Enigma-8B-Entity-Attribute / trainer_state.json
SeacowX's picture
Upload folder using huggingface_hub
232bf27 verified
Raw
History Blame Contribute Delete
130 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 747,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004016064257028112,
"grad_norm": 3.3572497367858887,
"learning_rate": 6.666666666666667e-07,
"loss": 1.0681,
"step": 1
},
{
"epoch": 0.008032128514056224,
"grad_norm": 3.1153993606567383,
"learning_rate": 1.3333333333333334e-06,
"loss": 1.1112,
"step": 2
},
{
"epoch": 0.012048192771084338,
"grad_norm": 3.1947457790374756,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.1329,
"step": 3
},
{
"epoch": 0.01606425702811245,
"grad_norm": 3.0621848106384277,
"learning_rate": 2.666666666666667e-06,
"loss": 1.0596,
"step": 4
},
{
"epoch": 0.020080321285140562,
"grad_norm": 3.211728811264038,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.1193,
"step": 5
},
{
"epoch": 0.024096385542168676,
"grad_norm": 3.1836931705474854,
"learning_rate": 4.000000000000001e-06,
"loss": 1.2089,
"step": 6
},
{
"epoch": 0.028112449799196786,
"grad_norm": 3.296353816986084,
"learning_rate": 4.666666666666667e-06,
"loss": 1.1122,
"step": 7
},
{
"epoch": 0.0321285140562249,
"grad_norm": 3.3933396339416504,
"learning_rate": 5.333333333333334e-06,
"loss": 1.0387,
"step": 8
},
{
"epoch": 0.03614457831325301,
"grad_norm": 3.0436792373657227,
"learning_rate": 6e-06,
"loss": 1.0777,
"step": 9
},
{
"epoch": 0.040160642570281124,
"grad_norm": 3.169412612915039,
"learning_rate": 6.666666666666667e-06,
"loss": 1.1535,
"step": 10
},
{
"epoch": 0.04417670682730924,
"grad_norm": 3.418894052505493,
"learning_rate": 7.333333333333334e-06,
"loss": 0.9688,
"step": 11
},
{
"epoch": 0.04819277108433735,
"grad_norm": 3.1774702072143555,
"learning_rate": 8.000000000000001e-06,
"loss": 1.1101,
"step": 12
},
{
"epoch": 0.05220883534136546,
"grad_norm": 3.19480562210083,
"learning_rate": 8.666666666666668e-06,
"loss": 1.1184,
"step": 13
},
{
"epoch": 0.05622489959839357,
"grad_norm": 3.0001771450042725,
"learning_rate": 9.333333333333334e-06,
"loss": 1.0127,
"step": 14
},
{
"epoch": 0.060240963855421686,
"grad_norm": 3.100181818008423,
"learning_rate": 1e-05,
"loss": 1.0335,
"step": 15
},
{
"epoch": 0.0642570281124498,
"grad_norm": 2.895311117172241,
"learning_rate": 1.0666666666666667e-05,
"loss": 1.043,
"step": 16
},
{
"epoch": 0.06827309236947791,
"grad_norm": 3.066803455352783,
"learning_rate": 1.1333333333333334e-05,
"loss": 0.9846,
"step": 17
},
{
"epoch": 0.07228915662650602,
"grad_norm": 2.89982533454895,
"learning_rate": 1.2e-05,
"loss": 0.9096,
"step": 18
},
{
"epoch": 0.07630522088353414,
"grad_norm": 2.5824337005615234,
"learning_rate": 1.2666666666666668e-05,
"loss": 1.046,
"step": 19
},
{
"epoch": 0.08032128514056225,
"grad_norm": 2.0172696113586426,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.9038,
"step": 20
},
{
"epoch": 0.08433734939759036,
"grad_norm": 2.5625598430633545,
"learning_rate": 1.4000000000000001e-05,
"loss": 0.8794,
"step": 21
},
{
"epoch": 0.08835341365461848,
"grad_norm": 1.9606741666793823,
"learning_rate": 1.4666666666666668e-05,
"loss": 0.8385,
"step": 22
},
{
"epoch": 0.09236947791164658,
"grad_norm": 2.228515625,
"learning_rate": 1.5333333333333334e-05,
"loss": 0.7899,
"step": 23
},
{
"epoch": 0.0963855421686747,
"grad_norm": 1.8769129514694214,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.7948,
"step": 24
},
{
"epoch": 0.10040160642570281,
"grad_norm": 1.9379593133926392,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.8842,
"step": 25
},
{
"epoch": 0.10441767068273092,
"grad_norm": 1.6009212732315063,
"learning_rate": 1.7333333333333336e-05,
"loss": 0.7737,
"step": 26
},
{
"epoch": 0.10843373493975904,
"grad_norm": 1.4680776596069336,
"learning_rate": 1.8e-05,
"loss": 0.7688,
"step": 27
},
{
