amburger66's picture
LoRA fine-tune on RobotSmith real world task02 reaching
5d1adb4 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.08602150537634409,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 8.602150537634408e-05,
"grad_norm": 26.13152503967285,
"learning_rate": 0.0,
"loss": 1.9334,
"step": 1
},
{
"epoch": 0.00017204301075268816,
"grad_norm": 12.308059692382812,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.8172,
"step": 2
},
{
"epoch": 0.00025806451612903227,
"grad_norm": 5.769585609436035,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.2786,
"step": 3
},
{
"epoch": 0.0003440860215053763,
"grad_norm": 15.32481861114502,
"learning_rate": 6.000000000000001e-07,
"loss": 1.5367,
"step": 4
},
{
"epoch": 0.00043010752688172043,
"grad_norm": 7.713790416717529,
"learning_rate": 8.000000000000001e-07,
"loss": 0.9053,
"step": 5
},
{
"epoch": 0.0005161290322580645,
"grad_norm": 10.92131519317627,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.2591,
"step": 6
},
{
"epoch": 0.0006021505376344086,
"grad_norm": 14.062760353088379,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.1718,
"step": 7
},
{
"epoch": 0.0006881720430107526,
"grad_norm": 12.738030433654785,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.416,
"step": 8
},
{
"epoch": 0.0007741935483870968,
"grad_norm": 7.097315311431885,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.9348,
"step": 9
},
{
"epoch": 0.0008602150537634409,
"grad_norm": 9.602299690246582,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.6063,
"step": 10
},
{
"epoch": 0.000946236559139785,
"grad_norm": 5.977272987365723,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3435,
"step": 11
},
{
"epoch": 0.001032258064516129,
"grad_norm": 9.875030517578125,
"learning_rate": 2.2e-06,
"loss": 0.809,
"step": 12
},
{
"epoch": 0.001118279569892473,
"grad_norm": 15.130753517150879,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.0979,
"step": 13
},
{
"epoch": 0.0012043010752688173,
"grad_norm": 10.204551696777344,
"learning_rate": 2.6e-06,
"loss": 0.5257,
"step": 14
},
{
"epoch": 0.0012903225806451613,
"grad_norm": 6.346858501434326,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.4956,
"step": 15
},
{
"epoch": 0.0013763440860215053,
"grad_norm": 5.456563949584961,
"learning_rate": 3e-06,
"loss": 0.3942,
"step": 16
},
{
"epoch": 0.0014623655913978495,
"grad_norm": 8.51014232635498,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.6383,
"step": 17
},
{
"epoch": 0.0015483870967741935,
"grad_norm": 10.209907531738281,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.0255,
"step": 18
},
{
"epoch": 0.0016344086021505377,
"grad_norm": 21.3107852935791,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.3,
"step": 19
},
{
"epoch": 0.0017204301075268817,
"grad_norm": 5.358498573303223,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.3525,
"step": 20
},
{
"epoch": 0.0018064516129032257,
"grad_norm": 17.639127731323242,
"learning_rate": 4.000000000000001e-06,
"loss": 1.3008,
"step": 21
},
{
"epoch": 0.00189247311827957,
"grad_norm": 12.497448921203613,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.3028,
"step": 22
},
{
"epoch": 0.001978494623655914,
"grad_norm": 10.120162010192871,
"learning_rate": 4.4e-06,
"loss": 0.52,
"step": 23
},
{
"epoch": 0.002064516129032258,
"grad_norm": 5.192323684692383,
"learning_rate": 4.600000000000001e-06,
"loss": 0.3929,
"step": 24
},
{
"epoch": 0.002150537634408602,
"grad_norm": 5.372901439666748,
"learning_rate": 4.800000000000001e-06,
"loss": 0.4032,
"step": 25
},
{
"epoch": 0.002236559139784946,
"grad_norm": 20.389095306396484,
"learning_rate": 5e-06,
"loss": 1.4283,
"step": 26
},
{
"epoch": 0.00232258064516129,
"grad_norm": 7.011532306671143,
"learning_rate": 5.2e-06,
"loss": 0.2889,
"step": 27
},
{
"epoch": 0.0024086021505376346,
"grad_norm": 5.003702640533447,
"learning_rate": 5.400000000000001e-06,
"loss": 0.3988,
"step": 28
},
{
"epoch": 0.0024946236559139786,
"grad_norm": 14.830155372619629,
"learning_rate": 5.600000000000001e-06,
"loss": 1.4922,
"step": 29
},
{
"epoch": 0.0025806451612903226,
"grad_norm": 9.541350364685059,
"learning_rate": 5.8e-06,
"loss": 0.77,
"step": 30
},
{
"epoch": 0.0026666666666666666,
"grad_norm": 6.875884532928467,
"learning_rate": 6e-06,
"loss": 0.667,
"step": 31
},
{
"epoch": 0.0027526881720430106,
"grad_norm": 15.122583389282227,
"learning_rate": 6.200000000000001e-06,
"loss": 1.7123,
"step": 32
},
{
"epoch": 0.002838709677419355,
"grad_norm": 6.474270820617676,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.4232,
"step": 33
},
{
"epoch": 0.002924731182795699,
"grad_norm": 10.105985641479492,
"learning_rate": 6.600000000000001e-06,
"loss": 0.6005,
"step": 34
},
{
"epoch": 0.003010752688172043,
"grad_norm": 4.490959167480469,
"learning_rate": 6.800000000000001e-06,
"loss": 0.3575,
"step": 35
},
{
"epoch": 0.003096774193548387,
"grad_norm": 9.392435073852539,
"learning_rate": 7e-06,
"loss": 0.3663,
"step": 36
},
{
"epoch": 0.003182795698924731,
"grad_norm": 9.383217811584473,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.8748,
"step": 37
},
{
"epoch": 0.0032688172043010754,
"grad_norm": 8.383386611938477,
"learning_rate": 7.4e-06,
"loss": 0.7348,
"step": 38
},
{
"epoch": 0.0033548387096774194,
"grad_norm": 6.199663162231445,
"learning_rate": 7.600000000000001e-06,
"loss": 0.4644,
"step": 39
},
{
"epoch": 0.0034408602150537634,
"grad_norm": 9.572957992553711,
"learning_rate": 7.800000000000002e-06,
"loss": 0.976,
"step": 40
},
{
"epoch": 0.0035268817204301074,
"grad_norm": 9.266552925109863,
"learning_rate": 8.000000000000001e-06,
"loss": 0.7412,
"step": 41
},
{
"epoch": 0.0036129032258064514,
"grad_norm": 9.287247657775879,
"learning_rate": 8.2e-06,
"loss": 0.537,
"step": 42
},
{
"epoch": 0.003698924731182796,
"grad_norm": 6.134600639343262,
"learning_rate": 8.400000000000001e-06,
"loss": 0.3733,
"step": 43
},
{
"epoch": 0.00378494623655914,
"grad_norm": 9.29589557647705,
"learning_rate": 8.6e-06,
"loss": 0.5229,
"step": 44
},
{
"epoch": 0.003870967741935484,
"grad_norm": 4.620382308959961,
"learning_rate": 8.8e-06,
"loss": 0.1926,
"step": 45
},
{
"epoch": 0.003956989247311828,
"grad_norm": 9.61703872680664,
"learning_rate": 9e-06,
"loss": 0.8871,
"step": 46
},
{
"epoch": 0.004043010752688172,
"grad_norm": 6.898414134979248,
"learning_rate": 9.200000000000002e-06,
"loss": 0.5726,
"step": 47
},
{
"epoch": 0.004129032258064516,
"grad_norm": 8.194817543029785,
"learning_rate": 9.4e-06,
"loss": 0.4448,
"step": 48
},
{
"epoch": 0.00421505376344086,
"grad_norm": 12.40204906463623,
"learning_rate": 9.600000000000001e-06,
"loss": 0.4421,
"step": 49
},
{
"epoch": 0.004301075268817204,
"grad_norm": 10.6871337890625,
"learning_rate": 9.800000000000001e-06,
"loss": 0.8173,
"step": 50
},
{
"epoch": 0.004387096774193549,
"grad_norm": 11.042154312133789,
"learning_rate": 1e-05,
"loss": 0.8205,
"step": 51
},
{
"epoch": 0.004473118279569892,
"grad_norm": 6.640468120574951,
"learning_rate": 1.02e-05,
"loss": 0.307,
"step": 52
},
{
"epoch": 0.004559139784946237,
"grad_norm": 6.990267276763916,
"learning_rate": 1.04e-05,
"loss": 0.5263,
"step": 53
},
{
"epoch": 0.00464516129032258,
"grad_norm": 3.496103525161743,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.2546,
"step": 54
},
{
"epoch": 0.004731182795698925,
"grad_norm": 3.179828405380249,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.2393,
"step": 55
},
{
"epoch": 0.004817204301075269,
"grad_norm": 11.997007369995117,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.8857,
"step": 56
},
{
"epoch": 0.004903225806451613,
"grad_norm": 6.33434534072876,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.4108,
"step": 57
},
{
"epoch": 0.004989247311827957,
"grad_norm": 5.966276168823242,
"learning_rate": 1.14e-05,
"loss": 0.3833,
"step": 58
},
{
"epoch": 0.005075268817204301,
"grad_norm": 7.068697929382324,
"learning_rate": 1.16e-05,
"loss": 0.4835,
"step": 59
},
{
"epoch": 0.005161290322580645,
"grad_norm": 5.894850730895996,
"learning_rate": 1.18e-05,
"loss": 0.3125,
"step": 60
},
{
"epoch": 0.00524731182795699,
"grad_norm": 2.986966848373413,
"learning_rate": 1.2e-05,
"loss": 0.2675,
"step": 61
},
{
"epoch": 0.005333333333333333,
"grad_norm": 6.198732376098633,
"learning_rate": 1.22e-05,
"loss": 0.3122,
"step": 62
},
{
"epoch": 0.005419354838709678,
"grad_norm": 6.080263614654541,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.3036,
"step": 63
},
{
"epoch": 0.005505376344086021,
"grad_norm": 7.634750843048096,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.4761,
"step": 64
},
{
"epoch": 0.005591397849462366,
"grad_norm": 12.790021896362305,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.4377,
"step": 65
},
{
"epoch": 0.00567741935483871,
"grad_norm": 12.705551147460938,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.5418,
"step": 66
},
{
"epoch": 0.005763440860215054,
"grad_norm": 4.59014368057251,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.1789,
"step": 67
},
{
"epoch": 0.005849462365591398,
"grad_norm": 5.9041829109191895,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.2809,
"step": 68
},
{
"epoch": 0.005935483870967742,
"grad_norm": 7.209511756896973,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.4706,
"step": 69
},
{
"epoch": 0.006021505376344086,
"grad_norm": 6.8474602699279785,
"learning_rate": 1.38e-05,
"loss": 0.3197,
"step": 70
},
{
"epoch": 0.0061075268817204304,
"grad_norm": 3.906287908554077,
"learning_rate": 1.4e-05,
"loss": 0.2689,
"step": 71
},
{
"epoch": 0.006193548387096774,
"grad_norm": 4.254907131195068,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.2714,
"step": 72
},
{
"epoch": 0.0062795698924731184,
"grad_norm": 8.126392364501953,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.547,
"step": 73
},
{
"epoch": 0.006365591397849462,
"grad_norm": 5.2069525718688965,
"learning_rate": 1.46e-05,
"loss": 0.2573,
"step": 74
},
{
"epoch": 0.0064516129032258064,
"grad_norm": 2.18611741065979,
"learning_rate": 1.48e-05,
"loss": 0.176,
"step": 75
},
{
"epoch": 0.006537634408602151,
"grad_norm": 5.081924915313721,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.2773,
"step": 76
},
{
"epoch": 0.0066236559139784944,
"grad_norm": 3.5612785816192627,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.2679,
"step": 77
},
{
"epoch": 0.006709677419354839,
"grad_norm": 8.817394256591797,
"learning_rate": 1.54e-05,
"loss": 0.5226,
"step": 78
},
{
"epoch": 0.0067956989247311824,
"grad_norm": 2.6454527378082275,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.1483,
"step": 79
},
{
"epoch": 0.006881720430107527,
"grad_norm": 6.943172454833984,
"learning_rate": 1.58e-05,
"loss": 0.5444,
"step": 80
},
{
"epoch": 0.006967741935483871,
"grad_norm": 2.6451799869537354,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.1383,
"step": 81
},
{
"epoch": 0.007053763440860215,
"grad_norm": 6.003849029541016,
"learning_rate": 1.62e-05,
"loss": 0.5836,
"step": 82
},
{
"epoch": 0.007139784946236559,
"grad_norm": 1.6628965139389038,
"learning_rate": 1.64e-05,
"loss": 0.0621,
"step": 83
},
{
"epoch": 0.007225806451612903,
"grad_norm": 5.394771099090576,
"learning_rate": 1.66e-05,
"loss": 0.1605,
"step": 84
},
{
"epoch": 0.007311827956989247,
"grad_norm": 5.427704334259033,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.1553,
"step": 85
},
{
"epoch": 0.007397849462365592,
"grad_norm": 2.023010492324829,
"learning_rate": 1.7e-05,
"loss": 0.1294,
"step": 86
},
{
"epoch": 0.007483870967741935,
"grad_norm": 1.4516578912734985,
"learning_rate": 1.72e-05,
"loss": 0.0285,
"step": 87
},
{
"epoch": 0.00756989247311828,
"grad_norm": 4.412249565124512,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.2444,
"step": 88
},
{
"epoch": 0.007655913978494623,
"grad_norm": 3.5322859287261963,
"learning_rate": 1.76e-05,
"loss": 0.2281,
"step": 89
},
{
"epoch": 0.007741935483870968,
"grad_norm": 3.9227123260498047,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.2811,
"step": 90
},
{
"epoch": 0.007827956989247311,
"grad_norm": 2.3604328632354736,
"learning_rate": 1.8e-05,
"loss": 0.1989,
"step": 91
},
{
"epoch": 0.007913978494623657,
"grad_norm": 1.5397543907165527,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.1202,
"step": 92
},
{
"epoch": 0.008,
"grad_norm": 1.258698582649231,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.0977,
"step": 93
},
{
"epoch": 0.008086021505376344,
"grad_norm": 11.321172714233398,
"learning_rate": 1.86e-05,
"loss": 0.5742,
"step": 94
},
{
"epoch": 0.008172043010752689,
"grad_norm": 3.0486249923706055,
"learning_rate": 1.88e-05,
"loss": 0.1873,
"step": 95
},
{
"epoch": 0.008258064516129033,
"grad_norm": 5.444565773010254,
"learning_rate": 1.9e-05,
"loss": 0.6733,
"step": 96
},
{
"epoch": 0.008344086021505376,
"grad_norm": 2.4698333740234375,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.1483,
"step": 97
},
{
"epoch": 0.00843010752688172,
"grad_norm": 1.0990469455718994,
"learning_rate": 1.94e-05,
"loss": 0.0938,
"step": 98
},
{
"epoch": 0.008516129032258065,
"grad_norm": 19.15937042236328,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.9282,
"step": 99
},
{
"epoch": 0.008602150537634409,
"grad_norm": 1.6818559169769287,
"learning_rate": 1.98e-05,
"loss": 0.111,
"step": 100
},
{
"epoch": 0.008688172043010752,
"grad_norm": 2.062930107116699,
"learning_rate": 2e-05,
"loss": 0.1039,
"step": 101
},
{
"epoch": 0.008774193548387097,
"grad_norm": 1.6798441410064697,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.0848,
"step": 102
},
{
"epoch": 0.008860215053763441,
"grad_norm": 2.2785284519195557,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.136,
"step": 103
},
{
"epoch": 0.008946236559139785,
"grad_norm": 3.9607315063476562,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.144,
"step": 104
},
{
"epoch": 0.00903225806451613,
"grad_norm": 9.754727363586426,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.5352,
"step": 105
},
{
"epoch": 0.009118279569892473,
"grad_norm": 0.6848426461219788,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.0459,
"step": 106
},
{
"epoch": 0.009204301075268817,
"grad_norm": 9.159880638122559,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.2678,
"step": 107
},
{
"epoch": 0.00929032258064516,
"grad_norm": 1.5956013202667236,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.0979,
"step": 108
},
{
"epoch": 0.009376344086021506,
"grad_norm": 2.8924038410186768,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.2754,
"step": 109
},
{
"epoch": 0.00946236559139785,
"grad_norm": 0.6897431015968323,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.0132,
"step": 110
},
{
"epoch": 0.009548387096774193,
"grad_norm": 4.700838565826416,
"learning_rate": 1.999390827019096e-05,
"loss": 0.2679,
"step": 111
},
{
"epoch": 0.009634408602150538,
"grad_norm": 2.364711284637451,
"learning_rate": 1.999262916410621e-05,
"loss": 0.0943,
"step": 112
},
{
"epoch": 0.009720430107526882,
"grad_norm": 1.4274086952209473,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.1167,
"step": 113
},
{
"epoch": 0.009806451612903225,
"grad_norm": 6.709775447845459,
"learning_rate": 1.998970569790715e-05,
"loss": 0.752,
"step": 114
},
{
"epoch": 0.00989247311827957,
"grad_norm": 2.03126859664917,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.1144,
"step": 115
},
{
"epoch": 0.009978494623655914,
"grad_norm": 3.7494125366210938,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.1408,
"step": 116
},
{
"epoch": 0.010064516129032258,
"grad_norm": 1.9576760530471802,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1212,
"step": 117
},
{
"epoch": 0.010150537634408601,
"grad_norm": 6.121198654174805,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.5683,
"step": 118
},
{
"epoch": 0.010236559139784947,
"grad_norm": 8.39349365234375,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.4357,
"step": 119
},
{
"epoch": 0.01032258064516129,
"grad_norm": 4.752889156341553,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.1863,
"step": 120
},
{
"epoch": 0.010408602150537634,
"grad_norm": 1.3134794235229492,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.0887,
"step": 121
},
{
"epoch": 0.01049462365591398,
"grad_norm": 1.7149567604064941,
"learning_rate": 1.997314477224458e-05,
"loss": 0.1672,
"step": 122
},
{
"epoch": 0.010580645161290323,
"grad_norm": 2.789108991622925,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1178,
"step": 123
},
{
"epoch": 0.010666666666666666,
"grad_norm": 0.29963165521621704,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.0055,
"step": 124
},
{
"epoch": 0.010752688172043012,
"grad_norm": 8.834779739379883,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.2324,
"step": 125
},
{
"epoch": 0.010838709677419355,
"grad_norm": 3.4565048217773438,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.0792,
"step": 126
},
{
"epoch": 0.010924731182795699,
"grad_norm": 2.795158624649048,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.1156,
"step": 127
},
{
"epoch": 0.011010752688172042,
"grad_norm": 1.5416563749313354,
"learning_rate": 1.99556196460308e-05,
"loss": 0.1166,
"step": 128
},
{
"epoch": 0.011096774193548388,
