0.0052710892632603645,
      "learning_rate": 2.9154377291267674e-06,
      "loss": 0.0,
      "step": 654
    },
    {
      "epoch": 2.320637732506643,
      "grad_norm": 0.06341676414012909,
      "learning_rate": 2.790638039834668e-06,
      "loss": 0.0002,
      "step": 655
    },
    {
      "epoch": 2.324180690876882,
      "grad_norm": 0.018487611785531044,
      "learning_rate": 2.6685308098108106e-06,
      "loss": 0.0001,
      "step": 656
    },
    {
      "epoch": 2.327723649247121,
      "grad_norm": 0.19819696247577667,
      "learning_rate": 2.5491193866866025e-06,
      "loss": 0.0008,
      "step": 657
    },
    {
      "epoch": 2.3312666076173603,
      "grad_norm": 0.9627290964126587,
      "learning_rate": 2.432407044186509e-06,
      "loss": 0.0032,
      "step": 658
    },
    {
      "epoch": 2.3348095659876,
      "grad_norm": 0.26399293541908264,
      "learning_rate": 2.3183969820383735e-06,
      "loss": 0.0002,
      "step": 659
    },
    {
      "epoch": 2.338352524357839,
      "grad_norm": 0.0048317075707018375,
      "learning_rate": 2.2070923258856255e-06,
      "loss": 0.0,
      "step": 660
    },
    {
      "epoch": 2.341895482728078,
      "grad_norm": 0.019670935347676277,
      "learning_rate": 2.098496127201648e-06,
      "loss": 0.0001,
      "step": 661
    },
    {
      "epoch": 2.345438441098317,
      "grad_norm": 0.0395568385720253,
      "learning_rate": 1.992611363206103e-06,
      "loss": 0.0001,
      "step": 662
    },
    {
      "epoch": 2.348981399468556,
      "grad_norm": 0.2325470745563507,
      "learning_rate": 1.889440936783242e-06,
      "loss": 0.0026,
      "step": 663
    },
    {
      "epoch": 2.3525243578387953,
      "grad_norm": 0.019612450152635574,
      "learning_rate": 1.7889876764024505e-06,
      "loss": 0.0001,
      "step": 664
    },
    {
      "epoch": 2.3560673162090344,
      "grad_norm": 0.11476591974496841,
      "learning_rate": 1.691254336040595e-06,
      "loss": 0.0004,
      "step": 665
    },
    {
      "epoch": 2.3596102745792735,
      "grad_norm": 0.022656535729765892,
      "learning_rate": 1.59624359510657e-06,
      "loss": 0.0001,
      "step": 666
    },
    {
      "epoch": 2.3631532329495126,
      "grad_norm": 0.04269755259156227,
      "learning_rate": 1.5039580583678393e-06,
      "loss": 0.0003,
      "step": 667
    },
    {
      "epoch": 2.366696191319752,
      "grad_norm": 0.6594712734222412,
      "learning_rate": 1.414400255879008e-06,
      "loss": 0.0011,
      "step": 668
    },
    {
      "epoch": 2.3702391496899913,
      "grad_norm": 0.10037083178758621,
      "learning_rate": 1.327572642912468e-06,
      "loss": 0.0002,
      "step": 669
    },
    {
      "epoch": 2.3737821080602304,
      "grad_norm": 0.21993687748908997,
      "learning_rate": 1.2434775998910964e-06,
      "loss": 0.0002,
      "step": 670
    },
    {
      "epoch": 2.3773250664304695,
      "grad_norm": 0.018617277964949608,
      "learning_rate": 1.1621174323229612e-06,
      "loss": 0.0001,
      "step": 671
    },
    {
      "epoch": 2.3808680248007086,
      "grad_norm": 0.019156964495778084,
      "learning_rate": 1.0834943707381784e-06,
      "loss": 0.0001,
      "step": 672
    },
    {
      "epoch": 2.3844109831709477,
      "grad_norm": 0.017957083880901337,
      "learning_rate": 1.0076105706276888e-06,
      "loss": 0.0001,
      "step": 673
    },
    {
      "epoch": 2.387953941541187,
      "grad_norm": 0.0515175499022007,
      "learning_rate": 9.344681123841967e-07,
      "loss": 0.0002,
      "step": 674
    },
    {
      "epoch": 2.391496899911426,
      "grad_norm": 0.05943327769637108,
      "learning_rate": 8.640690012451515e-07,
      "loss": 0.0004,
      "step": 675
    },
    {
      "epoch": 2.395039858281665,
      "grad_norm": 0.24372032284736633,
      "learning_rate": 7.964151672377458e-07,
      "loss": 0.0015,
      "step": 676
    },
    {
      "epoch": 2.3985828166519045,
      "grad_norm": 0.07449764758348465,
      "learning_rate": 7.315084651260009e-07,
      "loss": 0.0002,
      "step": 677
    },
    {