"epoch": 0.11244979919678715,
"grad_norm": 1.3376935720443726,
"learning_rate": 1.866666666666667e-05,
"loss": 0.7702,
"step": 28
},
{
"epoch": 0.11646586345381527,
"grad_norm": 1.205298900604248,
"learning_rate": 1.9333333333333333e-05,
"loss": 0.7429,
"step": 29
},
{
"epoch": 0.12048192771084337,
"grad_norm": 1.2713260650634766,
"learning_rate": 2e-05,
"loss": 0.6816,
"step": 30
},
{
"epoch": 0.12449799196787148,
"grad_norm": 1.3442246913909912,
"learning_rate": 2.0666666666666666e-05,
"loss": 0.6816,
"step": 31
},
{
"epoch": 0.1285140562248996,
"grad_norm": 1.2449231147766113,
"learning_rate": 2.1333333333333335e-05,
"loss": 0.5845,
"step": 32
},
{
"epoch": 0.13253012048192772,
"grad_norm": 1.3410189151763916,
"learning_rate": 2.2000000000000003e-05,
"loss": 0.6201,
"step": 33
},
{
"epoch": 0.13654618473895583,
"grad_norm": 0.9493803977966309,
"learning_rate": 2.2666666666666668e-05,
"loss": 0.5436,
"step": 34
},
{
"epoch": 0.14056224899598393,
"grad_norm": 1.295204758644104,
"learning_rate": 2.3333333333333336e-05,
"loss": 0.5349,
"step": 35
},
{
"epoch": 0.14457831325301204,
"grad_norm": 1.548696756362915,
"learning_rate": 2.4e-05,
"loss": 0.6341,
"step": 36
},
{
"epoch": 0.14859437751004015,
"grad_norm": 1.1955889463424683,
"learning_rate": 2.466666666666667e-05,
"loss": 0.6453,
"step": 37
},
{
"epoch": 0.15261044176706828,
"grad_norm": 1.0581564903259277,
"learning_rate": 2.5333333333333337e-05,
"loss": 0.5878,
"step": 38
},
{
"epoch": 0.1566265060240964,
"grad_norm": 1.0585978031158447,
"learning_rate": 2.6000000000000002e-05,
"loss": 0.5434,
"step": 39
},
{
"epoch": 0.1606425702811245,
"grad_norm": 0.966687023639679,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.5521,
"step": 40
},
{
"epoch": 0.1646586345381526,
"grad_norm": 1.033333659172058,
"learning_rate": 2.733333333333333e-05,
"loss": 0.5705,
"step": 41
},
{
"epoch": 0.1686746987951807,
"grad_norm": 1.0497894287109375,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.6386,
"step": 42
},
{
"epoch": 0.17269076305220885,
"grad_norm": 0.7324492931365967,
"learning_rate": 2.8666666666666668e-05,
"loss": 0.5261,
"step": 43
},
{
"epoch": 0.17670682730923695,
"grad_norm": 1.0644015073776245,
"learning_rate": 2.9333333333333336e-05,
"loss": 0.5311,
"step": 44
},
{
"epoch": 0.18072289156626506,
"grad_norm": 1.040525197982788,
"learning_rate": 3e-05,
"loss": 0.4845,
"step": 45
},
{
"epoch": 0.18473895582329317,
"grad_norm": 1.0211730003356934,
"learning_rate": 3.066666666666667e-05,
"loss": 0.4406,
"step": 46
},
{
"epoch": 0.18875502008032127,
"grad_norm": 0.8270606398582458,
"learning_rate": 3.1333333333333334e-05,
"loss": 0.4598,
"step": 47
},
{
"epoch": 0.1927710843373494,
"grad_norm": 0.6945549249649048,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.4978,
"step": 48
},
{
"epoch": 0.19678714859437751,
"grad_norm": 1.003801941871643,
"learning_rate": 3.266666666666667e-05,
"loss": 0.4607,
"step": 49
},
{
"epoch": 0.20080321285140562,
"grad_norm": 0.6564232110977173,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.4691,
"step": 50
},
{
"epoch": 0.20481927710843373,
"grad_norm": 1.2288357019424438,
"learning_rate": 3.4000000000000007e-05,
"loss": 0.4611,
"step": 51
},
{
"epoch": 0.20883534136546184,
"grad_norm": 0.7254999279975891,
"learning_rate": 3.466666666666667e-05,
"loss": 0.4266,
"step": 52
},
{
"epoch": 0.21285140562248997,
"grad_norm": 0.6874690055847168,
"learning_rate": 3.5333333333333336e-05,
"loss": 0.4364,
"step": 53
},
{
"epoch": 0.21686746987951808,
"grad_norm": 1.439631462097168,
"learning_rate": 3.6e-05,
"loss": 0.4736,
"step": 54
},
{
"epoch": 0.22088353413654618,
"grad_norm": 1.390151858329773,
"learning_rate": 3.6666666666666666e-05,
"loss": 0.4937,
"step": 55
},
{
"epoch": 0.2248995983935743,
"grad_norm": 1.206671118736267,
"learning_rate": 3.733333333333334e-05,
"loss": 0.5432,
"step": 56
},