"grad_norm": 11.173118591308594,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.2825,
"step": 129
},
{
"epoch": 0.011182795698924731,
"grad_norm": 0.791736364364624,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.074,
"step": 130
},
{
"epoch": 0.011268817204301075,
"grad_norm": 2.8545353412628174,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.1841,
"step": 131
},
{
"epoch": 0.01135483870967742,
"grad_norm": 3.312082529067993,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.2529,
"step": 132
},
{
"epoch": 0.011440860215053764,
"grad_norm": 4.397865295410156,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.0579,
"step": 133
},
{
"epoch": 0.011526881720430107,
"grad_norm": 1.5323413610458374,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.0352,
"step": 134
},
{
"epoch": 0.011612903225806452,
"grad_norm": 2.2003939151763916,
"learning_rate": 1.992965508106537e-05,
"loss": 0.081,
"step": 135
},
{
"epoch": 0.011698924731182796,
"grad_norm": 2.3644931316375732,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.1151,
"step": 136
},
{
"epoch": 0.01178494623655914,
"grad_norm": 3.616933584213257,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.2247,
"step": 137
},
{
"epoch": 0.011870967741935483,
"grad_norm": 3.2906737327575684,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.1754,
"step": 138
},
{
"epoch": 0.011956989247311828,
"grad_norm": 2.3769936561584473,
"learning_rate": 1.991215540251542e-05,
"loss": 0.1575,
"step": 139
},
{
"epoch": 0.012043010752688172,
"grad_norm": 13.670448303222656,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.2348,
"step": 140
},
{
"epoch": 0.012129032258064516,
"grad_norm": 5.556253433227539,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.1438,
"step": 141
},
{
"epoch": 0.012215053763440861,
"grad_norm": 4.107612609863281,
"learning_rate": 1.989776230907789e-05,
"loss": 0.1341,
"step": 142
},
{
"epoch": 0.012301075268817204,
"grad_norm": 2.5551977157592773,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1329,
"step": 143
},
{
"epoch": 0.012387096774193548,
"grad_norm": 1.317958116531372,
"learning_rate": 1.988756381047006e-05,
"loss": 0.1073,
"step": 144
},
{
"epoch": 0.012473118279569893,
"grad_norm": 11.63919734954834,
"learning_rate": 1.988228381446553e-05,
"loss": 0.2892,
"step": 145
},
{
"epoch": 0.012559139784946237,
"grad_norm": 0.17914678156375885,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.004,
"step": 146
},
{
"epoch": 0.01264516129032258,
"grad_norm": 2.019073009490967,
"learning_rate": 1.987136265072988e-05,
"loss": 0.0881,
"step": 147
},
{
"epoch": 0.012731182795698924,
"grad_norm": 16.270954132080078,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.1936,
"step": 148
},
{
"epoch": 0.01281720430107527,
"grad_norm": 5.838150978088379,
"learning_rate": 1.985996037070505e-05,
"loss": 0.0482,
"step": 149
},
{
"epoch": 0.012903225806451613,
"grad_norm": 2.0501770973205566,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.0904,
"step": 150
},
{
"epoch": 0.012989247311827956,
"grad_norm": 1.1647562980651855,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.0931,
"step": 151
},
{
"epoch": 0.013075268817204302,
"grad_norm": 0.5622814297676086,
"learning_rate": 1.984195607969242e-05,
"loss": 0.0377,
"step": 152
},
{
"epoch": 0.013161290322580645,
"grad_norm": 1.6728405952453613,
"learning_rate": 1.983571470813386e-05,
"loss": 0.0993,
"step": 153
},
{
"epoch": 0.013247311827956989,
"grad_norm": 8.484197616577148,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.265,
"step": 154
},
{
"epoch": 0.013333333333333334,
"grad_norm": 1.1448677778244019,
"learning_rate": 1.982287250728689e-05,
"loss": 0.051,
"step": 155
},
{
"epoch": 0.013419354838709678,
"grad_norm": 3.8463077545166016,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.1084,
"step": 156
},
{
"epoch": 0.013505376344086021,
"grad_norm": 1.0600382089614868,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.0914,
"step": 157
},
{
"epoch": 0.013591397849462365,
"grad_norm": 0.9652776122093201,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.0537,
"step": 158
},
{
"epoch": 0.01367741935483871,
"grad_norm": 1.5798885822296143,
"learning_rate": 1.979575249599344e-05,
"loss": 0.0576,
"step": 159
},
{
"epoch": 0.013763440860215054,
"grad_norm": 3.926121950149536,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.1463,
"step": 160
},
{
"epoch": 0.013849462365591397,
"grad_norm": 0.9733953475952148,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.0823,
"step": 161
},
{
"epoch": 0.013935483870967743,
"grad_norm": 15.145951271057129,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.2998,
"step": 162
},
{
"epoch": 0.014021505376344086,
"grad_norm": 2.3089499473571777,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.1374,
"step": 163
},
{
"epoch": 0.01410752688172043,
"grad_norm": 4.4648895263671875,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.101,
"step": 164
},
{
"epoch": 0.014193548387096775,
"grad_norm": 7.4126152992248535,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.1396,
"step": 165
},
{
"epoch": 0.014279569892473119,
"grad_norm": 1.0811303853988647,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.0486,
"step": 166
},
{
"epoch": 0.014365591397849462,
"grad_norm": 6.099431037902832,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.1715,
"step": 167
},
{
"epoch": 0.014451612903225806,
"grad_norm": 10.455945014953613,
"learning_rate": 1.972775878209397e-05,
"loss": 0.6786,
"step": 168
},
{
"epoch": 0.014537634408602151,
"grad_norm": 5.669197082519531,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.2454,
"step": 169
},
{
"epoch": 0.014623655913978495,
"grad_norm": 7.427839279174805,
"learning_rate": 1.971134279909636e-05,
"loss": 0.1512,
"step": 170
},
{
"epoch": 0.014709677419354838,
"grad_norm": 12.540517807006836,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.4858,
"step": 171
},
{
"epoch": 0.014795698924731183,
"grad_norm": 8.503724098205566,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.1673,
"step": 172
},
{
"epoch": 0.014881720430107527,
"grad_norm": 0.4525596499443054,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.036,
"step": 173
},
{
"epoch": 0.01496774193548387,
"grad_norm": 1.3618741035461426,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.0494,
"step": 174
},
{
"epoch": 0.015053763440860216,
"grad_norm": 1.278050184249878,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.1177,
"step": 175
},
{
"epoch": 0.01513978494623656,
"grad_norm": 2.4488964080810547,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.1521,
"step": 176
},
{
"epoch": 0.015225806451612903,
"grad_norm": 1.580557107925415,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.082,
"step": 177
},
{
"epoch": 0.015311827956989247,
"grad_norm": 0.9505684971809387,
"learning_rate": 1.96409540423411e-05,
"loss": 0.0589,
"step": 178
},
{
"epoch": 0.015397849462365592,
"grad_norm": 12.36058235168457,
"learning_rate": 1.9631625667976584e-05,
"loss": 1.2206,
"step": 179
},
{
"epoch": 0.015483870967741935,
"grad_norm": 0.4550093114376068,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.057,
"step": 180
},
{
"epoch": 0.015569892473118279,
"grad_norm": 0.6297346949577332,
"learning_rate": 1.961261695938319e-05,
"loss": 0.0367,
"step": 181
},
{
"epoch": 0.015655913978494623,
"grad_norm": 1.9497625827789307,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.1119,
"step": 182
},
{
"epoch": 0.015741935483870966,
"grad_norm": 1.0460658073425293,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.073,
"step": 183
},
{
"epoch": 0.015827956989247313,
"grad_norm": 7.174893856048584,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.0942,
"step": 184
},
{
"epoch": 0.015913978494623657,
"grad_norm": 0.34728294610977173,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.0025,
"step": 185
},
{
"epoch": 0.016,
"grad_norm": 13.906850814819336,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.3912,
"step": 186
},
{
"epoch": 0.016086021505376344,
"grad_norm": 0.7030720114707947,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.0673,
"step": 187
},
{
"epoch": 0.016172043010752687,
"grad_norm": 1.1968787908554077,
"learning_rate": 1.954240328516277e-05,
"loss": 0.0795,
"step": 188
},
{
"epoch": 0.01625806451612903,
"grad_norm": 0.7793338298797607,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.0618,
"step": 189
},
{
"epoch": 0.016344086021505378,
"grad_norm": 1.7036898136138916,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.09,
"step": 190
},
{
"epoch": 0.01643010752688172,
"grad_norm": 7.776175022125244,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.3275,
"step": 191
},
{
"epoch": 0.016516129032258065,
"grad_norm": 0.7343016862869263,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.0523,
"step": 192
},
{
"epoch": 0.01660215053763441,
"grad_norm": 0.8714408278465271,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.0643,
"step": 193
},
{
"epoch": 0.016688172043010752,
"grad_norm": 0.9974550008773804,
"learning_rate": 1.947768410009586e-05,
"loss": 0.0801,
"step": 194
},
{
"epoch": 0.016774193548387096,
"grad_norm": 3.0199031829833984,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.1322,
"step": 195
},
{
"epoch": 0.01686021505376344,
"grad_norm": 1.3036985397338867,
"learning_rate": 1.945518575599317e-05,
"loss": 0.0785,
"step": 196
},
{
"epoch": 0.016946236559139786,
"grad_norm": 2.6375234127044678,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0499,
"step": 197
},
{
"epoch": 0.01703225806451613,
"grad_norm": 1.7725839614868164,
"learning_rate": 1.943222657947601e-05,
"loss": 0.0754,
"step": 198
},
{
"epoch": 0.017118279569892474,
"grad_norm": 6.20302677154541,
"learning_rate": 1.942057452787297e-05,
"loss": 0.0933,
"step": 199
},
{
"epoch": 0.017204301075268817,
"grad_norm": 1.3552567958831787,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0556,
"step": 200
},
{
"epoch": 0.01729032258064516,
"grad_norm": 0.5432722568511963,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.0109,
"step": 201
},
{
"epoch": 0.017376344086021504,
"grad_norm": 0.1880321502685547,
"learning_rate": 1.938493022759556e-05,
"loss": 0.002,
"step": 202
},
{
"epoch": 0.017462365591397848,
"grad_norm": 1.4618467092514038,
"learning_rate": 1.937281989491892e-05,
"loss": 0.1144,
"step": 203
},
{
"epoch": 0.017548387096774195,
"grad_norm": 1.386118769645691,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.0894,
"step": 204
},
{
"epoch": 0.01763440860215054,
"grad_norm": 0.4225877523422241,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.0124,
"step": 205
},
{
"epoch": 0.017720430107526882,
"grad_norm": 8.801976203918457,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.1658,
"step": 206
},
{
"epoch": 0.017806451612903226,
"grad_norm": 17.360960006713867,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.5391,
"step": 207
},
{
"epoch": 0.01789247311827957,
"grad_norm": 1.4718023538589478,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.1041,
"step": 208
},
{
"epoch": 0.017978494623655913,
"grad_norm": 2.3638532161712646,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.0196,
"step": 209
},
{
"epoch": 0.01806451612903226,
"grad_norm": 3.099367141723633,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.0729,
"step": 210
},
{
"epoch": 0.018150537634408603,
"grad_norm": 10.05279541015625,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.1395,
"step": 211
},
{
"epoch": 0.018236559139784947,
"grad_norm": 6.7756547927856445,
"learning_rate": 1.925870584809995e-05,
"loss": 0.1824,
"step": 212
},
{
"epoch": 0.01832258064516129,
"grad_norm": 1.9750202894210815,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.1094,
"step": 213
},
{
"epoch": 0.018408602150537634,
"grad_norm": 0.8687901496887207,
"learning_rate": 1.923210217112981e-05,
"loss": 0.0666,
"step": 214
},
{
"epoch": 0.018494623655913978,
"grad_norm": 0.823858380317688,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.063,
"step": 215
},
{
"epoch": 0.01858064516129032,
"grad_norm": 1.7087424993515015,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.1478,
"step": 216
},
{
"epoch": 0.018666666666666668,
"grad_norm": 0.3196397125720978,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.0056,
"step": 217
},
{
"epoch": 0.018752688172043012,
"grad_norm": 1.2545502185821533,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.0441,
"step": 218
},
{
"epoch": 0.018838709677419355,
"grad_norm": 1.5336922407150269,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.0495,
"step": 219
},
{
"epoch": 0.0189247311827957,
"grad_norm": 2.144378662109375,
"learning_rate": 1.914959667849825e-05,
"loss": 0.1613,
"step": 220
},
{
"epoch": 0.019010752688172042,
"grad_norm": 1.5358353853225708,
"learning_rate": 1.913545457642601e-05,
"loss": 0.1424,
"step": 221
},
{
"epoch": 0.019096774193548386,
"grad_norm": 0.7785419225692749,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.0548,
"step": 222
},
{
"epoch": 0.01918279569892473,
"grad_norm": 1.4889520406723022,
"learning_rate": 1.910683660806177e-05,
"loss": 0.0796,
"step": 223
},
{
"epoch": 0.019268817204301077,
"grad_norm": 2.869166135787964,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.082,
"step": 224
},
{
"epoch": 0.01935483870967742,
"grad_norm": 2.2500557899475098,
"learning_rate": 1.907777478532909e-05,
"loss": 0.0727,
"step": 225
},
{
"epoch": 0.019440860215053764,
"grad_norm": 16.548152923583984,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.4665,
"step": 226
},
{
"epoch": 0.019526881720430107,
"grad_norm": 18.88007164001465,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.2271,
"step": 227
},
{
"epoch": 0.01961290322580645,
"grad_norm": 1.0665011405944824,
"learning_rate": 1.903335292863301e-05,
"loss": 0.0609,
"step": 228
},
{
"epoch": 0.019698924731182794,
"grad_norm": 1.0349040031433105,
"learning_rate": 1.901832526405114e-05,
"loss": 0.036,
"step": 229
},
{
"epoch": 0.01978494623655914,
"grad_norm": 1.1308789253234863,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0226,
"step": 230
},
{
"epoch": 0.019870967741935485,
"grad_norm": 1.4125988483428955,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0232,
"step": 231
},
{
"epoch": 0.01995698924731183,
"grad_norm": 0.5785567760467529,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0532,
"step": 232
},
{
"epoch": 0.020043010752688172,
"grad_norm": 9.862194061279297,
"learning_rate": 1.895711760239413e-05,
"loss": 0.1466,
"step": 233
},
{
"epoch": 0.020129032258064516,
"grad_norm": 13.802783966064453,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.4622,
"step": 234
},
{
"epoch": 0.02021505376344086,
"grad_norm": 10.582741737365723,
"learning_rate": 1.892585818452126e-05,
"loss": 0.8757,
"step": 235
},
{
"epoch": 0.020301075268817203,
"grad_norm": 5.641706943511963,
"learning_rate": 1.891006524188368e-05,
"loss": 0.376,
"step": 236
},
{
"epoch": 0.02038709677419355,
"grad_norm": 5.477917671203613,
"learning_rate": 1.889416373291298e-05,
"loss": 0.0701,
"step": 237
},
{
"epoch": 0.020473118279569893,
"grad_norm": 1.678926944732666,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.08,
"step": 238
},
{
"epoch": 0.020559139784946237,
"grad_norm": 1.2497512102127075,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.0632,
"step": 239
},
{
"epoch": 0.02064516129032258,
"grad_norm": 5.993684768676758,
"learning_rate": 1.884580975215084e-05,
"loss": 0.0438,
"step": 240
},
{
"epoch": 0.020731182795698924,
"grad_norm": 16.17313575744629,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.1304,
"step": 241
},
{
"epoch": 0.020817204301075268,
"grad_norm": 0.5322088599205017,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.0021,
"step": 242
},
{
"epoch": 0.02090322580645161,
"grad_norm": 4.3276848793029785,
"learning_rate": 1.879648572866617e-05,
"loss": 0.1046,
"step": 243
},
{
"epoch": 0.02098924731182796,
"grad_norm": 0.5298567414283752,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.0582,
"step": 244
},
{
"epoch": 0.021075268817204302,
"grad_norm": 26.181636810302734,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.4294,
"step": 245
},
{
"epoch": 0.021161290322580645,
"grad_norm": 4.328470706939697,
"learning_rate": 1.874619707139396e-05,
"loss": 0.0375,
"step": 246
},
{
"epoch": 0.02124731182795699,
"grad_norm": 24.821365356445312,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.2591,
"step": 247
},
{
"epoch": 0.021333333333333333,
"grad_norm": 0.639571487903595,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0206,
"step": 248
},
{
"epoch": 0.021419354838709676,
"grad_norm": 1.795882225036621,
"learning_rate": 1.869494929505219e-05,
"loss": 0.0727,
"step": 249
},
{
"epoch": 0.021505376344086023,
"grad_norm": 4.359352111816406,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.1199,
"step": 250
},
{
"epoch": 0.021591397849462367,
"grad_norm": 5.080965042114258,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0737,
"step": 251
},
{
"epoch": 0.02167741935483871,
"grad_norm": 1.3430877923965454,
"learning_rate": 1.864274801953705e-05,
"loss": 0.0592,
"step": 252
},
{
"epoch": 0.021763440860215054,
"grad_norm": 1.503000020980835,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.0444,