      "epoch": 2.4021257750221436,
      "grad_norm": 0.021373214200139046,
      "learning_rate": 6.69350674359959e-07,
      "loss": 0.0001,
      "step": 678
    },
    {
      "epoch": 2.4056687333923827,
      "grad_norm": 0.008149687200784683,
      "learning_rate": 6.099434990268609e-07,
      "loss": 0.0001,
      "step": 679
    },
    {
      "epoch": 2.409211691762622,
      "grad_norm": 0.10549762845039368,
      "learning_rate": 5.532885678043977e-07,
      "loss": 0.0003,
      "step": 680
    },
    {
      "epoch": 2.412754650132861,
      "grad_norm": 0.08827023208141327,
      "learning_rate": 4.9938743391615e-07,
      "loss": 0.0003,
      "step": 681
    },
    {
      "epoch": 2.4162976085031,
      "grad_norm": 0.011174162849783897,
      "learning_rate": 4.482415750889204e-07,
      "loss": 0.0,
      "step": 682
    },
    {
      "epoch": 2.419840566873339,
      "grad_norm": 0.0026113265193998814,
      "learning_rate": 3.998523935122772e-07,
      "loss": 0.0,
      "step": 683
    },
    {
      "epoch": 2.4233835252435783,
      "grad_norm": 0.03312141075730324,
      "learning_rate": 3.5422121580005864e-07,
      "loss": 0.0001,
      "step": 684
    },
    {
      "epoch": 2.4269264836138174,
      "grad_norm": 0.08379507064819336,
      "learning_rate": 3.1134929295407564e-07,
      "loss": 0.0003,
      "step": 685
    },
    {
      "epoch": 2.430469441984057,
      "grad_norm": 0.025159165263175964,
      "learning_rate": 2.7123780032973235e-07,
      "loss": 0.0001,
      "step": 686
    },
    {
      "epoch": 2.434012400354296,
      "grad_norm": 0.025299694389104843,
      "learning_rate": 2.3388783760386601e-07,
      "loss": 0.0001,
      "step": 687
    },
    {
      "epoch": 2.437555358724535,
      "grad_norm": 0.6148785352706909,
      "learning_rate": 1.9930042874457254e-07,
      "loss": 0.0163,
      "step": 688
    },
    {
      "epoch": 2.441098317094774,
      "grad_norm": 0.0018735540797933936,
      "learning_rate": 1.6747652198313957e-07,
      "loss": 0.0,
      "step": 689
    },
    {
      "epoch": 2.4446412754650133,
      "grad_norm": 0.11185324192047119,
      "learning_rate": 1.3841698978804285e-07,
      "loss": 0.0003,
      "step": 690
    },
    {
      "epoch": 2.4481842338352524,
      "grad_norm": 0.47679388523101807,
      "learning_rate": 1.1212262884103974e-07,
      "loss": 0.001,
      "step": 691
    },
    {
      "epoch": 2.4517271922054915,
      "grad_norm": 0.017726508900523186,
      "learning_rate": 8.85941600153033e-08,
      "loss": 0.0001,
      "step": 692
    },
    {
      "epoch": 2.4552701505757306,
      "grad_norm": 0.40522223711013794,
      "learning_rate": 6.783222835572055e-08,
      "loss": 0.0018,
      "step": 693
    },
    {
      "epoch": 2.4588131089459697,
      "grad_norm": 0.01428899448364973,
      "learning_rate": 4.98374030611084e-08,
      "loss": 0.0,
      "step": 694
    },
    {
      "epoch": 2.4623560673162093,
      "grad_norm": 1.708270788192749,
      "learning_rate": 3.461017746871675e-08,
      "loss": 0.0083,
      "step": 695
    },
    {
      "epoch": 2.4658990256864484,
      "grad_norm": 0.04217607527971268,
      "learning_rate": 2.215096904060454e-08,
      "loss": 0.0001,
      "step": 696
    },
    {
      "epoch": 2.4694419840566875,
      "grad_norm": 0.007940036244690418,
      "learning_rate": 1.246011935228064e-08,
      "loss": 0.0,
      "step": 697
    },
    {
      "epoch": 2.4729849424269266,
      "grad_norm": 0.14349356293678284,
      "learning_rate": 5.537894083273543e-09,
      "loss": 0.0007,
      "step": 698
    },
    {
      "epoch": 2.4765279007971657,
      "grad_norm": 0.017691675573587418,
      "learning_rate": 1.384483009898796e-09,
      "loss": 0.0001,
      "step": 699
    },
    {
      "epoch": 2.4800708591674048,
      "grad_norm": 0.371564656496048,
      "learning_rate": 0.0,
      "loss": 0.0009,
      "step": 700
    }
  ],
  "logging_steps": 1,
  "max_steps": 700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.77336688291532e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}