{
"epoch": 0.2289156626506024,
"grad_norm": 0.7038202881813049,
"learning_rate": 3.8e-05,
"loss": 0.4592,
"step": 57
},
{
"epoch": 0.23293172690763053,
"grad_norm": 1.7638537883758545,
"learning_rate": 3.866666666666667e-05,
"loss": 0.4848,
"step": 58
},
{
"epoch": 0.23694779116465864,
"grad_norm": 2.070549488067627,
"learning_rate": 3.933333333333333e-05,
"loss": 0.4893,
"step": 59
},
{
"epoch": 0.24096385542168675,
"grad_norm": 1.4182060956954956,
"learning_rate": 4e-05,
"loss": 0.4641,
"step": 60
},
{
"epoch": 0.24497991967871485,
"grad_norm": 0.7965697646141052,
"learning_rate": 4.066666666666667e-05,
"loss": 0.485,
"step": 61
},
{
"epoch": 0.24899598393574296,
"grad_norm": 0.9898649454116821,
"learning_rate": 4.133333333333333e-05,
"loss": 0.4654,
"step": 62
},
{
"epoch": 0.25301204819277107,
"grad_norm": 1.1060651540756226,
"learning_rate": 4.2e-05,
"loss": 0.4796,
"step": 63
},
{
"epoch": 0.2570281124497992,
"grad_norm": 0.8289003968238831,
"learning_rate": 4.266666666666667e-05,
"loss": 0.498,
"step": 64
},
{
"epoch": 0.26104417670682734,
"grad_norm": 0.7476992011070251,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.4447,
"step": 65
},
{
"epoch": 0.26506024096385544,
"grad_norm": 0.6724223494529724,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.4553,
"step": 66
},
{
"epoch": 0.26907630522088355,
"grad_norm": 0.7126241326332092,
"learning_rate": 4.466666666666667e-05,
"loss": 0.499,
"step": 67
},
{
"epoch": 0.27309236947791166,
"grad_norm": 0.6537957191467285,
"learning_rate": 4.5333333333333335e-05,
"loss": 0.4127,
"step": 68
},
{
"epoch": 0.27710843373493976,
"grad_norm": 0.7121750712394714,
"learning_rate": 4.600000000000001e-05,
"loss": 0.4978,
"step": 69
},
{
"epoch": 0.28112449799196787,
"grad_norm": 0.8718213438987732,
"learning_rate": 4.666666666666667e-05,
"loss": 0.438,
"step": 70
},
{
"epoch": 0.285140562248996,
"grad_norm": 0.6130560040473938,
"learning_rate": 4.7333333333333336e-05,
"loss": 0.5058,
"step": 71
},
{
"epoch": 0.2891566265060241,
"grad_norm": 0.5417210459709167,
"learning_rate": 4.8e-05,
"loss": 0.425,
"step": 72
},
{
"epoch": 0.2931726907630522,
"grad_norm": 0.6405383348464966,
"learning_rate": 4.866666666666667e-05,
"loss": 0.4396,
"step": 73
},
{
"epoch": 0.2971887550200803,
"grad_norm": 0.6718979477882385,
"learning_rate": 4.933333333333334e-05,
"loss": 0.4653,
"step": 74
},
{
"epoch": 0.30120481927710846,
"grad_norm": 0.7254752516746521,
"learning_rate": 5e-05,
"loss": 0.5035,
"step": 75
},
{
"epoch": 0.30522088353413657,
"grad_norm": 0.6335278749465942,
"learning_rate": 4.9999726806462766e-05,
"loss": 0.4235,
"step": 76
},
{
"epoch": 0.3092369477911647,
"grad_norm": 0.6676427721977234,
"learning_rate": 4.9998907231821824e-05,
"loss": 0.4406,
"step": 77
},
{
"epoch": 0.3132530120481928,
"grad_norm": 0.6101539731025696,
"learning_rate": 4.9997541293989384e-05,
"loss": 0.4266,
"step": 78
},
{
"epoch": 0.3172690763052209,
"grad_norm": 0.5556290149688721,
"learning_rate": 4.999562902281866e-05,
"loss": 0.4065,
"step": 79
},
{
"epoch": 0.321285140562249,
"grad_norm": 0.5451353192329407,
"learning_rate": 4.999317046010329e-05,
"loss": 0.3893,
"step": 80
},
{
"epoch": 0.3253012048192771,
"grad_norm": 0.5386996865272522,
"learning_rate": 4.999016565957633e-05,
"loss": 0.3812,
"step": 81
},
{
"epoch": 0.3293172690763052,
"grad_norm": 0.8385406136512756,
"learning_rate": 4.9986614686909146e-05,
"loss": 0.4352,
"step": 82
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.7156203389167786,
"learning_rate": 4.998251761970997e-05,
"loss": 0.4083,
"step": 83
},
{
"epoch": 0.3373493975903614,
"grad_norm": 0.6663508415222168,
"learning_rate": 4.9977874547522175e-05,
"loss": 0.4039,
"step": 84
},
{
"epoch": 0.3413654618473896,
"grad_norm": 0.6586647629737854,
"learning_rate": 4.997268557182235e-05,
"loss": 0.4088,
"step": 85
},
{