"step": 253
},
{
"epoch": 0.021849462365591397,
"grad_norm": 4.173727035522461,
"learning_rate": 1.860742027003944e-05,
"loss": 0.149,
"step": 254
},
{
"epoch": 0.02193548387096774,
"grad_norm": 2.155229091644287,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.0325,
"step": 255
},
{
"epoch": 0.022021505376344085,
"grad_norm": 2.8601109981536865,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.0689,
"step": 256
},
{
"epoch": 0.02210752688172043,
"grad_norm": 1.7606431245803833,
"learning_rate": 1.855364260160507e-05,
"loss": 0.0874,
"step": 257
},
{
"epoch": 0.022193548387096775,
"grad_norm": 6.3607964515686035,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.1592,
"step": 258
},
{
"epoch": 0.02227956989247312,
"grad_norm": 29.71542739868164,
"learning_rate": 1.851726934143048e-05,
"loss": 0.1125,
"step": 259
},
{
"epoch": 0.022365591397849462,
"grad_norm": 10.089420318603516,
"learning_rate": 1.849892692986864e-05,
"loss": 0.0695,
"step": 260
},
{
"epoch": 0.022451612903225806,
"grad_norm": 2.771125078201294,
"learning_rate": 1.848048096156426e-05,
"loss": 0.0925,
"step": 261
},
{
"epoch": 0.02253763440860215,
"grad_norm": 1.3765859603881836,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.019,
"step": 262
},
{
"epoch": 0.022623655913978493,
"grad_norm": 2.852144479751587,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.0965,
"step": 263
},
{
"epoch": 0.02270967741935484,
"grad_norm": 1.1639028787612915,
"learning_rate": 1.842452397007148e-05,
"loss": 0.0459,
"step": 264
},
{
"epoch": 0.022795698924731184,
"grad_norm": 21.226099014282227,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.417,
"step": 265
},
{
"epoch": 0.022881720430107527,
"grad_norm": 1.7253221273422241,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.0747,
"step": 266
},
{
"epoch": 0.02296774193548387,
"grad_norm": 25.369264602661133,
"learning_rate": 1.836764313458962e-05,
"loss": 0.1782,
"step": 267
},
{
"epoch": 0.023053763440860214,
"grad_norm": 2.1202876567840576,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.1006,
"step": 268
},
{
"epoch": 0.023139784946236558,
"grad_norm": 1.3634047508239746,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.0777,
"step": 269
},
{
"epoch": 0.023225806451612905,
"grad_norm": 1.1996127367019653,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.0834,
"step": 270
},
{
"epoch": 0.02331182795698925,
"grad_norm": 16.200428009033203,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.1316,
"step": 271
},
{
"epoch": 0.023397849462365592,
"grad_norm": 0.9208644032478333,
"learning_rate": 1.827080574274562e-05,
"loss": 0.0827,
"step": 272
},
{
"epoch": 0.023483870967741936,
"grad_norm": 1.376083493232727,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.0117,
"step": 273
},
{
"epoch": 0.02356989247311828,
"grad_norm": 31.145734786987305,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1955,
"step": 274
},
{
"epoch": 0.023655913978494623,
"grad_norm": 1.3080297708511353,
"learning_rate": 1.821149209133704e-05,
"loss": 0.0593,
"step": 275
},
{
"epoch": 0.023741935483870966,
"grad_norm": 1.437355399131775,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0623,
"step": 276
},
{
"epoch": 0.023827956989247313,
"grad_norm": 2.2659268379211426,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.0412,
"step": 277
},
{
"epoch": 0.023913978494623657,
"grad_norm": 0.5584107041358948,
"learning_rate": 1.815127795728554e-05,
"loss": 0.0334,
"step": 278
},
{
"epoch": 0.024,
"grad_norm": 0.9149103164672852,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.0625,
"step": 279
},
{
"epoch": 0.024086021505376344,
"grad_norm": 0.21168604493141174,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.001,
"step": 280
},
{
"epoch": 0.024172043010752688,
"grad_norm": 0.28735315799713135,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0009,
"step": 281
},
{
"epoch": 0.02425806451612903,
"grad_norm": 5.056240558624268,
"learning_rate": 1.806960312143802e-05,
"loss": 0.0928,
"step": 282
},
{
"epoch": 0.024344086021505375,
"grad_norm": 15.460404396057129,
"learning_rate": 1.804893797355914e-05,
"loss": 0.4325,
"step": 283
},
{
"epoch": 0.024430107526881722,
"grad_norm": 171.36404418945312,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.4924,
"step": 284
},
{
"epoch": 0.024516129032258065,
"grad_norm": 0.531282901763916,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.0467,
"step": 285
},
{
"epoch": 0.02460215053763441,
"grad_norm": 0.18217715620994568,
"learning_rate": 1.798635510047293e-05,
"loss": 0.001,
"step": 286
},
{
"epoch": 0.024688172043010752,
"grad_norm": 1.6189045906066895,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0694,
"step": 287
},
{
"epoch": 0.024774193548387096,
"grad_norm": 0.8191723227500916,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.0487,
"step": 288
},
{
"epoch": 0.02486021505376344,
"grad_norm": 0.6113877296447754,
"learning_rate": 1.792289643355191e-05,
"loss": 0.0532,
"step": 289
},
{
"epoch": 0.024946236559139787,
"grad_norm": 0.856691837310791,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0483,
"step": 290
},
{
"epoch": 0.02503225806451613,
"grad_norm": 1.3021553754806519,
"learning_rate": 1.788010753606722e-05,
"loss": 0.0473,
"step": 291
},
{
"epoch": 0.025118279569892474,
"grad_norm": 1.4489821195602417,
"learning_rate": 1.785856893175402e-05,
"loss": 0.091,
"step": 292
},
{
"epoch": 0.025204301075268817,
"grad_norm": 1.8007919788360596,
"learning_rate": 1.78369345732584e-05,
"loss": 0.093,
"step": 293
},
{
"epoch": 0.02529032258064516,
"grad_norm": 0.210431769490242,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0015,
"step": 294
},
{
"epoch": 0.025376344086021504,
"grad_norm": 1.2820943593978882,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.0715,
"step": 295
},
{
"epoch": 0.025462365591397848,
"grad_norm": 0.6511804461479187,
"learning_rate": 1.777145961456971e-05,
"loss": 0.0476,
"step": 296
},
{
"epoch": 0.025548387096774195,
"grad_norm": 0.1344589740037918,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.0017,
"step": 297
},
{
"epoch": 0.02563440860215054,
"grad_norm": 0.40938836336135864,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.012,
"step": 298
},
{
"epoch": 0.025720430107526882,
"grad_norm": 0.05873025581240654,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.0012,
"step": 299
},
{
"epoch": 0.025806451612903226,
"grad_norm": 9.811320304870605,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.1022,
"step": 300
},
{
"epoch": 0.02589247311827957,
"grad_norm": 12.113297462463379,
"learning_rate": 1.766044443118978e-05,
"loss": 0.5176,
"step": 301
},
{
"epoch": 0.025978494623655913,
"grad_norm": 6.134032726287842,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.0597,
"step": 302
},
{
"epoch": 0.026064516129032256,
"grad_norm": 1.1919665336608887,
"learning_rate": 1.761538307536737e-05,
"loss": 0.0758,
"step": 303
},
{
"epoch": 0.026150537634408604,
"grad_norm": 1.3780089616775513,
"learning_rate": 1.759271307334881e-05,
"loss": 0.0713,
"step": 304
},
{
"epoch": 0.026236559139784947,
"grad_norm": 1.070957064628601,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0642,
"step": 305
},
{
"epoch": 0.02632258064516129,
"grad_norm": 0.8633624315261841,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0531,
"step": 306
},
{
"epoch": 0.026408602150537634,
"grad_norm": 10.743792533874512,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.1123,
"step": 307
},
{
"epoch": 0.026494623655913978,
"grad_norm": 13.648173332214355,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.1435,
"step": 308
},
{
"epoch": 0.02658064516129032,
"grad_norm": 1.3393679857254028,
"learning_rate": 1.747798090498532e-05,
"loss": 0.0772,
"step": 309
},
{
"epoch": 0.02666666666666667,
"grad_norm": 0.5580834150314331,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.0336,
"step": 310
},
{
"epoch": 0.026752688172043012,
"grad_norm": 14.132169723510742,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.5032,
"step": 311
},
{
"epoch": 0.026838709677419356,
"grad_norm": 17.469453811645508,
"learning_rate": 1.74080459628675e-05,
"loss": 0.1998,
"step": 312
},
{
"epoch": 0.0269247311827957,
"grad_norm": 0.7411155700683594,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.0288,
"step": 313
},
{
"epoch": 0.027010752688172043,
"grad_norm": 2.3204100131988525,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.0203,
"step": 314
},
{
"epoch": 0.027096774193548386,
"grad_norm": 2.6321027278900146,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.1212,
"step": 315
},
{
"epoch": 0.02718279569892473,
"grad_norm": 0.7601708173751831,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.0534,
"step": 316
},
{
"epoch": 0.027268817204301077,
"grad_norm": 6.11430025100708,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.042,
"step": 317
},
{
"epoch": 0.02735483870967742,
"grad_norm": 0.3483811914920807,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.0341,
"step": 318
},
{
"epoch": 0.027440860215053764,
"grad_norm": 14.013130187988281,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.4467,
"step": 319
},
{
"epoch": 0.027526881720430108,
"grad_norm": 0.8994920253753662,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.0294,
"step": 320
},
{
"epoch": 0.02761290322580645,
"grad_norm": 4.1907477378845215,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.0982,
"step": 321
},
{
"epoch": 0.027698924731182795,
"grad_norm": 0.8169042468070984,
"learning_rate": 1.716910607650483e-05,
"loss": 0.0495,
"step": 322
},
{
"epoch": 0.027784946236559138,
"grad_norm": 0.7887750267982483,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.0519,
"step": 323
},
{
"epoch": 0.027870967741935485,
"grad_norm": 1.678784728050232,
"learning_rate": 1.712026045990997e-05,
"loss": 0.0742,
"step": 324
},
{
"epoch": 0.02795698924731183,
"grad_norm": 0.22454331815242767,
"learning_rate": 1.709570736536521e-05,
"loss": 0.0002,
"step": 325
},
{
"epoch": 0.028043010752688172,
"grad_norm": 1.7833493947982788,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.0546,
"step": 326
},
{
"epoch": 0.028129032258064516,
"grad_norm": 1.4376364946365356,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.0478,
"step": 327
},
{
"epoch": 0.02821505376344086,
"grad_norm": 9.607670783996582,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.1383,
"step": 328
},
{
"epoch": 0.028301075268817203,
"grad_norm": 0.12965631484985352,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0006,
"step": 329
},
{
"epoch": 0.02838709677419355,
"grad_norm": 0.47277605533599854,
"learning_rate": 1.697165102854565e-05,
"loss": 0.0513,
"step": 330
},
{
"epoch": 0.028473118279569894,
"grad_norm": 0.7704019546508789,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.078,
"step": 331
},
{
"epoch": 0.028559139784946237,
"grad_norm": 18.713659286499023,
"learning_rate": 1.692143173870407e-05,
"loss": 0.4574,
"step": 332
},
{
"epoch": 0.02864516129032258,
"grad_norm": 0.08445344865322113,
"learning_rate": 1.68961954373567e-05,
"loss": 0.0016,
"step": 333
},
{
"epoch": 0.028731182795698924,
"grad_norm": 0.4566827118396759,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.0373,
"step": 334
},
{
"epoch": 0.028817204301075268,
"grad_norm": 4.861818313598633,
"learning_rate": 1.684547105928689e-05,
"loss": 0.1263,
"step": 335
},
{
"epoch": 0.02890322580645161,
"grad_norm": 0.8080859184265137,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.0752,
"step": 336
},
{
"epoch": 0.02898924731182796,
"grad_norm": 0.16214898228645325,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0014,
"step": 337
},
{
"epoch": 0.029075268817204302,
"grad_norm": 8.505428314208984,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.0889,
"step": 338
},
{
"epoch": 0.029161290322580646,
"grad_norm": 1.1777197122573853,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.0738,
"step": 339
},
{
"epoch": 0.02924731182795699,
"grad_norm": 0.16665324568748474,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.003,
"step": 340
},
{
"epoch": 0.029333333333333333,
"grad_norm": 2.379326105117798,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.0987,
"step": 341
},
{
"epoch": 0.029419354838709676,
"grad_norm": 1.551870584487915,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.0737,
"step": 342
},
{
"epoch": 0.02950537634408602,
"grad_norm": 1.9003819227218628,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.063,
"step": 343
},
{
"epoch": 0.029591397849462367,
"grad_norm": 12.26793384552002,
"learning_rate": 1.661311865323652e-05,
"loss": 0.5357,
"step": 344
},
{
"epoch": 0.02967741935483871,
"grad_norm": 0.12813951075077057,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.0019,
"step": 345
},
{
"epoch": 0.029763440860215054,
"grad_norm": 6.388195514678955,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.0322,
"step": 346
},
{
"epoch": 0.029849462365591398,
"grad_norm": 8.91196346282959,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.6304,
"step": 347
},
{
"epoch": 0.02993548387096774,
"grad_norm": 2.6920552253723145,
"learning_rate": 1.650774217265851e-05,
"loss": 0.0826,
"step": 348
},
{
"epoch": 0.030021505376344085,
"grad_norm": 22.753618240356445,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.283,
"step": 349
},
{
"epoch": 0.030107526881720432,
"grad_norm": 0.8969531655311584,
"learning_rate": 1.645457687723951e-05,
"loss": 0.0555,
"step": 350
},
{
"epoch": 0.030193548387096775,
"grad_norm": 1.018851399421692,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0534,
"step": 351
},
{
"epoch": 0.03027956989247312,
"grad_norm": 15.589733123779297,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.1539,
"step": 352
},
{
"epoch": 0.030365591397849463,
"grad_norm": 0.7832201719284058,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0469,
"step": 353
},
{
"epoch": 0.030451612903225806,
"grad_norm": 0.6544471979141235,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.0338,
"step": 354
},
{
"epoch": 0.03053763440860215,
"grad_norm": 2.684368133544922,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0654,
"step": 355
},
{
"epoch": 0.030623655913978493,
"grad_norm": 1.5461190938949585,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.0668,
"step": 356
},
{
"epoch": 0.03070967741935484,
"grad_norm": 0.4829937815666199,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.0086,
"step": 357
},
{
"epoch": 0.030795698924731184,
"grad_norm": 0.8205646872520447,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.0125,
"step": 358
},
{
"epoch": 0.030881720430107527,
"grad_norm": 0.5053842663764954,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0505,
"step": 359
},
{
"epoch": 0.03096774193548387,
"grad_norm": 1.4946280717849731,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.0523,
"step": 360
},
{
"epoch": 0.031053763440860215,
"grad_norm": 1.72577965259552,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.019,
"step": 361
},
{
"epoch": 0.031139784946236558,
"grad_norm": 3.57784104347229,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.1146,
"step": 362
},
{
"epoch": 0.0312258064516129,
"grad_norm": 0.8708614706993103,
"learning_rate": 1.610145163901268e-05,
"loss": 0.0718,
"step": 363
},
{
"epoch": 0.031311827956989245,
"grad_norm": 16.201126098632812,
"learning_rate": 1.607375839723287e-05,
"loss": 0.219,
"step": 364
},
{
"epoch": 0.03139784946236559,
"grad_norm": 1.5675890445709229,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.0687,
"step": 365
},
{
"epoch": 0.03148387096774193,
"grad_norm": 0.6107972264289856,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.0692,
"step": 366
},
{
"epoch": 0.03156989247311828,
"grad_norm": 1.3218944072723389,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0407,
"step": 367
},
{
"epoch": 0.031655913978494626,
"grad_norm": 0.8572149276733398,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.0544,
"step": 368
},
{
"epoch": 0.03174193548387097,
"grad_norm": 2.889328718185425,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.1244,
"step": 369
},
{
"epoch": 0.031827956989247314,
"grad_norm": 2.046905517578125,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0222,
"step": 370
},
{
"epoch": 0.03191397849462366,
"grad_norm": 1.0577274560928345,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.046,
"step": 371
},
{
"epoch": 0.032,
"grad_norm": 7.275178909301758,
"learning_rate": 1.584957674987216e-05,
"loss": 0.0849,
"step": 372
},
{
"epoch": 0.032086021505376344,
"grad_norm": 19.529382705688477,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.2363,
"step": 373
},
{
"epoch": 0.03217204301075269,
"grad_norm": 8.86138916015625,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.2585,
"step": 374
},
{
"epoch": 0.03225806451612903,
"grad_norm": 3.18966007232666,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.1019,
"step": 375
},
{
"epoch": 0.032344086021505375,
"grad_norm": 0.3039686679840088,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0021,
"step": 376
},
{
"epoch": 0.03243010752688172,
"grad_norm": 0.38623660802841187,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0414,
"step": 377
},
{
"epoch": 0.03251612903225806,
"grad_norm": 1.0705783367156982,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.0557,
"step": 378
},
{
"epoch": 0.032602150537634406,
"grad_norm": 1.44542396068573,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0715,
"step": 379
},
{
"epoch": 0.032688172043010756,
"grad_norm": 2.654113531112671,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.039,
"step": 380
},
{
"epoch": 0.0327741935483871,
"grad_norm": 3.973752975463867,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.125,
"step": 381
},
{
"epoch": 0.03286021505376344,
"grad_norm": 2.304202079772949,
"learning_rate": 1.556295615500305e-05,
"loss": 0.073,
"step": 382
},
{
"epoch": 0.03294623655913979,
"grad_norm": 4.54976749420166,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0547,
"step": 383
},
{
"epoch": 0.03303225806451613,
"grad_norm": 16.492666244506836,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.3752,
"step": 384
},
{
"epoch": 0.033118279569892474,
"grad_norm": 10.679157257080078,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0795,
"step": 385
},
{
"epoch": 0.03320430107526882,
"grad_norm": 1.0078972578048706,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0553,
"step": 386
},
{
"epoch": 0.03329032258064516,
"grad_norm": 0.36279061436653137,
"learning_rate": 1.54170821028274e-05,
"loss": 0.0368,
"step": 387
},
{
"epoch": 0.033376344086021505,
"grad_norm": 0.7490427494049072,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0569,
"step": 388
},
{
"epoch": 0.03346236559139785,
"grad_norm": 7.000235080718994,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0799,
"step": 389
},
{
"epoch": 0.03354838709677419,
"grad_norm": 1.766234040260315,
"learning_rate": 1.53287627607073e-05,
"loss": 0.0117,
"step": 390
},
{
"epoch": 0.033634408602150535,
"grad_norm": 0.556225597858429,
"learning_rate": 1.529919264233205e-05,
"loss": 0.0287,
"step": 391
},
{
"epoch": 0.03372043010752688,
"grad_norm": 0.4689476191997528,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0698,
"step": 392
},
{
"epoch": 0.03380645161290322,
"grad_norm": 1.5458521842956543,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.0651,
"step": 393
},
{
"epoch": 0.03389247311827957,
"grad_norm": 2.3166942596435547,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.0326,
"step": 394
},
{
"epoch": 0.03397849462365592,
"grad_norm": 0.10857067257165909,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0012,
"step": 395
},
{
"epoch": 0.03406451612903226,
"grad_norm": 0.6150857210159302,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.0452,
"step": 396
},
{
"epoch": 0.034150537634408604,
"grad_norm": 5.561111927032471,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.2051,
"step": 397
},
{
"epoch": 0.03423655913978495,
"grad_norm": 0.11002308875322342,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.001,
"step": 398
},
{
"epoch": 0.03432258064516129,
"grad_norm": 0.324336975812912,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.0217,
"step": 399
},
{
"epoch": 0.034408602150537634,
"grad_norm": 0.15216132998466492,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.0019,
"step": 400
},
{
"epoch": 0.03449462365591398,
"grad_norm": 6.830773830413818,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.1076,
"step": 401
},
{
"epoch": 0.03458064516129032,
"grad_norm": 0.6550801396369934,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.0515,
"step": 402
},
{
"epoch": 0.034666666666666665,
"grad_norm": 2.0629918575286865,
"learning_rate": 1.493941866584231e-05,
"loss": 0.0196,
"step": 403
},
{
"epoch": 0.03475268817204301,
"grad_norm": 0.4882068932056427,
"learning_rate": 1.490903753615141e-05,
"loss": 0.0224,
"step": 404
},
{
"epoch": 0.03483870967741935,
"grad_norm": 0.5143284797668457,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.0281,
"step": 405
},
{
"epoch": 0.034924731182795696,
"grad_norm": 11.54061508178711,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1232,
"step": 406
},
{
"epoch": 0.035010752688172046,
"grad_norm": 7.367484092712402,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.2078,
"step": 407
},
{
"epoch": 0.03509677419354839,
"grad_norm": 0.18851612508296967,
"learning_rate": 1.478691857940607e-05,
"loss": 0.0182,
"step": 408
},
{
"epoch": 0.03518279569892473,
"grad_norm": 4.6408514976501465,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0903,
"step": 409
},
{
"epoch": 0.03526881720430108,
"grad_norm": 1.5257534980773926,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0789,
"step": 410
},
{
"epoch": 0.03535483870967742,
"grad_norm": 0.5931143164634705,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0365,
"step": 411
},
{
"epoch": 0.035440860215053764,
"grad_norm": 0.2917311489582062,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0034,
"step": 412
},
{
"epoch": 0.03552688172043011,
"grad_norm": 0.18733984231948853,
"learning_rate": 1.463296035119862e-05,
"loss": 0.0216,
"step": 413
},
{
"epoch": 0.03561290322580645,
"grad_norm": 0.9559271931648254,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.0417,
"step": 414
},
{
"epoch": 0.035698924731182795,
"grad_norm": 0.46000197529792786,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0277,
"step": 415
},
{
"epoch": 0.03578494623655914,
"grad_norm": 0.7068843245506287,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.0399,
"step": 416
},
{
"epoch": 0.03587096774193548,
"grad_norm": 1.2146211862564087,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.0679,
"step": 417
},
{
"epoch": 0.035956989247311825,
"grad_norm": 0.09786419570446014,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.001,
"step": 418
},
{
"epoch": 0.03604301075268817,
"grad_norm": 23.267200469970703,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.2187,
"step": 419
},
{
"epoch": 0.03612903225806452,
"grad_norm": 0.08939865231513977,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.0018,
"step": 420
},
{
"epoch": 0.03621505376344086,
"grad_norm": 0.21663102507591248,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.0201,
"step": 421
},
{
"epoch": 0.03630107526881721,
"grad_norm": 0.4929128587245941,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0566,
"step": 422
},
{
"epoch": 0.03638709677419355,
"grad_norm": 0.20295441150665283,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.0168,
"step": 423
},
{
"epoch": 0.036473118279569894,
"grad_norm": 11.831048965454102,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.1475,
"step": 424
},
{
"epoch": 0.03655913978494624,
"grad_norm": 0.9440916776657104,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0714,
"step": 425
},
{
"epoch": 0.03664516129032258,
"grad_norm": 0.5353071093559265,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.049,
"step": 426
},
{
"epoch": 0.036731182795698925,
"grad_norm": 0.3578296899795532,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.0299,
"step": 427
},
{
"epoch": 0.03681720430107527,
"grad_norm": 1.1037670373916626,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.0718,
"step": 428
},
{
"epoch": 0.03690322580645161,
"grad_norm": 0.7938814163208008,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0739,
"step": 429
},
{
"epoch": 0.036989247311827955,
"grad_norm": 1.74843168258667,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.076,
"step": 430
},
{
"epoch": 0.0370752688172043,
"grad_norm": 0.7622953653335571,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0715,
"step": 431
},
{
"epoch": 0.03716129032258064,
"grad_norm": 1.1367182731628418,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.0591,
"step": 432
},
{
"epoch": 0.037247311827956986,
"grad_norm": 0.8989241123199463,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0073,
"step": 433
},
{
"epoch": 0.037333333333333336,
"grad_norm": 0.15885065495967865,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0004,
"step": 434
},
{
"epoch": 0.03741935483870968,
"grad_norm": 0.8541908860206604,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0658,
"step": 435
},
{
"epoch": 0.037505376344086024,
"grad_norm": 0.7835508584976196,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.0673,
"step": 436
},
{
"epoch": 0.03759139784946237,
"grad_norm": 0.3358050584793091,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.0463,
"step": 437
},
{
"epoch": 0.03767741935483871,
"grad_norm": 0.9010152816772461,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.0099,
"step": 438
},
{
"epoch": 0.037763440860215054,
"grad_norm": 1.7610527276992798,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.0532,
"step": 439
},
{
"epoch": 0.0378494623655914,
"grad_norm": 2.229600429534912,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0828,
"step": 440
},
{
"epoch": 0.03793548387096774,
"grad_norm": 1.5150995254516602,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0154,
"step": 441
},
{
"epoch": 0.038021505376344085,
"grad_norm": 0.10854873806238174,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0018,
"step": 442
},
{
"epoch": 0.03810752688172043,
"grad_norm": 0.2967485785484314,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.0469,
"step": 443
},
{
"epoch": 0.03819354838709677,
"grad_norm": 45.047271728515625,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.5587,
"step": 444
},
{
"epoch": 0.038279569892473116,
"grad_norm": 17.297653198242188,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.0828,
"step": 445
},
{
"epoch": 0.03836559139784946,
"grad_norm": 0.7074285745620728,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0407,
"step": 446
},
{
"epoch": 0.03845161290322581,
"grad_norm": 0.05841834470629692,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0016,
"step": 447
},
{
"epoch": 0.03853763440860215,
"grad_norm": 0.8092709183692932,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.042,
"step": 448
},
{
"epoch": 0.0386236559139785,
"grad_norm": 1.1940699815750122,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.0999,
"step": 449
},
{
"epoch": 0.03870967741935484,
"grad_norm": 1.5089001655578613,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.1066,
"step": 450
},
{
"epoch": 0.038795698924731184,
"grad_norm": 0.06807069480419159,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0002,
"step": 451
},
{
"epoch": 0.03888172043010753,
"grad_norm": 0.6153876185417175,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0568,
"step": 452
},
{
"epoch": 0.03896774193548387,
"grad_norm": 0.08948861062526703,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.0013,
"step": 453
},
{
"epoch": 0.039053763440860215,
"grad_norm": 0.6557666063308716,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.0478,
"step": 454
},
{
"epoch": 0.03913978494623656,
"grad_norm": 18.694232940673828,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.1014,
"step": 455
},
{
"epoch": 0.0392258064516129,
"grad_norm": 22.71377182006836,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.1433,
"step": 456
},
{
"epoch": 0.039311827956989245,
"grad_norm": 0.13822661340236664,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0165,
"step": 457
},
{
"epoch": 0.03939784946236559,
"grad_norm": 18.860660552978516,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.2886,
"step": 458
},
{
"epoch": 0.03948387096774193,
"grad_norm": 0.3118293285369873,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0422,
"step": 459
},
{
"epoch": 0.03956989247311828,
"grad_norm": 0.7687340974807739,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0582,
"step": 460
},
{
"epoch": 0.03965591397849463,
"grad_norm": 0.10931949317455292,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0007,
"step": 461
},
{
"epoch": 0.03974193548387097,
"grad_norm": 0.5360721349716187,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.0468,
"step": 462
},
{
"epoch": 0.039827956989247314,
"grad_norm": 0.29420220851898193,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0434,
"step": 463
},
{
"epoch": 0.03991397849462366,
"grad_norm": 0.17784754931926727,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.0045,
"step": 464
},
{
"epoch": 0.04,
"grad_norm": 0.6907579898834229,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.0227,
"step": 465
},
{
"epoch": 0.040086021505376344,
"grad_norm": 0.1385393887758255,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.0019,
"step": 466
},
{
"epoch": 0.04017204301075269,
"grad_norm": 4.646353244781494,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.0735,
"step": 467
},
{
"epoch": 0.04025806451612903,
"grad_norm": 0.2532256245613098,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0178,
"step": 468
},
{
"epoch": 0.040344086021505375,
"grad_norm": 0.30014610290527344,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0024,
"step": 469
},
{
"epoch": 0.04043010752688172,
"grad_norm": 0.8456259369850159,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0513,
"step": 470
},
{
"epoch": 0.04051612903225806,
"grad_norm": 2.1403234004974365,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.0582,
"step": 471
},
{
"epoch": 0.040602150537634406,
"grad_norm": 19.30840301513672,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.2168,
"step": 472
},
{
"epoch": 0.04068817204301075,
"grad_norm": 0.949839174747467,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.028,
"step": 473
},
{
"epoch": 0.0407741935483871,
"grad_norm": 0.46832913160324097,
"learning_rate": 1.265556117486809e-05,
"loss": 0.0455,
"step": 474
},
{
"epoch": 0.04086021505376344,
"grad_norm": 0.8065933585166931,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.0577,
"step": 475
},
{
"epoch": 0.04094623655913979,
"grad_norm": 1.0928422212600708,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.0109,
"step": 476
},
{
"epoch": 0.04103225806451613,
"grad_norm": 0.5261349081993103,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.0386,
"step": 477
},
{
"epoch": 0.041118279569892474,
"grad_norm": 0.7469334006309509,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0284,
"step": 478
},
{
"epoch": 0.04120430107526882,
"grad_norm": 0.9594560861587524,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.0581,
"step": 479
},
{
"epoch": 0.04129032258064516,
"grad_norm": 1.0573279857635498,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.0114,
"step": 480
},
{
"epoch": 0.041376344086021505,
"grad_norm": 0.053066685795784,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.0006,
"step": 481
},
{
"epoch": 0.04146236559139785,
"grad_norm": 0.14365120232105255,
"learning_rate": 1.238533457578581e-05,
"loss": 0.0188,
"step": 482
},
{
"epoch": 0.04154838709677419,
"grad_norm": 1.643015742301941,
"learning_rate": 1.23514211310259e-05,
"loss": 0.0103,
"step": 483
},
{
"epoch": 0.041634408602150536,
"grad_norm": 0.7614791393280029,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.0227,
"step": 484
},
{
"epoch": 0.04172043010752688,
"grad_norm": 0.12262050062417984,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.0012,
"step": 485
},
{
"epoch": 0.04180645161290322,
"grad_norm": 0.10699264705181122,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.0004,
"step": 486
},
{
"epoch": 0.04189247311827957,
"grad_norm": 0.1653907299041748,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.001,
"step": 487
},
{
"epoch": 0.04197849462365592,
"grad_norm": 0.926601767539978,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.0409,
"step": 488
},
{
"epoch": 0.04206451612903226,
"grad_norm": 18.835126876831055,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.3754,
"step": 489
},
{
"epoch": 0.042150537634408604,
"grad_norm": 0.5505790710449219,
"learning_rate": 1.211324796455389e-05,
"loss": 0.0404,
"step": 490
},
{
"epoch": 0.04223655913978495,
"grad_norm": 0.11535416543483734,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0004,
"step": 491
},
{
"epoch": 0.04232258064516129,
"grad_norm": 0.5526607036590576,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0331,
"step": 492
},
{
"epoch": 0.042408602150537635,
"grad_norm": 6.443247318267822,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.073,
"step": 493
},
{
"epoch": 0.04249462365591398,
"grad_norm": 2.4193997383117676,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.078,
"step": 494
},
{
"epoch": 0.04258064516129032,
"grad_norm": 15.975000381469727,
"learning_rate": 1.194234351219972e-05,
"loss": 0.1913,
"step": 495
},
{
"epoch": 0.042666666666666665,
"grad_norm": 0.46803516149520874,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0441,
"step": 496
},
{
"epoch": 0.04275268817204301,
"grad_norm": 0.8427488207817078,
"learning_rate": 1.187381314585725e-05,
"loss": 0.0586,
"step": 497
},
{
"epoch": 0.04283870967741935,
"grad_norm": 0.8448715806007385,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0574,
"step": 498
},
{
"epoch": 0.042924731182795696,
"grad_norm": 0.30947965383529663,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.0264,
"step": 499
},
{
"epoch": 0.043010752688172046,
"grad_norm": 0.4901565611362457,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0454,
"step": 500
},
{
"epoch": 0.04309677419354839,
"grad_norm": 0.21561990678310394,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0149,
"step": 501
},
{