"epoch": 0.3453815261044177,
"grad_norm": 0.7833324074745178,
"learning_rate": 4.9966950806018075e-05,
"loss": 0.3846,
"step": 86
},
{
"epoch": 0.3493975903614458,
"grad_norm": 0.712884247303009,
"learning_rate": 4.996067037544542e-05,
"loss": 0.3763,
"step": 87
},
{
"epoch": 0.3534136546184739,
"grad_norm": 0.8896615505218506,
"learning_rate": 4.995384441736621e-05,
"loss": 0.4284,
"step": 88
},
{
"epoch": 0.357429718875502,
"grad_norm": 0.9067183136940002,
"learning_rate": 4.994647308096509e-05,
"loss": 0.4275,
"step": 89
},
{
"epoch": 0.3614457831325301,
"grad_norm": 0.604626476764679,
"learning_rate": 4.9938556527346155e-05,
"loss": 0.3933,
"step": 90
},
{
"epoch": 0.3654618473895582,
"grad_norm": 0.783541202545166,
"learning_rate": 4.9930094929529506e-05,
"loss": 0.4127,
"step": 91
},
{
"epoch": 0.36947791164658633,
"grad_norm": 1.0507838726043701,
"learning_rate": 4.992108847244746e-05,
"loss": 0.4376,
"step": 92
},
{
"epoch": 0.37349397590361444,
"grad_norm": 0.7194378972053528,
"learning_rate": 4.991153735294049e-05,
"loss": 0.557,
"step": 93
},
{
"epoch": 0.37751004016064255,
"grad_norm": 0.6300519108772278,
"learning_rate": 4.99014417797529e-05,
"loss": 0.39,
"step": 94
},
{
"epoch": 0.3815261044176707,
"grad_norm": 0.8160332441329956,
"learning_rate": 4.989080197352834e-05,
"loss": 0.3785,
"step": 95
},
{
"epoch": 0.3855421686746988,
"grad_norm": 0.8603289723396301,
"learning_rate": 4.987961816680492e-05,
"loss": 0.3574,
"step": 96
},
{
"epoch": 0.3895582329317269,
"grad_norm": 0.7826237082481384,
"learning_rate": 4.9867890604010125e-05,
"loss": 0.49,
"step": 97
},
{
"epoch": 0.39357429718875503,
"grad_norm": 0.7405675649642944,
"learning_rate": 4.9855619541455514e-05,
"loss": 0.4629,
"step": 98
},
{
"epoch": 0.39759036144578314,
"grad_norm": 0.6662713885307312,
"learning_rate": 4.984280524733107e-05,
"loss": 0.3972,
"step": 99
},
{
"epoch": 0.40160642570281124,
"grad_norm": 0.915530264377594,
"learning_rate": 4.9829448001699384e-05,
"loss": 0.444,
"step": 100
},
{
"epoch": 0.40562248995983935,
"grad_norm": 0.697108268737793,
"learning_rate": 4.981554809648952e-05,
"loss": 0.4647,
"step": 101
},
{
"epoch": 0.40963855421686746,
"grad_norm": 0.739271342754364,
"learning_rate": 4.980110583549062e-05,
"loss": 0.377,
"step": 102
},
{
"epoch": 0.41365461847389556,
"grad_norm": 1.2061580419540405,
"learning_rate": 4.9786121534345265e-05,
"loss": 0.4533,
"step": 103
},
{
"epoch": 0.41767068273092367,
"grad_norm": 0.7795490026473999,
"learning_rate": 4.9770595520542594e-05,
"loss": 0.4601,
"step": 104
},
{
"epoch": 0.42168674698795183,
"grad_norm": 0.6940763592720032,
"learning_rate": 4.975452813341114e-05,
"loss": 0.4315,
"step": 105
},
{
"epoch": 0.42570281124497994,
"grad_norm": 0.8994523882865906,
"learning_rate": 4.973791972411141e-05,
"loss": 0.3887,
"step": 106
},
{
"epoch": 0.42971887550200805,
"grad_norm": 1.9176881313323975,
"learning_rate": 4.972077065562821e-05,
"loss": 0.455,
"step": 107
},
{
"epoch": 0.43373493975903615,
"grad_norm": 0.7927311062812805,
"learning_rate": 4.970308130276272e-05,
"loss": 0.4399,
"step": 108
},
{
"epoch": 0.43775100401606426,
"grad_norm": 0.6954143047332764,
"learning_rate": 4.9684852052124285e-05,
"loss": 0.3544,
"step": 109
},
{
"epoch": 0.44176706827309237,
"grad_norm": 0.8691537380218506,
"learning_rate": 4.966608330212198e-05,
"loss": 0.3807,
"step": 110
},
{
"epoch": 0.4457831325301205,
"grad_norm": 0.7648891806602478,
"learning_rate": 4.96467754629559e-05,
"loss": 0.3797,
"step": 111
},
{
"epoch": 0.4497991967871486,
"grad_norm": 0.7721076607704163,
"learning_rate": 4.96269289566082e-05,
"loss": 0.4214,
"step": 112
},
{
"epoch": 0.4538152610441767,
"grad_norm": 0.5924379229545593,
"learning_rate": 4.960654421683386e-05,
"loss": 0.4101,
"step": 113
},
{
"epoch": 0.4578313253012048,
"grad_norm": 0.9550824761390686,
"learning_rate": 4.9585621689151216e-05,