"epoch": 0.043182795698924734,
"grad_norm": 0.05099567025899887,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.0003,
"step": 502
},
{
"epoch": 0.04326881720430108,
"grad_norm": 0.494913786649704,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.0458,
"step": 503
},
{
"epoch": 0.04335483870967742,
"grad_norm": 0.30785197019577026,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.0368,
"step": 504
},
{
"epoch": 0.043440860215053764,
"grad_norm": 0.525658905506134,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0524,
"step": 505
},
{
"epoch": 0.04352688172043011,
"grad_norm": 1.2069048881530762,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0745,
"step": 506
},
{
"epoch": 0.04361290322580645,
"grad_norm": 0.46624451875686646,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0557,
"step": 507
},
{
"epoch": 0.043698924731182795,
"grad_norm": 1.360707402229309,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.059,
"step": 508
},
{
"epoch": 0.04378494623655914,
"grad_norm": 0.4702546298503876,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.0593,
"step": 509
},
{
"epoch": 0.04387096774193548,
"grad_norm": 1.5979403257369995,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.0872,
"step": 510
},
{
"epoch": 0.043956989247311826,
"grad_norm": 0.23515485227108002,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.003,
"step": 511
},
{
"epoch": 0.04404301075268817,
"grad_norm": 0.05853600427508354,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.0005,
"step": 512
},
{
"epoch": 0.04412903225806451,
"grad_norm": 0.39573636651039124,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0447,
"step": 513
},
{
"epoch": 0.04421505376344086,
"grad_norm": 8.23908519744873,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0737,
"step": 514
},
{
"epoch": 0.04430107526881721,
"grad_norm": 0.43761834502220154,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.0061,
"step": 515
},
{
"epoch": 0.04438709677419355,
"grad_norm": 23.628995895385742,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.2414,
"step": 516
},
{
"epoch": 0.044473118279569894,
"grad_norm": 1.0565506219863892,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0342,
"step": 517
},
{
"epoch": 0.04455913978494624,
"grad_norm": 0.14458726346492767,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0161,
"step": 518
},
{
"epoch": 0.04464516129032258,
"grad_norm": 0.10527769476175308,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0258,
"step": 519
},
{
"epoch": 0.044731182795698925,
"grad_norm": 1.5733777284622192,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.0692,
"step": 520
},
{
"epoch": 0.04481720430107527,
"grad_norm": 0.18575237691402435,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0002,
"step": 521
},
{
"epoch": 0.04490322580645161,
"grad_norm": 4.343693256378174,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.1294,
"step": 522
},
{
"epoch": 0.044989247311827955,
"grad_norm": 0.40545737743377686,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.0331,
"step": 523
},
{
"epoch": 0.0450752688172043,
"grad_norm": 2.2450573444366455,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0937,
"step": 524
},
{
"epoch": 0.04516129032258064,
"grad_norm": 33.10220718383789,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.2721,
"step": 525
},
{
"epoch": 0.045247311827956986,
"grad_norm": 2.528806209564209,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0497,
"step": 526
},
{
"epoch": 0.04533333333333334,
"grad_norm": 0.40581417083740234,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0451,
"step": 527
},
{
"epoch": 0.04541935483870968,
"grad_norm": 1.3152446746826172,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0598,
"step": 528
},
{
"epoch": 0.045505376344086024,
"grad_norm": 0.2866791784763336,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.0348,
"step": 529
},
{
"epoch": 0.04559139784946237,
"grad_norm": 1.03348708152771,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0103,
"step": 530
},
{
"epoch": 0.04567741935483871,
"grad_norm": 2.065891981124878,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.0553,
"step": 531
},
{
"epoch": 0.045763440860215054,
"grad_norm": 0.47983378171920776,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.0511,
"step": 532
},
{
"epoch": 0.0458494623655914,
"grad_norm": 0.057472631335258484,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.0003,
"step": 533
},
{
"epoch": 0.04593548387096774,
"grad_norm": 5.160360813140869,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.065,
"step": 534
},
{
"epoch": 0.046021505376344085,
"grad_norm": 10.316766738891602,
"learning_rate": 1.055821504993164e-05,
"loss": 0.6754,
"step": 535
},
{
"epoch": 0.04610752688172043,
"grad_norm": 0.13856926560401917,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.0022,
"step": 536
},
{
"epoch": 0.04619354838709677,
"grad_norm": 1.2394936084747314,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0665,
"step": 537
},
{
"epoch": 0.046279569892473116,
"grad_norm": 0.18689517676830292,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.0148,
"step": 538
},
{
"epoch": 0.04636559139784946,
"grad_norm": 3.4537198543548584,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.0907,
"step": 539
},
{
"epoch": 0.04645161290322581,
"grad_norm": 0.5184764862060547,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0054,
"step": 540
},
{
"epoch": 0.04653763440860215,
"grad_norm": 1.899298071861267,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.0283,
"step": 541
},
{
"epoch": 0.0466236559139785,
"grad_norm": 1.5816041231155396,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.057,
"step": 542
},
{
"epoch": 0.04670967741935484,
"grad_norm": 0.4355829060077667,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0213,
"step": 543
},
{
"epoch": 0.046795698924731184,
"grad_norm": 14.712191581726074,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.4919,
"step": 544
},
{
"epoch": 0.04688172043010753,
"grad_norm": 0.45089730620384216,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0457,
"step": 545
},
{
"epoch": 0.04696774193548387,
"grad_norm": 0.33708882331848145,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.0327,
"step": 546
},
{
"epoch": 0.047053763440860215,
"grad_norm": 8.833731651306152,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.1788,
"step": 547
},
{
"epoch": 0.04713978494623656,
"grad_norm": 0.16485916078090668,
"learning_rate": 1.010471784116246e-05,
"loss": 0.0019,
"step": 548
},
{
"epoch": 0.0472258064516129,
"grad_norm": 27.31016731262207,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.1536,
"step": 549
},
{
"epoch": 0.047311827956989246,
"grad_norm": 0.17711074650287628,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0184,
"step": 550
},
{
"epoch": 0.04739784946236559,
"grad_norm": 0.7526916861534119,
"learning_rate": 1e-05,
"loss": 0.0491,
"step": 551
},
{
"epoch": 0.04748387096774193,
"grad_norm": 0.4844660460948944,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0332,
"step": 552
},
{
"epoch": 0.047569892473118276,
"grad_norm": 44.50306701660156,
"learning_rate": 9.930187397020385e-06,
"loss": 0.2554,
"step": 553
},
{
"epoch": 0.04765591397849463,
"grad_norm": 0.11765478551387787,
"learning_rate": 9.895282158837545e-06,
"loss": 0.0189,
"step": 554
},
{
"epoch": 0.04774193548387097,
"grad_norm": 0.35154959559440613,
"learning_rate": 9.860378196608549e-06,
"loss": 0.0485,
"step": 555
},
{
"epoch": 0.047827956989247314,
"grad_norm": 0.26784801483154297,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0033,
"step": 556
},
{
"epoch": 0.04791397849462366,
"grad_norm": 17.311376571655273,
"learning_rate": 9.790575801166432e-06,
"loss": 0.4305,
"step": 557
},
{
"epoch": 0.048,
"grad_norm": 0.6146430373191833,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0042,
"step": 558
},
{
"epoch": 0.048086021505376345,
"grad_norm": 0.25335821509361267,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0331,
"step": 559
},
{
"epoch": 0.04817204301075269,
"grad_norm": 0.28571268916130066,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0163,
"step": 560
},
{
"epoch": 0.04825806451612903,
"grad_norm": 0.1584693193435669,
"learning_rate": 9.651005032974994e-06,
"loss": 0.0183,
"step": 561
},
{
"epoch": 0.048344086021505375,
"grad_norm": 0.8895270824432373,
"learning_rate": 9.616121909124801e-06,
"loss": 0.0483,
"step": 562
},
{
"epoch": 0.04843010752688172,
"grad_norm": 0.579006552696228,
"learning_rate": 9.581243462708007e-06,
"loss": 0.0476,
"step": 563
},
{
"epoch": 0.04851612903225806,
"grad_norm": 20.041345596313477,
"learning_rate": 9.546370118707463e-06,
"loss": 0.0836,
"step": 564
},
{
"epoch": 0.048602150537634406,
"grad_norm": 17.9771728515625,
"learning_rate": 9.511502302043867e-06,
"loss": 0.5601,
"step": 565
},
{
"epoch": 0.04868817204301075,
"grad_norm": 0.6775204539299011,
"learning_rate": 9.476640437570562e-06,
"loss": 0.035,
"step": 566
},
{
"epoch": 0.0487741935483871,
"grad_norm": 1.7222801446914673,
"learning_rate": 9.441784950068362e-06,
"loss": 0.0639,
"step": 567
},
{
"epoch": 0.048860215053763444,
"grad_norm": 2.4061119556427,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0166,
"step": 568
},
{
"epoch": 0.04894623655913979,
"grad_norm": 1.297899842262268,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0629,
"step": 569
},
{
"epoch": 0.04903225806451613,
"grad_norm": 0.08678483963012695,
"learning_rate": 9.337260996000002e-06,
"loss": 0.0,
"step": 570
},
{
"epoch": 0.049118279569892474,
"grad_norm": 0.2054297924041748,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0174,
"step": 571
},
{
"epoch": 0.04920430107526882,
"grad_norm": 3.0709662437438965,
"learning_rate": 9.267618028723687e-06,
"loss": 0.0743,
"step": 572
},
{
"epoch": 0.04929032258064516,
"grad_norm": 0.7435240745544434,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0499,
"step": 573
},
{
"epoch": 0.049376344086021505,
"grad_norm": 0.2106350213289261,
"learning_rate": 9.198010756711413e-06,
"loss": 0.003,
"step": 574
},
{
"epoch": 0.04946236559139785,
"grad_norm": 1.1141198873519897,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0617,
"step": 575
},
{
"epoch": 0.04954838709677419,
"grad_norm": 0.472577303647995,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0406,
"step": 576
},
{
"epoch": 0.049634408602150536,
"grad_norm": 1.7352815866470337,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0637,
"step": 577
},
{
"epoch": 0.04972043010752688,
"grad_norm": 0.0677686333656311,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0011,
"step": 578
},
{
"epoch": 0.04980645161290322,
"grad_norm": 1.7300328016281128,
"learning_rate": 9.024171002408507e-06,
"loss": 0.0198,
"step": 579
},
{
"epoch": 0.04989247311827957,
"grad_norm": 0.04986424744129181,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0004,
"step": 580
},
{
"epoch": 0.04997849462365592,
"grad_norm": 24.545026779174805,
"learning_rate": 8.954715367323468e-06,
"loss": 0.1554,
"step": 581
},
{
"epoch": 0.05006451612903226,
"grad_norm": 0.34065911173820496,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0191,
"step": 582
},
{
"epoch": 0.050150537634408604,
"grad_norm": 1.6574269533157349,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0314,
"step": 583
},
{
"epoch": 0.05023655913978495,
"grad_norm": 0.7514144778251648,
"learning_rate": 8.850628495071336e-06,
"loss": 0.047,
"step": 584
},
{
"epoch": 0.05032258064516129,
"grad_norm": 0.9866738319396973,
"learning_rate": 8.815960316934991e-06,
"loss": 0.0215,
"step": 585
},
{
"epoch": 0.050408602150537635,
"grad_norm": 0.30546078085899353,
"learning_rate": 8.781306565948528e-06,
"loss": 0.0164,
"step": 586
},
{
"epoch": 0.05049462365591398,
"grad_norm": 0.4999809265136719,
"learning_rate": 8.746667664356957e-06,
"loss": 0.049,
"step": 587
},
{
"epoch": 0.05058064516129032,
"grad_norm": 0.8216047286987305,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0087,
"step": 588
},
{
"epoch": 0.050666666666666665,
"grad_norm": 1.9540270566940308,
"learning_rate": 8.677436097428775e-06,
"loss": 0.0802,
"step": 589
},
{
"epoch": 0.05075268817204301,
"grad_norm": 0.18779006600379944,
"learning_rate": 8.642844275656957e-06,
"loss": 0.0187,
"step": 590
},
{
"epoch": 0.05083870967741935,
"grad_norm": 0.6665785908699036,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0227,
"step": 591
},
{
"epoch": 0.050924731182795696,
"grad_norm": 0.2524818480014801,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0224,
"step": 592
},
{
"epoch": 0.05101075268817204,
"grad_norm": 1.1699833869934082,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0585,
"step": 593
},
{
"epoch": 0.05109677419354839,
"grad_norm": 1.0949643850326538,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0482,
"step": 594
},
{
"epoch": 0.051182795698924734,
"grad_norm": 0.34720727801322937,
"learning_rate": 8.47014163715962e-06,
"loss": 0.0426,
"step": 595
},
{
"epoch": 0.05126881720430108,
"grad_norm": 0.7867182493209839,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0434,
"step": 596
},
{
"epoch": 0.05135483870967742,
"grad_norm": 1.5878329277038574,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0534,
"step": 597
},
{
"epoch": 0.051440860215053764,
"grad_norm": 0.39357221126556396,
"learning_rate": 8.366740377583781e-06,
"loss": 0.0459,
"step": 598
},
{
"epoch": 0.05152688172043011,
"grad_norm": 21.886886596679688,
"learning_rate": 8.332312532838978e-06,
"loss": 0.4427,
"step": 599
},
{
"epoch": 0.05161290322580645,
"grad_norm": 0.07870791107416153,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0011,
"step": 600
},
{
"epoch": 0.051698924731182795,
"grad_norm": 0.30214324593544006,
"learning_rate": 8.263518223330698e-06,
"loss": 0.0436,
"step": 601
},
{
"epoch": 0.05178494623655914,
"grad_norm": 0.1962364763021469,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0024,
"step": 602
},
{
"epoch": 0.05187096774193548,
"grad_norm": 0.32249677181243896,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0447,
"step": 603
},
{
"epoch": 0.051956989247311826,
"grad_norm": 5.703609466552734,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0214,
"step": 604
},
{
"epoch": 0.05204301075268817,
"grad_norm": 0.8382378220558167,
"learning_rate": 8.126186854142752e-06,
"loss": 0.0599,
"step": 605
},
{
"epoch": 0.05212903225806451,
"grad_norm": 0.4358929395675659,
"learning_rate": 8.091910046234552e-06,
"loss": 0.0171,
"step": 606
},
{
"epoch": 0.05221505376344086,
"grad_norm": 0.12581054866313934,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0139,
"step": 607
},
{
"epoch": 0.05230107526881721,
"grad_norm": 0.465304434299469,
"learning_rate": 8.023426596208739e-06,
"loss": 0.0404,
"step": 608
},
{
"epoch": 0.05238709677419355,
"grad_norm": 0.09958583861589432,
"learning_rate": 7.989220788540356e-06,
"loss": 0.001,
"step": 609
},
{
"epoch": 0.052473118279569894,
"grad_norm": 0.2937449812889099,
"learning_rate": 7.955039481582098e-06,
"loss": 0.0166,
"step": 610
},
{
"epoch": 0.05255913978494624,
"grad_norm": 0.07765782624483109,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0017,
"step": 611
},
{
"epoch": 0.05264516129032258,
"grad_norm": 1.644150972366333,
"learning_rate": 7.886752035446116e-06,
"loss": 0.0595,
"step": 612
},
{
"epoch": 0.052731182795698925,
"grad_norm": 0.7550148963928223,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0491,
"step": 613
},
{
"epoch": 0.05281720430107527,
"grad_norm": 0.9756588339805603,
"learning_rate": 7.818567586034578e-06,
"loss": 0.0238,
"step": 614
},
{
"epoch": 0.05290322580645161,
"grad_norm": 0.6819542646408081,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0414,
"step": 615
},
{
"epoch": 0.052989247311827956,
"grad_norm": 0.3407612442970276,
"learning_rate": 7.750489456561351e-06,
"loss": 0.0313,
"step": 616
},
{
"epoch": 0.0530752688172043,
"grad_norm": 0.11561661958694458,
"learning_rate": 7.716491298893443e-06,
"loss": 0.0187,
"step": 617
},
{
"epoch": 0.05316129032258064,
"grad_norm": 0.1793968677520752,
"learning_rate": 7.68252096505843e-06,
"loss": 0.016,
"step": 618
},
{
"epoch": 0.053247311827956986,
"grad_norm": 0.4549241065979004,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0391,
"step": 619
},
{
"epoch": 0.05333333333333334,
"grad_norm": 0.5826348662376404,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0509,
"step": 620
},
{
"epoch": 0.05341935483870968,
"grad_norm": 0.15304617583751678,
"learning_rate": 7.580781044003324e-06,
"loss": 0.0157,
"step": 621
},
{
"epoch": 0.053505376344086024,
"grad_norm": 0.22097612917423248,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0171,
"step": 622
},
{
"epoch": 0.05359139784946237,
"grad_norm": 0.9715512990951538,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0347,
"step": 623
},
{
"epoch": 0.05367741935483871,
"grad_norm": 0.39079171419143677,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0398,
"step": 624
},
{
"epoch": 0.053763440860215055,
"grad_norm": 1.746471881866455,
"learning_rate": 7.445542420642097e-06,
"loss": 0.0302,
"step": 625
},
{
"epoch": 0.0538494623655914,
"grad_norm": 12.523533821105957,
"learning_rate": 7.411809548974792e-06,