"loss": 0.426,
"step": 114
},
{
"epoch": 0.46184738955823296,
"grad_norm": 0.8562377095222473,
"learning_rate": 4.956416183083221e-05,
"loss": 0.3875,
"step": 115
},
{
"epoch": 0.46586345381526106,
"grad_norm": 0.6500234007835388,
"learning_rate": 4.9542165110892414e-05,
"loss": 0.4688,
"step": 116
},
{
"epoch": 0.46987951807228917,
"grad_norm": 0.596043586730957,
"learning_rate": 4.951963201008076e-05,
"loss": 0.434,
"step": 117
},
{
"epoch": 0.4738955823293173,
"grad_norm": 0.6146607398986816,
"learning_rate": 4.9496563020869064e-05,
"loss": 0.3896,
"step": 118
},
{
"epoch": 0.4779116465863454,
"grad_norm": 0.6440965533256531,
"learning_rate": 4.947295864744121e-05,
"loss": 0.4451,
"step": 119
},
{
"epoch": 0.4819277108433735,
"grad_norm": 0.7583677768707275,
"learning_rate": 4.9448819405682193e-05,
"loss": 0.3887,
"step": 120
},
{
"epoch": 0.4859437751004016,
"grad_norm": 0.6874030232429504,
"learning_rate": 4.9424145823166794e-05,
"loss": 0.3614,
"step": 121
},
{
"epoch": 0.4899598393574297,
"grad_norm": 0.644295334815979,
"learning_rate": 4.9398938439148077e-05,
"loss": 0.3961,
"step": 122
},
{
"epoch": 0.4939759036144578,
"grad_norm": 0.9563642144203186,
"learning_rate": 4.937319780454559e-05,
"loss": 0.3955,
"step": 123
},
{
"epoch": 0.4979919678714859,
"grad_norm": 0.5275897979736328,
"learning_rate": 4.934692448193334e-05,
"loss": 0.3378,
"step": 124
},
{
"epoch": 0.5020080321285141,
"grad_norm": 0.6813479661941528,
"learning_rate": 4.9320119045527487e-05,
"loss": 0.4176,
"step": 125
},
{
"epoch": 0.5060240963855421,
"grad_norm": 0.7719974517822266,
"learning_rate": 4.929278208117378e-05,
"loss": 0.3983,
"step": 126
},
{
"epoch": 0.5100401606425703,
"grad_norm": 0.627138614654541,
"learning_rate": 4.9264914186334775e-05,
"loss": 0.3915,
"step": 127
},
{
"epoch": 0.5140562248995983,
"grad_norm": 0.7721793055534363,
"learning_rate": 4.9236515970076785e-05,
"loss": 0.3566,
"step": 128
},
{
"epoch": 0.5180722891566265,
"grad_norm": 0.6848064661026001,
"learning_rate": 4.9207588053056545e-05,
"loss": 0.381,
"step": 129
},
{
"epoch": 0.5220883534136547,
"grad_norm": 0.578819215297699,
"learning_rate": 4.9178131067507625e-05,
"loss": 0.4308,
"step": 130
},
{
"epoch": 0.5261044176706827,
"grad_norm": 0.608585000038147,
"learning_rate": 4.914814565722671e-05,
"loss": 0.3979,
"step": 131
},
{
"epoch": 0.5301204819277109,
"grad_norm": 0.74711012840271,
"learning_rate": 4.91176324775594e-05,
"loss": 0.4111,
"step": 132
},
{
"epoch": 0.5341365461847389,
"grad_norm": 0.6809923052787781,
"learning_rate": 4.9086592195385974e-05,
"loss": 0.3647,
"step": 133
},
{
"epoch": 0.5381526104417671,
"grad_norm": 0.5864577889442444,
"learning_rate": 4.90550254891068e-05,
"loss": 0.3698,
"step": 134
},
{
"epoch": 0.5421686746987951,
"grad_norm": 0.7553641200065613,
"learning_rate": 4.9022933048627496e-05,
"loss": 0.4233,
"step": 135
},
{
"epoch": 0.5461847389558233,
"grad_norm": 0.7890750765800476,
"learning_rate": 4.899031557534383e-05,
"loss": 0.3721,
"step": 136
},
{
"epoch": 0.5502008032128514,
"grad_norm": 0.778347909450531,
"learning_rate": 4.895717378212644e-05,
"loss": 0.3746,
"step": 137
},
{
"epoch": 0.5542168674698795,
"grad_norm": 0.8391472101211548,
"learning_rate": 4.892350839330522e-05,
"loss": 0.3745,
"step": 138
},
{
"epoch": 0.5582329317269076,
"grad_norm": 0.6306824088096619,
"learning_rate": 4.888932014465352e-05,
"loss": 0.4011,
"step": 139
},
{
"epoch": 0.5622489959839357,
"grad_norm": 0.6019623875617981,
"learning_rate": 4.8854609783372014e-05,
"loss": 0.3061,
"step": 140
},
{
"epoch": 0.5662650602409639,
"grad_norm": 0.6332250833511353,
"learning_rate": 4.881937806807241e-05,
"loss": 0.3296,
"step": 141
},
{
"epoch": 0.570281124497992,
"grad_norm": 0.8003882765769958,
"learning_rate": 4.878362576876087e-05,
"loss": 0.3772,
"step": 142
},
{
"epoch": 0.5742971887550201,
"grad_norm": 1.5326228141784668,
"learning_rate": 4.874735366682115e-05,
"loss": 0.3881,
"step": 143
},
{
"epoch": 0.5783132530120482,
"grad_norm": 0.5541295409202576,
"learning_rate": 4.871056255499757e-05,
"loss": 0.3395,
"step": 144
},
{
"epoch": 0.5823293172690763,
"grad_norm": 0.7162477374076843,
"learning_rate": 4.867325323737765e-05,
"loss": 0.436,
"step": 145
},
{
"epoch": 0.5863453815261044,
"grad_norm": 1.1130383014678955,
"learning_rate": 4.863542652937453e-05,
"loss": 0.4736,
"step": 146
},
{
"epoch": 0.5903614457831325,
"grad_norm": 0.7125911712646484,
"learning_rate": 4.8597083257709194e-05,
"loss": 0.3639,
"step": 147
},
{
"epoch": 0.5943775100401606,
"grad_norm": 0.6064804196357727,
"learning_rate": 4.855822426039236e-05,
"loss": 0.4083,
"step": 148
},
{
"epoch": 0.5983935742971888,
"grad_norm": 0.6795706748962402,
"learning_rate": 4.851885038670618e-05,
"loss": 0.4017,
"step": 149
},
{
"epoch": 0.6024096385542169,
"grad_norm": 0.9338403940200806,
"learning_rate": 4.84789624971857e-05,
"loss": 0.3676,
"step": 150
},
{
"epoch": 0.606425702811245,
"grad_norm": 0.6994291543960571,
"learning_rate": 4.8438561463599984e-05,
"loss": 0.3956,
"step": 151
},
{
"epoch": 0.6104417670682731,
"grad_norm": 0.6114183068275452,
"learning_rate": 4.839764816893315e-05,
"loss": 0.3762,
"step": 152
},
{
"epoch": 0.6144578313253012,
"grad_norm": 0.6106306314468384,
"learning_rate": 4.8356223507364996e-05,
"loss": 0.5335,
"step": 153
},
{
"epoch": 0.6184738955823293,
"grad_norm": 0.773564875125885,
"learning_rate": 4.831428838425153e-05,
"loss": 0.3427,
"step": 154
},
{
"epoch": 0.6224899598393574,
"grad_norm": 0.7435568571090698,
"learning_rate": 4.827184371610511e-05,
"loss": 0.4057,
"step": 155
},
{
"epoch": 0.6265060240963856,
"grad_norm": 0.8714675903320312,
"learning_rate": 4.822889043057446e-05,
"loss": 0.3568,
"step": 156
},
{
"epoch": 0.6305220883534136,
"grad_norm": 0.5461587905883789,
"learning_rate": 4.818542946642438e-05,
"loss": 0.3559,
"step": 157
},
{
"epoch": 0.6345381526104418,
"grad_norm": 0.5777304768562317,
"learning_rate": 4.814146177351523e-05,
"loss": 0.3202,
"step": 158
},
{
"epoch": 0.6385542168674698,
"grad_norm": 0.7063300013542175,
"learning_rate": 4.8096988312782174e-05,
"loss": 0.4611,
"step": 159
},
{
"epoch": 0.642570281124498,
"grad_norm": 1.1671459674835205,
"learning_rate": 4.805201005621418e-05,
"loss": 0.4838,
"step": 160
},
{
"epoch": 0.6465863453815262,
"grad_norm": 0.8122550845146179,
"learning_rate": 4.800652798683277e-05,
"loss": 0.4497,
"step": 161
},
{
"epoch": 0.6506024096385542,
"grad_norm": 0.6299118399620056,
"learning_rate": 4.796054309867053e-05,
"loss": 0.3697,
"step": 162
},
{
"epoch": 0.6546184738955824,
"grad_norm": 0.7938956022262573,
"learning_rate": 4.791405639674941e-05,
"loss": 0.3955,
"step": 163
},
{
"epoch": 0.6586345381526104,
"grad_norm": 0.6391373872756958,
"learning_rate": 4.786706889705872e-05,
"loss": 0.3969,
"step": 164
},
{
"epoch": 0.6626506024096386,
"grad_norm": 0.7443742156028748,
"learning_rate": 4.781958162653297e-05,
"loss": 0.3577,
"step": 165
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.5645174384117126,
"learning_rate": 4.7771595623029394e-05,
"loss": 0.3535,
"step": 166
},
{
"epoch": 0.6706827309236948,
"grad_norm": 0.7596954107284546,
"learning_rate": 4.7723111935305274e-05,
"loss": 0.2914,
"step": 167
},
{
"epoch": 0.6746987951807228,
"grad_norm": 0.8062319159507751,
"learning_rate": 4.767413162299501e-05,
"loss": 0.4092,
"step": 168
},
{
"epoch": 0.678714859437751,
"grad_norm": 0.5653491616249084,
"learning_rate": 4.7624655756587e-05,
"loss": 0.3556,
"step": 169
},
{
"epoch": 0.6827309236947792,
"grad_norm": 1.0263564586639404,
"learning_rate": 4.757468541740019e-05,
"loss": 0.4144,
"step": 170
},
{
"epoch": 0.6867469879518072,
"grad_norm": 0.7135465145111084,
"learning_rate": 4.752422169756048e-05,