"loss": 0.095,
"step": 626
},
{
"epoch": 0.05393548387096774,
"grad_norm": 0.3867488503456116,
"learning_rate": 7.378108213591355e-06,
"loss": 0.042,
"step": 627
},
{
"epoch": 0.054021505376344085,
"grad_norm": 0.1662452220916748,
"learning_rate": 7.344438825131912e-06,
"loss": 0.0177,
"step": 628
},
{
"epoch": 0.05410752688172043,
"grad_norm": 0.25189900398254395,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0144,
"step": 629
},
{
"epoch": 0.05419354838709677,
"grad_norm": 0.22375847399234772,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0161,
"step": 630
},
{
"epoch": 0.054279569892473116,
"grad_norm": 0.08189908415079117,
"learning_rate": 7.243626441830009e-06,
"loss": 0.0014,
"step": 631
},
{
"epoch": 0.05436559139784946,
"grad_norm": 0.12618860602378845,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0157,
"step": 632
},
{
"epoch": 0.0544516129032258,
"grad_norm": 0.7748531699180603,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0477,
"step": 633
},
{
"epoch": 0.054537634408602154,
"grad_norm": 0.15700079500675201,
"learning_rate": 7.143116325950266e-06,
"loss": 0.0028,
"step": 634
},
{
"epoch": 0.0546236559139785,
"grad_norm": 0.8115957379341125,
"learning_rate": 7.109682030555283e-06,
"loss": 0.0488,
"step": 635
},
{
"epoch": 0.05470967741935484,
"grad_norm": 1.2920669317245483,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0551,
"step": 636
},
{
"epoch": 0.054795698924731184,
"grad_norm": 0.517973005771637,
"learning_rate": 7.042919499559538e-06,
"loss": 0.0064,
"step": 637
},
{
"epoch": 0.05488172043010753,
"grad_norm": 0.6357753872871399,
"learning_rate": 7.009592077439135e-06,
"loss": 0.0491,
"step": 638
},
{
"epoch": 0.05496774193548387,
"grad_norm": 1.5114226341247559,
"learning_rate": 6.976301092495556e-06,
"loss": 0.0404,
"step": 639
},
{
"epoch": 0.055053763440860215,
"grad_norm": 0.26915115118026733,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0436,
"step": 640
},
{
"epoch": 0.05513978494623656,
"grad_norm": 0.11945857852697372,
"learning_rate": 6.909830056250527e-06,
"loss": 0.0165,
"step": 641
},
{
"epoch": 0.0552258064516129,
"grad_norm": 0.06642992049455643,
"learning_rate": 6.876650814877675e-06,
"loss": 0.0014,
"step": 642
},
{
"epoch": 0.055311827956989246,
"grad_norm": 0.04409955441951752,
"learning_rate": 6.843509630528977e-06,
"loss": 0.0008,
"step": 643
},
{
"epoch": 0.05539784946236559,
"grad_norm": 0.37977316975593567,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0407,
"step": 644
},
{
"epoch": 0.05548387096774193,
"grad_norm": 1.3953419923782349,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0271,
"step": 645
},
{
"epoch": 0.055569892473118276,
"grad_norm": 0.3367500901222229,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0029,
"step": 646
},
{
"epoch": 0.05565591397849463,
"grad_norm": 0.34404292702674866,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0271,
"step": 647
},
{
"epoch": 0.05574193548387097,
"grad_norm": 0.600295901298523,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0488,
"step": 648
},
{
"epoch": 0.055827956989247314,
"grad_norm": 0.4797069728374481,
"learning_rate": 6.645484302497452e-06,
"loss": 0.0444,
"step": 649
},
{
"epoch": 0.05591397849462366,
"grad_norm": 0.6412423849105835,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0521,
"step": 650
},
{
"epoch": 0.056,
"grad_norm": 0.3579300343990326,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0342,
"step": 651
},
{
"epoch": 0.056086021505376345,
"grad_norm": 1.2016092538833618,
"learning_rate": 6.547018010014654e-06,
"loss": 0.0455,
"step": 652
},
{
"epoch": 0.05617204301075269,
"grad_norm": 0.15595698356628418,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.0002,
"step": 653
},
{
"epoch": 0.05625806451612903,
"grad_norm": 16.75366973876953,
"learning_rate": 6.481583515952983e-06,
"loss": 0.2699,
"step": 654
},
{
"epoch": 0.056344086021505375,
"grad_norm": 1.124528408050537,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0617,
"step": 655
},
{
"epoch": 0.05643010752688172,
"grad_norm": 0.15359826385974884,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0009,
"step": 656
},
{
"epoch": 0.05651612903225806,
"grad_norm": 0.9680391550064087,
"learning_rate": 6.383754299179079e-06,
"loss": 0.0084,
"step": 657
},
{
"epoch": 0.056602150537634406,
"grad_norm": 2.3303120136260986,
"learning_rate": 6.351232156623803e-06,
"loss": 0.0347,
"step": 658
},
{
"epoch": 0.05668817204301075,
"grad_norm": 0.28156599402427673,
"learning_rate": 6.318754473153221e-06,
"loss": 0.0414,
"step": 659
},
{
"epoch": 0.0567741935483871,
"grad_norm": 0.05627167224884033,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0016,
"step": 660
},
{
"epoch": 0.056860215053763444,
"grad_norm": 0.44349533319473267,
"learning_rate": 6.25393406584088e-06,
"loss": 0.0448,
"step": 661
},
{
"epoch": 0.05694623655913979,
"grad_norm": 0.5382404923439026,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0535,
"step": 662
},
{
"epoch": 0.05703225806451613,
"grad_norm": 1.6657644510269165,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0537,
"step": 663
},
{
"epoch": 0.057118279569892474,
"grad_norm": 1.438772439956665,
"learning_rate": 6.157046773401964e-06,
"loss": 0.0607,
"step": 664
},
{
"epoch": 0.05720430107526882,
"grad_norm": 1.2115646600723267,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0461,
"step": 665
},
{
"epoch": 0.05729032258064516,
"grad_norm": 0.17756454646587372,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0163,
"step": 666
},
{
"epoch": 0.057376344086021505,
"grad_norm": 0.5177412629127502,
"learning_rate": 6.06058090409049e-06,
"loss": 0.0196,
"step": 667
},
{
"epoch": 0.05746236559139785,
"grad_norm": 0.5366353392601013,
"learning_rate": 6.028521093652195e-06,
"loss": 0.035,
"step": 668
},
{
"epoch": 0.05754838709677419,
"grad_norm": 1.3713760375976562,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0672,
"step": 669
},
{
"epoch": 0.057634408602150536,
"grad_norm": 0.7184767127037048,
"learning_rate": 5.9645470364761e-06,
"loss": 0.052,
"step": 670
},
{
"epoch": 0.05772043010752688,
"grad_norm": 0.612748920917511,
"learning_rate": 5.932633569242e-06,
"loss": 0.0408,
"step": 671
},
{
"epoch": 0.05780645161290322,
"grad_norm": 0.40397384762763977,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0328,
"step": 672
},
{
"epoch": 0.05789247311827957,
"grad_norm": 1.0806492567062378,
"learning_rate": 5.868955701754584e-06,
"loss": 0.0051,
"step": 673
},
{
"epoch": 0.05797849462365592,
"grad_norm": 7.067903995513916,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0626,
"step": 674
},
{
"epoch": 0.05806451612903226,
"grad_norm": 0.10974466800689697,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0128,
"step": 675
},
{
"epoch": 0.058150537634408604,
"grad_norm": 0.23326830565929413,
"learning_rate": 5.773817382593008e-06,
"loss": 0.0184,
"step": 676
},
{
"epoch": 0.05823655913978495,
"grad_norm": 0.2011391669511795,
"learning_rate": 5.742207084349274e-06,
"loss": 0.0043,
"step": 677
},
{
"epoch": 0.05832258064516129,
"grad_norm": 0.15894480049610138,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0287,
"step": 678
},
{
"epoch": 0.058408602150537635,
"grad_norm": 0.055516887456178665,
"learning_rate": 5.679142511980176e-06,
"loss": 0.001,
"step": 679
},
{
"epoch": 0.05849462365591398,
"grad_norm": 0.11871472001075745,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0021,
"step": 680
},
{
"epoch": 0.05858064516129032,
"grad_norm": 0.3745734691619873,
"learning_rate": 5.616288532109225e-06,
"loss": 0.0395,
"step": 681
},
{
"epoch": 0.058666666666666666,
"grad_norm": 0.20747460424900055,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0408,
"step": 682
},
{
"epoch": 0.05875268817204301,
"grad_norm": 0.09483140707015991,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0128,
"step": 683
},
{
"epoch": 0.05883870967741935,
"grad_norm": 0.9139610528945923,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0754,
"step": 684
},
{
"epoch": 0.058924731182795696,
"grad_norm": 25.397844314575195,
"learning_rate": 5.491224593105695e-06,
"loss": 0.1776,
"step": 685
},
{
"epoch": 0.05901075268817204,
"grad_norm": 0.7194263339042664,
"learning_rate": 5.460095002604533e-06,
"loss": 0.0331,
"step": 686
},
{
"epoch": 0.05909677419354839,
"grad_norm": 0.5842320322990417,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0086,
"step": 687
},
{
"epoch": 0.059182795698924734,
"grad_norm": 0.3831823170185089,
"learning_rate": 5.398002152161484e-06,
"loss": 0.0276,
"step": 688
},
{
"epoch": 0.05926881720430108,
"grad_norm": 3.5154924392700195,
"learning_rate": 5.367039648801386e-06,
"loss": 0.0186,
"step": 689
},
{
"epoch": 0.05935483870967742,
"grad_norm": 9.187335968017578,
"learning_rate": 5.336133596601089e-06,
"loss": 0.6087,
"step": 690
},
{
"epoch": 0.059440860215053765,
"grad_norm": 12.233358383178711,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0536,
"step": 691
},
{
"epoch": 0.05952688172043011,
"grad_norm": 0.238552987575531,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0095,
"step": 692
},
{
"epoch": 0.05961290322580645,
"grad_norm": 1.0100759267807007,
"learning_rate": 5.243757909297247e-06,
"loss": 0.0478,
"step": 693
},
{
"epoch": 0.059698924731182795,
"grad_norm": 0.4475638270378113,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0177,
"step": 694
},
{
"epoch": 0.05978494623655914,
"grad_norm": 1.522112250328064,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0114,
"step": 695
},
{
"epoch": 0.05987096774193548,
"grad_norm": 1.8416249752044678,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0916,
"step": 696
},
{
"epoch": 0.059956989247311826,
"grad_norm": 1.2784309387207031,
"learning_rate": 5.121403408612672e-06,
"loss": 0.0808,
"step": 697
},
{
"epoch": 0.06004301075268817,
"grad_norm": 0.36965733766555786,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0458,
"step": 698
},
{
"epoch": 0.06012903225806451,
"grad_norm": 0.4033626616001129,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0433,
"step": 699
},
{
"epoch": 0.060215053763440864,
"grad_norm": 0.4812996983528137,
"learning_rate": 5.030260389724447e-06,
"loss": 0.0342,
"step": 700
},
{
"epoch": 0.06030107526881721,
"grad_norm": 0.5300344228744507,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0041,
"step": 701
},
{
"epoch": 0.06038709677419355,
"grad_norm": 1.568294882774353,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0132,
"step": 702
},
{
"epoch": 0.060473118279569894,
"grad_norm": 0.3441486358642578,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0044,
"step": 703
},
{
"epoch": 0.06055913978494624,
"grad_norm": 1.082911491394043,
"learning_rate": 4.909585842496287e-06,
"loss": 0.0553,
"step": 704
},
{
"epoch": 0.06064516129032258,
"grad_norm": 7.256593227386475,
"learning_rate": 4.879571351294287e-06,
"loss": 0.1265,
"step": 705
},
{
"epoch": 0.060731182795698925,
"grad_norm": 0.25618690252304077,
"learning_rate": 4.849619250899458e-06,
"loss": 0.0414,
"step": 706
},
{
"epoch": 0.06081720430107527,
"grad_norm": 1.0205395221710205,
"learning_rate": 4.8197299062687e-06,
"loss": 0.03,
"step": 707
},
{
"epoch": 0.06090322580645161,
"grad_norm": 0.21015413105487823,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0288,
"step": 708
},
{
"epoch": 0.060989247311827956,
"grad_norm": 0.4064849317073822,
"learning_rate": 4.76014094029921e-06,
"loss": 0.0063,
"step": 709
},
{
"epoch": 0.0610752688172043,
"grad_norm": 1.340759038925171,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.0722,
"step": 710
},
{
"epoch": 0.06116129032258064,
"grad_norm": 0.2745679020881653,
"learning_rate": 4.700807357667953e-06,
"loss": 0.0031,
"step": 711
},
{
"epoch": 0.061247311827956986,
"grad_norm": 0.323463499546051,
"learning_rate": 4.671237239292699e-06,
"loss": 0.0261,
"step": 712
},
{
"epoch": 0.06133333333333333,
"grad_norm": 0.33008670806884766,
"learning_rate": 4.641732050210032e-06,
"loss": 0.0429,
"step": 713
},
{
"epoch": 0.06141935483870968,
"grad_norm": 0.13211064040660858,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0017,
"step": 714
},
{
"epoch": 0.061505376344086024,
"grad_norm": 0.9314747452735901,
"learning_rate": 4.582917897172603e-06,
"loss": 0.0547,
"step": 715
},
{
"epoch": 0.06159139784946237,
"grad_norm": 0.847343921661377,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0523,
"step": 716
},
{
"epoch": 0.06167741935483871,
"grad_norm": 0.6814939975738525,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0285,
"step": 717
},
{
"epoch": 0.061763440860215055,
"grad_norm": 1.3675930500030518,
"learning_rate": 4.495192599150045e-06,
"loss": 0.0577,
"step": 718
},
{
"epoch": 0.0618494623655914,
"grad_norm": 0.13373658061027527,
"learning_rate": 4.46608450756656e-06,
"loss": 0.0392,
"step": 719
},
{
"epoch": 0.06193548387096774,
"grad_norm": 11.578421592712402,
"learning_rate": 4.437043844996952e-06,
"loss": 0.1956,
"step": 720
},
{
"epoch": 0.062021505376344085,
"grad_norm": 0.5449861884117126,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0334,
"step": 721
},
{
"epoch": 0.06210752688172043,
"grad_norm": 0.681031346321106,
"learning_rate": 4.379166221478697e-06,
"loss": 0.0344,
"step": 722
},
{
"epoch": 0.06219354838709677,
"grad_norm": 0.6701276302337646,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0357,
"step": 723
},
{
"epoch": 0.062279569892473116,
"grad_norm": 0.24201197922229767,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0409,
"step": 724
},
{
"epoch": 0.06236559139784946,
"grad_norm": 0.19946308434009552,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0135,
"step": 725
},
{
"epoch": 0.0624516129032258,
"grad_norm": 0.7804461121559143,
"learning_rate": 4.264235636489542e-06,
"loss": 0.0431,
"step": 726
},
{
"epoch": 0.06253763440860215,
"grad_norm": 0.10509052127599716,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0016,
"step": 727
},
{
"epoch": 0.06262365591397849,
"grad_norm": 29.797351837158203,
"learning_rate": 4.207188276573214e-06,
"loss": 0.2524,
"step": 728
},
{
"epoch": 0.06270967741935483,
"grad_norm": 0.12196232378482819,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0163,
"step": 729
},
{
"epoch": 0.06279569892473118,
"grad_norm": 0.2932770550251007,
"learning_rate": 4.150423250127846e-06,
"loss": 0.0172,
"step": 730
},
{
"epoch": 0.06288172043010752,
"grad_norm": 0.41895270347595215,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0053,
"step": 731
},
{
"epoch": 0.06296774193548386,
"grad_norm": 1.9735920429229736,
"learning_rate": 4.093943323800746e-06,
"loss": 0.0228,
"step": 732
},
{
"epoch": 0.06305376344086021,
"grad_norm": 0.10854923725128174,
"learning_rate": 4.065811133962987e-06,
"loss": 0.0002,
"step": 733
},
{
"epoch": 0.06313978494623657,
"grad_norm": 1.2445595264434814,
"learning_rate": 4.037751250343841e-06,
"loss": 0.07,
"step": 734
},
{
"epoch": 0.06322580645161291,
"grad_norm": 0.8963171243667603,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0735,
"step": 735
},
{
"epoch": 0.06331182795698925,
"grad_norm": 0.06893549859523773,
"learning_rate": 3.981849768479516e-06,
"loss": 0.0011,
"step": 736
},
{
"epoch": 0.0633978494623656,
"grad_norm": 0.7781766057014465,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0396,
"step": 737
},
{
"epoch": 0.06348387096774194,
"grad_norm": 0.08416116237640381,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.0173,
"step": 738
},
{
"epoch": 0.06356989247311828,
"grad_norm": 0.08717093616724014,
"learning_rate": 3.898548360987325e-06,
"loss": 0.0167,
"step": 739
},
{
"epoch": 0.06365591397849463,
"grad_norm": 0.2122025340795517,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.0184,
"step": 740
},
{
"epoch": 0.06374193548387097,
"grad_norm": 11.800824165344238,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.3558,
"step": 741
},
{
"epoch": 0.06382795698924731,
"grad_norm": 0.04387858137488365,
"learning_rate": 3.81591604642446e-06,
"loss": 0.0008,
"step": 742
},
{
"epoch": 0.06391397849462366,
"grad_norm": 3.4724667072296143,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.064,
"step": 743
},
{
"epoch": 0.064,
"grad_norm": 6.270655632019043,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.1091,
"step": 744
},
{
"epoch": 0.06408602150537634,
"grad_norm": 0.6801609396934509,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0432,
"step": 745
},
{
"epoch": 0.06417204301075269,
"grad_norm": 0.7614120244979858,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0493,
"step": 746
},
{
"epoch": 0.06425806451612903,
"grad_norm": 0.4645804464817047,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0472,
"step": 747
},
{
"epoch": 0.06434408602150538,
"grad_norm": 0.9466047883033752,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.0745,
"step": 748
},
{
"epoch": 0.06443010752688172,
"grad_norm": 0.18377548456192017,
"learning_rate": 3.625760102513103e-06,
"loss": 0.0413,
"step": 749
},
{
"epoch": 0.06451612903225806,
"grad_norm": 0.2968127131462097,
"learning_rate": 3.598903005150444e-06,
"loss": 0.041,
"step": 750
},
{
"epoch": 0.0646021505376344,
"grad_norm": 0.20752929151058197,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.0144,
"step": 751
},