"loss": 0.3471,
"step": 171
},
{
"epoch": 0.6907630522088354,
"grad_norm": 0.5874855518341064,
"learning_rate": 4.747326569997684e-05,
"loss": 0.3333,
"step": 172
},
{
"epoch": 0.6947791164658634,
"grad_norm": 0.7225725650787354,
"learning_rate": 4.742181853831721e-05,
"loss": 0.4247,
"step": 173
},
{
"epoch": 0.6987951807228916,
"grad_norm": 0.8238813877105713,
"learning_rate": 4.7369881336984153e-05,
"loss": 0.4537,
"step": 174
},
{
"epoch": 0.7028112449799196,
"grad_norm": 0.825269341468811,
"learning_rate": 4.731745523109029e-05,
"loss": 0.4262,
"step": 175
},
{
"epoch": 0.7068273092369478,
"grad_norm": 0.7465764284133911,
"learning_rate": 4.7264541366433494e-05,
"loss": 0.3829,
"step": 176
},
{
"epoch": 0.7108433734939759,
"grad_norm": 0.5821319818496704,
"learning_rate": 4.721114089947181e-05,
"loss": 0.3412,
"step": 177
},
{
"epoch": 0.714859437751004,
"grad_norm": 0.66474449634552,
"learning_rate": 4.715725499729826e-05,
"loss": 0.4036,
"step": 178
},
{
"epoch": 0.7188755020080321,
"grad_norm": 0.7002742290496826,
"learning_rate": 4.7102884837615244e-05,
"loss": 0.3917,
"step": 179
},
{
"epoch": 0.7228915662650602,
"grad_norm": 0.7928130626678467,
"learning_rate": 4.7048031608708876e-05,
"loss": 0.3368,
"step": 180
},
{
"epoch": 0.7269076305220884,
"grad_norm": 0.7208336591720581,
"learning_rate": 4.699269650942296e-05,
"loss": 0.4802,
"step": 181
},
{
"epoch": 0.7309236947791165,
"grad_norm": 0.5339462757110596,
"learning_rate": 4.693688074913282e-05,
"loss": 0.3881,
"step": 182
},
{
"epoch": 0.7349397590361446,
"grad_norm": 0.6376134753227234,
"learning_rate": 4.6880585547718845e-05,
"loss": 0.3136,
"step": 183
},
{
"epoch": 0.7389558232931727,
"grad_norm": 0.7227637767791748,
"learning_rate": 4.682381213553986e-05,
"loss": 0.3777,
"step": 184
},
{
"epoch": 0.7429718875502008,
"grad_norm": 0.7167319059371948,
"learning_rate": 4.676656175340621e-05,
"loss": 0.3804,
"step": 185
},
{
"epoch": 0.7469879518072289,
"grad_norm": 0.7428449392318726,
"learning_rate": 4.670883565255264e-05,
"loss": 0.3574,
"step": 186
},
{
"epoch": 0.751004016064257,
"grad_norm": 0.700078010559082,
"learning_rate": 4.665063509461097e-05,
"loss": 0.3891,
"step": 187
},
{
"epoch": 0.7550200803212851,
"grad_norm": 0.6233504414558411,
"learning_rate": 4.65919613515825e-05,
"loss": 0.3581,
"step": 188
},
{
"epoch": 0.7590361445783133,
"grad_norm": 0.7311496138572693,
"learning_rate": 4.653281570581023e-05,
"loss": 0.4229,
"step": 189
},
{
"epoch": 0.7630522088353414,
"grad_norm": 0.7398768663406372,
"learning_rate": 4.64731994499508e-05,
"loss": 0.3893,
"step": 190
},
{
"epoch": 0.7670682730923695,
"grad_norm": 0.6437134742736816,
"learning_rate": 4.641311388694629e-05,
"loss": 0.3548,
"step": 191
},
{
"epoch": 0.7710843373493976,
"grad_norm": 0.6008043885231018,
"learning_rate": 4.6352560329995686e-05,
"loss": 0.3488,
"step": 192
},
{
"epoch": 0.7751004016064257,
"grad_norm": 0.6803656220436096,
"learning_rate": 4.6291540102526235e-05,
"loss": 0.3711,
"step": 193
},
{
"epoch": 0.7791164658634538,
"grad_norm": 0.6243724822998047,
"learning_rate": 4.6230054538164474e-05,
"loss": 0.3355,
"step": 194
},
{
"epoch": 0.7831325301204819,
"grad_norm": 0.7050710916519165,
"learning_rate": 4.6168104980707107e-05,
"loss": 0.4309,
"step": 195
},
{
"epoch": 0.7871485943775101,
"grad_norm": 0.6469572186470032,
"learning_rate": 4.6105692784091636e-05,
"loss": 0.3517,
"step": 196
},
{
"epoch": 0.7911646586345381,
"grad_norm": 0.6431530117988586,
"learning_rate": 4.6042819312366755e-05,
"loss": 0.3819,
"step": 197
},
{
"epoch": 0.7951807228915663,
"grad_norm": 0.6341055035591125,
"learning_rate": 4.597948593966256e-05,
"loss": 0.2941,
"step": 198
},
{
"epoch": 0.7991967871485943,
"grad_norm": 0.5747613906860352,
"learning_rate": 4.591569405016049e-05,
"loss": 0.3735,
"step": 199
},
{
"epoch": 0.8032128514056225,