{
"epoch": 0.06468817204301075,
"grad_norm": 0.42166468501091003,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0318,
"step": 752
},
{
"epoch": 0.0647741935483871,
"grad_norm": 0.4380362033843994,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0177,
"step": 753
},
{
"epoch": 0.06486021505376344,
"grad_norm": 1.2207481861114502,
"learning_rate": 3.492257827341492e-06,
"loss": 0.0385,
"step": 754
},
{
"epoch": 0.06494623655913978,
"grad_norm": 3.291860818862915,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.0504,
"step": 755
},
{
"epoch": 0.06503225806451612,
"grad_norm": 0.46458402276039124,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.0316,
"step": 756
},
{
"epoch": 0.06511827956989247,
"grad_norm": 0.648389995098114,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0427,
"step": 757
},
{
"epoch": 0.06520430107526881,
"grad_norm": 0.04757985100150108,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.0004,
"step": 758
},
{
"epoch": 0.06529032258064515,
"grad_norm": 0.10170404613018036,
"learning_rate": 3.360737873477584e-06,
"loss": 0.0016,
"step": 759
},
{
"epoch": 0.06537634408602151,
"grad_norm": 15.877498626708984,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.3548,
"step": 760
},
{
"epoch": 0.06546236559139786,
"grad_norm": 0.8153650760650635,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0364,
"step": 761
},
{
"epoch": 0.0655483870967742,
"grad_norm": 0.06549742817878723,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.001,
"step": 762
},
{
"epoch": 0.06563440860215054,
"grad_norm": 0.19394999742507935,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.0139,
"step": 763
},
{
"epoch": 0.06572043010752689,
"grad_norm": 1.0987496376037598,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0392,
"step": 764
},
{
"epoch": 0.06580645161290323,
"grad_norm": 0.45911332964897156,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.0356,
"step": 765
},
{
"epoch": 0.06589247311827957,
"grad_norm": 8.916311264038086,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.1403,
"step": 766
},
{
"epoch": 0.06597849462365592,
"grad_norm": 0.24517515301704407,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0026,
"step": 767
},
{
"epoch": 0.06606451612903226,
"grad_norm": 0.27730047702789307,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.0316,
"step": 768
},
{
"epoch": 0.0661505376344086,
"grad_norm": 0.07554157078266144,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0007,
"step": 769
},
{
"epoch": 0.06623655913978495,
"grad_norm": 0.5018495917320251,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0454,
"step": 770
},
{
"epoch": 0.06632258064516129,
"grad_norm": 1.3560161590576172,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.0474,
"step": 771
},
{
"epoch": 0.06640860215053764,
"grad_norm": 0.3747791051864624,
"learning_rate": 3.028348971454356e-06,
"loss": 0.05,
"step": 772
},
{
"epoch": 0.06649462365591398,
"grad_norm": 1.5765891075134277,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0809,
"step": 773
},
{
"epoch": 0.06658064516129032,
"grad_norm": 0.3581109642982483,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0418,
"step": 774
},
{
"epoch": 0.06666666666666667,
"grad_norm": 0.151851624250412,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0027,
"step": 775
},
{
"epoch": 0.06675268817204301,
"grad_norm": 1.5862480401992798,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0103,
"step": 776
},
{
"epoch": 0.06683870967741935,
"grad_norm": 0.4262717664241791,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0472,
"step": 777
},
{
"epoch": 0.0669247311827957,
"grad_norm": 2.163163661956787,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.0329,
"step": 778
},
{
"epoch": 0.06701075268817204,
"grad_norm": 1.3016917705535889,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0523,
"step": 779
},
{
"epoch": 0.06709677419354838,
"grad_norm": 1.8324052095413208,
"learning_rate": 2.830893923495173e-06,
"loss": 0.0289,
"step": 780
},
{
"epoch": 0.06718279569892473,
"grad_norm": 0.6432943940162659,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.0436,
"step": 781
},
{
"epoch": 0.06726881720430107,
"grad_norm": 0.1624809056520462,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.0011,
"step": 782
},
{
"epoch": 0.06735483870967741,
"grad_norm": 0.17541979253292084,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0025,
"step": 783
},
{
"epoch": 0.06744086021505376,
"grad_norm": 1.027767539024353,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.0568,
"step": 784
},
{
"epoch": 0.0675268817204301,
"grad_norm": 17.318443298339844,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.2648,
"step": 785
},
{
"epoch": 0.06761290322580644,
"grad_norm": 0.5130524039268494,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.0471,
"step": 786
},
{
"epoch": 0.0676989247311828,
"grad_norm": 0.026903638616204262,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.0008,
"step": 787
},
{
"epoch": 0.06778494623655915,
"grad_norm": 0.19596929848194122,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0142,
"step": 788
},
{
"epoch": 0.06787096774193549,
"grad_norm": 0.33592498302459717,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0321,
"step": 789
},
{
"epoch": 0.06795698924731183,
"grad_norm": 0.19687753915786743,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0395,
"step": 790
},
{
"epoch": 0.06804301075268818,
"grad_norm": 0.6360945105552673,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0431,
"step": 791
},
{
"epoch": 0.06812903225806452,
"grad_norm": 0.15658673644065857,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0151,
"step": 792
},
{
"epoch": 0.06821505376344086,
"grad_norm": 0.6448573470115662,
"learning_rate": 2.522019095014683e-06,
"loss": 0.0552,
"step": 793
},
{
"epoch": 0.06830107526881721,
"grad_norm": 0.2615584433078766,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0462,
"step": 794
},
{
"epoch": 0.06838709677419355,
"grad_norm": 0.24448485672473907,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0416,
"step": 795
},
{
"epoch": 0.0684731182795699,
"grad_norm": 1.0664888620376587,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0408,
"step": 796
},
{
"epoch": 0.06855913978494624,
"grad_norm": 6.5837554931640625,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0592,
"step": 797
},
{
"epoch": 0.06864516129032258,
"grad_norm": 6.526048183441162,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0748,
"step": 798
},
{
"epoch": 0.06873118279569893,
"grad_norm": 0.2031853049993515,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0009,
"step": 799
},
{
"epoch": 0.06881720430107527,
"grad_norm": 0.5432092547416687,
"learning_rate": 2.362039713653581e-06,
"loss": 0.042,
"step": 800
},
{
"epoch": 0.06890322580645161,
"grad_norm": 0.1580670028924942,
"learning_rate": 2.339555568810221e-06,
"loss": 0.0405,
"step": 801
},
{
"epoch": 0.06898924731182796,
"grad_norm": 0.24623772501945496,
"learning_rate": 2.317164764064769e-06,
"loss": 0.0174,
"step": 802
},
{
"epoch": 0.0690752688172043,
"grad_norm": 1.9072779417037964,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.068,
"step": 803
},
{
"epoch": 0.06916129032258064,
"grad_norm": 0.42932820320129395,
"learning_rate": 2.27266426502649e-06,
"loss": 0.0166,
"step": 804
},
{
"epoch": 0.06924731182795699,
"grad_norm": 0.2034042775630951,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.009,
"step": 805
},
{
"epoch": 0.06933333333333333,
"grad_norm": 0.14331719279289246,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0384,
"step": 806
},
{
"epoch": 0.06941935483870967,
"grad_norm": 0.06686019897460938,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0006,
"step": 807
},
{
"epoch": 0.06950537634408602,
"grad_norm": 0.23538827896118164,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0177,
"step": 808
},
{
"epoch": 0.06959139784946236,
"grad_norm": 0.3997952342033386,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0151,
"step": 809
},
{
"epoch": 0.0696774193548387,
"grad_norm": 0.5086291432380676,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0522,
"step": 810
},
{
"epoch": 0.06976344086021505,
"grad_norm": 0.12174627929925919,
"learning_rate": 2.119892463932781e-06,
"loss": 0.017,
"step": 811
},
{
"epoch": 0.06984946236559139,
"grad_norm": 0.9089086055755615,
"learning_rate": 2.098449876243096e-06,
"loss": 0.04,
"step": 812
},
{
"epoch": 0.06993548387096774,
"grad_norm": 16.298641204833984,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.3347,
"step": 813
},
{
"epoch": 0.07002150537634409,
"grad_norm": 0.07275247573852539,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0007,
"step": 814
},
{
"epoch": 0.07010752688172044,
"grad_norm": 0.7558987140655518,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0523,
"step": 815
},
{
"epoch": 0.07019354838709678,
"grad_norm": 0.21539367735385895,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0147,
"step": 816
},
{
"epoch": 0.07027956989247312,
"grad_norm": 0.9567487835884094,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0501,
"step": 817
},
{
"epoch": 0.07036559139784947,
"grad_norm": 0.6094244718551636,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0374,
"step": 818
},
{
"epoch": 0.07045161290322581,
"grad_norm": 0.11600513011217117,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0297,
"step": 819
},
{
"epoch": 0.07053763440860215,
"grad_norm": 0.050217825919389725,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0006,
"step": 820
},
{
"epoch": 0.0706236559139785,
"grad_norm": 2.584566831588745,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0482,
"step": 821
},
{
"epoch": 0.07070967741935484,
"grad_norm": 0.9137936234474182,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0514,
"step": 822
},
{
"epoch": 0.07079569892473118,
"grad_norm": 0.47950270771980286,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.0191,
"step": 823
},
{
"epoch": 0.07088172043010753,
"grad_norm": 0.08608151227235794,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0008,
"step": 824
},
{
"epoch": 0.07096774193548387,
"grad_norm": 0.521852970123291,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0453,
"step": 825
},
{
"epoch": 0.07105376344086022,
"grad_norm": 0.10574334114789963,
"learning_rate": 1.808479557110081e-06,
"loss": 0.0137,
"step": 826
},
{
"epoch": 0.07113978494623656,
"grad_norm": 2.210048198699951,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0705,
"step": 827
},
{
"epoch": 0.0712258064516129,
"grad_norm": 0.7452006340026855,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.035,
"step": 828
},
{
"epoch": 0.07131182795698925,
"grad_norm": 20.14090919494629,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.2476,
"step": 829
},
{
"epoch": 0.07139784946236559,
"grad_norm": 6.750422477722168,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.0991,
"step": 830
},
{
"epoch": 0.07148387096774193,
"grad_norm": 1.4149521589279175,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0783,
"step": 831
},
{
"epoch": 0.07156989247311828,
"grad_norm": 0.6280036568641663,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0412,
"step": 832
},
{
"epoch": 0.07165591397849462,
"grad_norm": 0.0805603563785553,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.017,
"step": 833
},
{
"epoch": 0.07174193548387096,
"grad_norm": 0.32414036989212036,
"learning_rate": 1.651521367365936e-06,
"loss": 0.0371,
"step": 834
},
{
"epoch": 0.07182795698924731,
"grad_norm": 0.05950913205742836,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0006,
"step": 835
},
{
"epoch": 0.07191397849462365,
"grad_norm": 0.14364291727542877,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.023,
"step": 836
},
{
"epoch": 0.072,
"grad_norm": 2.638017177581787,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.0159,
"step": 837
},
{
"epoch": 0.07208602150537634,
"grad_norm": 0.18320181965827942,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.0019,
"step": 838
},
{
"epoch": 0.07217204301075268,
"grad_norm": 0.10719311982393265,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.025,
"step": 839
},
{
"epoch": 0.07225806451612904,
"grad_norm": 1.1186127662658691,
"learning_rate": 1.538068338724361e-06,
"loss": 0.0549,
"step": 840
},
{
"epoch": 0.07234408602150538,
"grad_norm": 0.4036846160888672,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0504,
"step": 841
},
{
"epoch": 0.07243010752688173,
"grad_norm": 0.03188015893101692,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.0004,
"step": 842
},
{
"epoch": 0.07251612903225807,
"grad_norm": 39.9023323059082,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.3616,
"step": 843
},
{
"epoch": 0.07260215053763441,
"grad_norm": 0.4421975016593933,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0383,
"step": 844
},
{
"epoch": 0.07268817204301076,
"grad_norm": 1.2133095264434814,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0518,
"step": 845
},
{
"epoch": 0.0727741935483871,
"grad_norm": 0.0632069855928421,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0008,
"step": 846
},
{
"epoch": 0.07286021505376344,
"grad_norm": 0.5547972321510315,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.0509,
"step": 847
},
{
"epoch": 0.07294623655913979,
"grad_norm": 0.6142999529838562,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.0486,
"step": 848
},
{
"epoch": 0.07303225806451613,
"grad_norm": 0.45687606930732727,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.0281,
"step": 849
},
{
"epoch": 0.07311827956989247,
"grad_norm": 1.080584168434143,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0223,
"step": 850
},
{
"epoch": 0.07320430107526882,
"grad_norm": 1.677720308303833,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0093,
"step": 851
},
{
"epoch": 0.07329032258064516,
"grad_norm": 0.20620085299015045,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0283,
"step": 852
},
{
"epoch": 0.0733763440860215,
"grad_norm": 12.645028114318848,
"learning_rate": 1.30505070494781e-06,
"loss": 0.0804,
"step": 853
},
{
"epoch": 0.07346236559139785,
"grad_norm": 0.4959879219532013,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0385,
"step": 854
},
{
"epoch": 0.07354838709677419,
"grad_norm": 0.5698630809783936,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0179,
"step": 855
},
{
"epoch": 0.07363440860215054,
"grad_norm": 0.1390477865934372,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0024,
"step": 856
},
{
"epoch": 0.07372043010752688,
"grad_norm": 2.140608549118042,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.067,
"step": 857
},
{
"epoch": 0.07380645161290322,
"grad_norm": 1.5918720960617065,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0797,
"step": 858
},
{
"epoch": 0.07389247311827957,
"grad_norm": 0.5098928213119507,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0576,
"step": 859
},
{
"epoch": 0.07397849462365591,
"grad_norm": 0.48281189799308777,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0043,
"step": 860
},
{
"epoch": 0.07406451612903225,
"grad_norm": 2.570976495742798,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.0637,
"step": 861
},
{
"epoch": 0.0741505376344086,
"grad_norm": 1.006514310836792,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.0482,
"step": 862
},
{
"epoch": 0.07423655913978494,
"grad_norm": 0.6433727741241455,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0621,
"step": 863
},
{
"epoch": 0.07432258064516128,
"grad_norm": 0.573498547077179,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.0475,
"step": 864
},
{
"epoch": 0.07440860215053763,
"grad_norm": 0.0934264063835144,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.0149,
"step": 865
},
{
"epoch": 0.07449462365591397,
"grad_norm": 0.25375381112098694,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.0166,
"step": 866
},
{
"epoch": 0.07458064516129033,
"grad_norm": 0.3528170585632324,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0021,
"step": 867
},
{
"epoch": 0.07466666666666667,
"grad_norm": 0.09370582550764084,
"learning_rate": 1.058457631606319e-06,
"loss": 0.001,
"step": 868
},
{
"epoch": 0.07475268817204302,
"grad_norm": 0.4991992712020874,
"learning_rate": 1.042882397605871e-06,
"loss": 0.0061,
"step": 869
},
{
"epoch": 0.07483870967741936,
"grad_norm": 0.08745426684617996,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0011,
"step": 870
},
{
"epoch": 0.0749247311827957,
"grad_norm": 0.8025544285774231,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0483,
"step": 871
},
{
"epoch": 0.07501075268817205,
"grad_norm": 0.34770914912223816,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0364,
"step": 872
},
{
"epoch": 0.07509677419354839,
"grad_norm": 0.2074299305677414,
"learning_rate": 9.816747359488632e-07,
"loss": 0.0316,
"step": 873
},
{
"epoch": 0.07518279569892473,
"grad_norm": 0.04261157289147377,
"learning_rate": 9.666470713669918e-07,
"loss": 0.001,
"step": 874
},
{
"epoch": 0.07526881720430108,
"grad_norm": 1.2047606706619263,
"learning_rate": 9.517294753398066e-07,
"loss": 0.0083,
"step": 875
},
{
"epoch": 0.07535483870967742,
"grad_norm": 1.0817291736602783,
"learning_rate": 9.369221296335007e-07,
"loss": 0.0599,
"step": 876
},
{
"epoch": 0.07544086021505376,
"grad_norm": 0.3332619369029999,
"learning_rate": 9.222252146709143e-07,
"loss": 0.0455,
"step": 877
},
{
"epoch": 0.07552688172043011,
"grad_norm": 0.6034457683563232,
"learning_rate": 9.076389095293148e-07,
"loss": 0.047,
"step": 878
},
{
"epoch": 0.07561290322580645,
"grad_norm": 0.2265661209821701,
"learning_rate": 8.931633919382299e-07,
"loss": 0.029,
"step": 879
},
{
"epoch": 0.0756989247311828,
"grad_norm": 0.07586542516946793,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0012,
"step": 880
},
{
"epoch": 0.07578494623655914,
"grad_norm": 0.8949000239372253,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0411,
"step": 881
},
{
"epoch": 0.07587096774193548,
"grad_norm": 0.36459389328956604,
"learning_rate": 8.504033215017527e-07,
"loss": 0.0319,
"step": 882
},
{
"epoch": 0.07595698924731183,
"grad_norm": 0.7259718179702759,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0501,
"step": 883
},
{
"epoch": 0.07604301075268817,
"grad_norm": 0.09692778438329697,
"learning_rate": 8.224537431601886e-07,
"loss": 0.0167,
"step": 884
},
{
"epoch": 0.07612903225806451,
"grad_norm": 16.163185119628906,
"learning_rate": 8.086466074476562e-07,
"loss": 0.3967,
"step": 885
},
{
"epoch": 0.07621505376344086,
"grad_norm": 0.780622124671936,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0425,
"step": 886
},
{
"epoch": 0.0763010752688172,
"grad_norm": 0.4603826105594635,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0455,
"step": 887
},
{
"epoch": 0.07638709677419354,
"grad_norm": 0.28862085938453674,
"learning_rate": 7.678978288701911e-07,
"loss": 0.0443,
"step": 888
},
{
"epoch": 0.07647311827956989,
"grad_norm": 0.4884816110134125,
"learning_rate": 7.545396638768698e-07,
"loss": 0.0421,
"step": 889
},
{
"epoch": 0.07655913978494623,
"grad_norm": 0.10120411962270737,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0232,
"step": 890
},
{
"epoch": 0.07664516129032257,
"grad_norm": 0.49541473388671875,
"learning_rate": 7.281614543321269e-07,
"loss": 0.0449,
"step": 891
},
{
"epoch": 0.07673118279569892,
"grad_norm": 0.15878531336784363,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0145,
"step": 892
},
{
"epoch": 0.07681720430107526,
"grad_norm": 2.301058053970337,
"learning_rate": 7.022351411174866e-07,
"loss": 0.0286,
"step": 893
},
{
"epoch": 0.07690322580645162,
"grad_norm": 0.20692890882492065,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0401,
"step": 894
},
{
"epoch": 0.07698924731182796,
"grad_norm": 0.5964217185974121,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0306,
"step": 895
},
{
"epoch": 0.0770752688172043,
"grad_norm": 1.6041860580444336,
"learning_rate": 6.641957350279838e-07,
"loss": 0.0437,
"step": 896
},
{
"epoch": 0.07716129032258065,
"grad_norm": 0.5073415040969849,
"learning_rate": 6.517432360398556e-07,
"loss": 0.0458,
"step": 897
},
{
"epoch": 0.077247311827957,
"grad_norm": 0.5337507128715515,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0405,
"step": 898
},
{
"epoch": 0.07733333333333334,
"grad_norm": 0.07075466215610504,
"learning_rate": 6.271801050810856e-07,
"loss": 0.0012,
"step": 899
},
{
"epoch": 0.07741935483870968,
"grad_norm": 0.053345296531915665,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0004,
"step": 900
},
{
"epoch": 0.07750537634408602,
"grad_norm": 0.06711148470640182,
"learning_rate": 6.030737921409169e-07,
"loss": 0.001,
"step": 901
},
{
"epoch": 0.07759139784946237,
"grad_norm": 0.13335081934928894,
"learning_rate": 5.911923104577455e-07,
"loss": 0.0118,
"step": 902
},
{
"epoch": 0.07767741935483871,
"grad_norm": 0.4616944491863251,
"learning_rate": 5.794254721270331e-07,
"loss": 0.0386,
"step": 903
},
{
"epoch": 0.07776344086021506,
"grad_norm": 12.91114330291748,
"learning_rate": 5.677734205239904e-07,
"loss": 0.123,
"step": 904
},
{
"epoch": 0.0778494623655914,
"grad_norm": 0.5097482204437256,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0406,
"step": 905
},
{
"epoch": 0.07793548387096774,
"grad_norm": 0.634564995765686,
"learning_rate": 5.448142440068316e-07,
"loss": 0.042,
"step": 906
},
{
"epoch": 0.07802150537634409,
"grad_norm": 0.40189385414123535,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0469,
"step": 907
},
{
"epoch": 0.07810752688172043,
"grad_norm": 0.2616819739341736,
"learning_rate": 5.223158999041444e-07,
"loss": 0.0414,
"step": 908
},
{
"epoch": 0.07819354838709677,
"grad_norm": 0.6781029105186462,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0266,
"step": 909
},
{
"epoch": 0.07827956989247312,
"grad_norm": 0.2495325356721878,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0459,
"step": 910
},
{
"epoch": 0.07836559139784946,
"grad_norm": 0.7014648914337158,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0579,
"step": 911
},
{
"epoch": 0.0784516129032258,
"grad_norm": 0.35808250308036804,
"learning_rate": 4.787060725786141e-07,
"loss": 0.0355,
"step": 912
},
{
"epoch": 0.07853763440860215,
"grad_norm": 0.5349449515342712,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0427,
"step": 913
},
{
"epoch": 0.07862365591397849,
"grad_norm": 0.23833853006362915,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0406,
"step": 914
},
{
"epoch": 0.07870967741935483,
"grad_norm": 0.06135661154985428,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0009,
"step": 915
},
{
"epoch": 0.07879569892473118,
"grad_norm": 0.2621379494667053,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0397,
"step": 916
},
{
"epoch": 0.07888172043010752,
"grad_norm": 1.4375600814819336,
"learning_rate": 4.268050246793276e-07,
"loss": 0.0159,
"step": 917
},
{
"epoch": 0.07896774193548387,
"grad_norm": 0.2933567762374878,
"learning_rate": 4.167742553486676e-07,
"loss": 0.037,
"step": 918
},
{
"epoch": 0.07905376344086021,
"grad_norm": 2.5214295387268066,
"learning_rate": 4.068602545994249e-07,
"loss": 0.0516,
"step": 919
},
{
"epoch": 0.07913978494623657,
"grad_norm": 0.5769420862197876,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0307,
"step": 920
},
{
"epoch": 0.07922580645161291,
"grad_norm": 0.41045108437538147,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0385,
"step": 921
},
{
"epoch": 0.07931182795698925,
"grad_norm": 6.576826572418213,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0662,
"step": 922
},
{
"epoch": 0.0793978494623656,
"grad_norm": 0.5694219470024109,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0527,
"step": 923
},
{
"epoch": 0.07948387096774194,
"grad_norm": 0.8089065551757812,
"learning_rate": 3.590459576589e-07,
"loss": 0.0469,
"step": 924
},
{
"epoch": 0.07956989247311828,
"grad_norm": 0.8619657754898071,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0579,
"step": 925
},
{
"epoch": 0.07965591397849463,
"grad_norm": 1.522670865058899,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0142,
"step": 926
},
{
"epoch": 0.07974193548387097,
"grad_norm": 16.685001373291016,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.3098,
"step": 927
},
{
"epoch": 0.07982795698924731,
"grad_norm": 4.488400936126709,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.061,
"step": 928
},
{
"epoch": 0.07991397849462366,
"grad_norm": 0.354764461517334,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.0488,
"step": 929
},
{
"epoch": 0.08,
"grad_norm": 0.15365296602249146,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0134,
"step": 930
},
{
"epoch": 0.08008602150537635,
"grad_norm": 0.39477673172950745,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0414,
"step": 931
},
{
"epoch": 0.08017204301075269,
"grad_norm": 16.021862030029297,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.5693,
"step": 932
},
{
"epoch": 0.08025806451612903,
"grad_norm": 0.5453251004219055,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.0293,
"step": 933
},
{
"epoch": 0.08034408602150538,
"grad_norm": 0.1045374646782875,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.0122,
"step": 934
},
{
"epoch": 0.08043010752688172,
"grad_norm": 0.6302529573440552,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0077,
"step": 935
},
{
"epoch": 0.08051612903225806,
"grad_norm": 0.32932358980178833,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.0268,
"step": 936
},
{
"epoch": 0.0806021505376344,
"grad_norm": 0.175184428691864,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.0326,
"step": 937
},
{
"epoch": 0.08068817204301075,
"grad_norm": 0.05547989159822464,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0004,
"step": 938
},
{
"epoch": 0.0807741935483871,
"grad_norm": 0.11873364448547363,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0013,
"step": 939
},
{
"epoch": 0.08086021505376344,
"grad_norm": 0.5810778141021729,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0377,
"step": 940
},
{
"epoch": 0.08094623655913978,
"grad_norm": 0.17171287536621094,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0227,
"step": 941
},
{
"epoch": 0.08103225806451612,
"grad_norm": 0.05749901756644249,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.0007,
"step": 942
},
{
"epoch": 0.08111827956989247,
"grad_norm": 0.819510281085968,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.0506,
"step": 943
},
{
"epoch": 0.08120430107526881,
"grad_norm": 0.11859465390443802,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0229,
"step": 944
},
{
"epoch": 0.08129032258064516,
"grad_norm": 0.26614150404930115,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.0094,
"step": 945
},
{
"epoch": 0.0813763440860215,
"grad_norm": 0.3212937116622925,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0316,
"step": 946
},
{
"epoch": 0.08146236559139786,
"grad_norm": 0.07445504516363144,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0169,
"step": 947
},
{
"epoch": 0.0815483870967742,
"grad_norm": 0.25952017307281494,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0337,
"step": 948
},
{
"epoch": 0.08163440860215054,
"grad_norm": 0.4497785270214081,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.031,
"step": 949
},
{
"epoch": 0.08172043010752689,
"grad_norm": 0.0719318762421608,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0004,
"step": 950
},
{
"epoch": 0.08180645161290323,
"grad_norm": 0.14760330319404602,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0276,
"step": 951
},
{
"epoch": 0.08189247311827957,
"grad_norm": 0.4321660101413727,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0327,
"step": 952
},
{
"epoch": 0.08197849462365592,
"grad_norm": 0.08429975062608719,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0006,
"step": 953
},
{
"epoch": 0.08206451612903226,
"grad_norm": 0.08962405472993851,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.0119,
"step": 954
},
{
"epoch": 0.0821505376344086,
"grad_norm": 11.667732238769531,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0854,
"step": 955
},
{
"epoch": 0.08223655913978495,
"grad_norm": 0.05361225828528404,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0005,
"step": 956
},
{
"epoch": 0.08232258064516129,
"grad_norm": 21.52396583557129,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.078,
"step": 957
},
{
"epoch": 0.08240860215053764,
"grad_norm": 22.472484588623047,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.1567,
"step": 958
},
{
"epoch": 0.08249462365591398,
"grad_norm": 0.09041581302881241,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0017,
"step": 959
},
{
"epoch": 0.08258064516129032,
"grad_norm": 0.07607632875442505,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.0281,
"step": 960
},
{
"epoch": 0.08266666666666667,
"grad_norm": 0.37712419033050537,
"learning_rate": 9.731931258429638e-08,
"loss": 0.043,
"step": 961
},
{
"epoch": 0.08275268817204301,
"grad_norm": 0.09832365065813065,
"learning_rate": 9.252159528556404e-08,
"loss": 0.0249,
"step": 962
},
{
"epoch": 0.08283870967741935,
"grad_norm": 0.6085331439971924,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0444,
"step": 963
},
{
"epoch": 0.0829247311827957,
"grad_norm": 0.3004654049873352,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0029,
"step": 964
},
{
"epoch": 0.08301075268817204,
"grad_norm": 0.11092333495616913,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0382,
"step": 965
},
{
"epoch": 0.08309677419354838,
"grad_norm": 0.12472664564847946,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0391,
"step": 966
},
{
"epoch": 0.08318279569892473,
"grad_norm": 0.04608062282204628,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0006,
"step": 967
},
{
"epoch": 0.08326881720430107,
"grad_norm": 0.22627900540828705,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0306,
"step": 968
},
{
"epoch": 0.08335483870967741,
"grad_norm": 0.11409452557563782,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0136,
"step": 969
},
{
"epoch": 0.08344086021505376,
"grad_norm": 0.2183784544467926,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0011,
"step": 970
},
{
"epoch": 0.0835268817204301,
"grad_norm": 0.3493112623691559,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0402,
"step": 971
},
{
"epoch": 0.08361290322580645,
"grad_norm": 0.5157341957092285,
"learning_rate": 5.119291171211793e-08,
"loss": 0.0594,
"step": 972
},
{
"epoch": 0.08369892473118279,
"grad_norm": 0.06796043366193771,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0013,
"step": 973
},
{
"epoch": 0.08378494623655915,
"grad_norm": 0.8845805525779724,
"learning_rate": 4.438035396920004e-08,
"loss": 0.0638,
"step": 974
},
{
"epoch": 0.08387096774193549,
"grad_norm": 1.1733800172805786,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0717,
"step": 975
},
{
"epoch": 0.08395698924731183,
"grad_norm": 0.6708886623382568,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0318,
"step": 976
},
{
"epoch": 0.08404301075268818,
"grad_norm": 0.2008119374513626,
"learning_rate": 3.50714075049563e-08,
"loss": 0.0254,
"step": 977
},
{
"epoch": 0.08412903225806452,
"grad_norm": 0.5209158658981323,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0445,
"step": 978
},
{
"epoch": 0.08421505376344086,
"grad_norm": 0.18631288409233093,
"learning_rate": 2.947247773079753e-08,
"loss": 0.002,
"step": 979
},
{
"epoch": 0.08430107526881721,
"grad_norm": 0.11395297199487686,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0013,
"step": 980
},
{
"epoch": 0.08438709677419355,
"grad_norm": 0.058965783566236496,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0009,
"step": 981
},
{
"epoch": 0.0844731182795699,
"grad_norm": 0.5768077373504639,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0248,
"step": 982
},
{
"epoch": 0.08455913978494624,
"grad_norm": 37.326847076416016,
"learning_rate": 1.973271571728441e-08,
"loss": 0.8894,
"step": 983
},
{
"epoch": 0.08464516129032258,
"grad_norm": 0.7202150225639343,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0408,
"step": 984
},
{
"epoch": 0.08473118279569893,
"grad_norm": 0.3096824884414673,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.0014,
"step": 985
},
{
"epoch": 0.08481720430107527,
"grad_norm": 0.15485045313835144,
"learning_rate": 1.370465245426167e-08,
"loss": 0.0026,
"step": 986
},
{
"epoch": 0.08490322580645161,
"grad_norm": 0.21106913685798645,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0402,
"step": 987
},
{
"epoch": 0.08498924731182796,
"grad_norm": 0.06831831485033035,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.0016,
"step": 988
},
{
"epoch": 0.0850752688172043,
"grad_norm": 0.3873961567878723,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0476,
"step": 989
},
{
"epoch": 0.08516129032258064,
"grad_norm": 0.5188955068588257,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0451,
"step": 990
},
{
"epoch": 0.08524731182795699,
"grad_norm": 2.4936110973358154,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0322,
"step": 991
},
{
"epoch": 0.08533333333333333,
"grad_norm": 0.2841807007789612,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.028,
"step": 992
},
{
"epoch": 0.08541935483870967,
"grad_norm": 0.05753358080983162,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0006,
"step": 993
},
{
"epoch": 0.08550537634408602,
"grad_norm": 0.23528507351875305,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0164,
"step": 994
},
{
"epoch": 0.08559139784946236,
"grad_norm": 0.46501123905181885,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0453,
"step": 995
},
{
"epoch": 0.0856774193548387,
"grad_norm": 25.617706298828125,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.3965,
"step": 996
},
{
"epoch": 0.08576344086021505,
"grad_norm": 2.257042407989502,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0286,
"step": 997
},
{
"epoch": 0.08584946236559139,
"grad_norm": 0.16915833950042725,
"learning_rate": 5.483063448785686e-10,
"loss": 0.017,
"step": 998
},
{
"epoch": 0.08593548387096774,
"grad_norm": 0.08071893453598022,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0015,
"step": 999
},
{
"epoch": 0.08602150537634409,
"grad_norm": 0.12956438958644867,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0408,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}