"grad_norm": 0.658657968044281,
"learning_rate": 4.585144503806312e-05,
"loss": 0.3994,
"step": 200
},
{
"epoch": 0.8072289156626506,
"grad_norm": 0.6977449655532837,
"learning_rate": 4.5786740307563636e-05,
"loss": 0.3653,
"step": 201
},
{
"epoch": 0.8112449799196787,
"grad_norm": 0.6175029277801514,
"learning_rate": 4.5721581272815164e-05,
"loss": 0.3333,
"step": 202
},
{
"epoch": 0.8152610441767069,
"grad_norm": 0.6854798793792725,
"learning_rate": 4.5655969357899874e-05,
"loss": 0.3784,
"step": 203
},
{
"epoch": 0.8192771084337349,
"grad_norm": 0.6330843567848206,
"learning_rate": 4.558990599679787e-05,
"loss": 0.3559,
"step": 204
},
{
"epoch": 0.8232931726907631,
"grad_norm": 0.5856051445007324,
"learning_rate": 4.552339263335581e-05,
"loss": 0.3886,
"step": 205
},
{
"epoch": 0.8273092369477911,
"grad_norm": 0.6043461561203003,
"learning_rate": 4.545643072125538e-05,
"loss": 0.3141,
"step": 206
},
{
"epoch": 0.8313253012048193,
"grad_norm": 0.6707512736320496,
"learning_rate": 4.538902172398151e-05,
"loss": 0.3362,
"step": 207
},
{
"epoch": 0.8353413654618473,
"grad_norm": 0.5556585192680359,
"learning_rate": 4.5321167114790385e-05,
"loss": 0.3758,
"step": 208
},
{
"epoch": 0.8393574297188755,
"grad_norm": 0.6057583093643188,
"learning_rate": 4.525286837667727e-05,
"loss": 0.3374,
"step": 209
},
{
"epoch": 0.8433734939759037,
"grad_norm": 0.6368465423583984,
"learning_rate": 4.518412700234406e-05,
"loss": 0.3359,
"step": 210
},
{
"epoch": 0.8473895582329317,
"grad_norm": 0.5859608054161072,
"learning_rate": 4.511494449416671e-05,
"loss": 0.2977,
"step": 211
},
{
"epoch": 0.8514056224899599,
"grad_norm": 0.6386719346046448,
"learning_rate": 4.504532236416234e-05,
"loss": 0.349,
"step": 212
},
{
"epoch": 0.8554216867469879,
"grad_norm": 0.7043953537940979,
"learning_rate": 4.497526213395623e-05,
"loss": 0.35,
"step": 213
},
{
"epoch": 0.8594377510040161,
"grad_norm": 0.6599947214126587,
"learning_rate": 4.490476533474855e-05,
"loss": 0.3626,
"step": 214
},
{
"epoch": 0.8634538152610441,
"grad_norm": 0.709950864315033,
"learning_rate": 4.4833833507280884e-05,
"loss": 0.3953,
"step": 215
},
{
"epoch": 0.8674698795180723,
"grad_norm": 0.6372776031494141,
"learning_rate": 4.476246820180259e-05,
"loss": 0.3446,
"step": 216
},
{
"epoch": 0.8714859437751004,
"grad_norm": 0.6113653779029846,
"learning_rate": 4.4690670978036886e-05,
"loss": 0.3912,
"step": 217
},
{
"epoch": 0.8755020080321285,
"grad_norm": 0.6425395607948303,
"learning_rate": 4.461844340514678e-05,
"loss": 0.376,
"step": 218
},
{
"epoch": 0.8795180722891566,
"grad_norm": 0.6444389224052429,
"learning_rate": 4.454578706170075e-05,
"loss": 0.2973,
"step": 219
},
{
"epoch": 0.8835341365461847,
"grad_norm": 1.6042680740356445,
"learning_rate": 4.447270353563828e-05,
"loss": 0.3803,
"step": 220
},
{
"epoch": 0.8875502008032129,
"grad_norm": 0.7146402597427368,
"learning_rate": 4.439919442423513e-05,
"loss": 0.3579,
"step": 221
},
{
"epoch": 0.891566265060241,
"grad_norm": 0.6526602506637573,
"learning_rate": 4.4325261334068426e-05,
"loss": 0.3645,
"step": 222
},
{
"epoch": 0.8955823293172691,
"grad_norm": 0.8015785217285156,
"learning_rate": 4.4250905880981574e-05,
"loss": 0.4367,
"step": 223
},
{
"epoch": 0.8995983935742972,
"grad_norm": 1.0051459074020386,
"learning_rate": 4.417612969004891e-05,
"loss": 0.3873,
"step": 224
},
{
"epoch": 0.9036144578313253,
"grad_norm": 0.7158998847007751,
"learning_rate": 4.410093439554019e-05,
"loss": 0.3457,
"step": 225
},
{
"epoch": 0.9076305220883534,
"grad_norm": 0.6931036114692688,
"learning_rate": 4.4025321640884905e-05,
"loss": 0.2949,
"step": 226
},
{
"epoch": 0.9116465863453815,
"grad_norm": 0.7364740371704102,
"learning_rate": 4.394929307863633e-05,
"loss": 0.3245,
"step": 227
},
{
"epoch": 0.9156626506024096,
"grad_norm": 0.8082252144813538,
"learning_rate": 4.38728503704354e-05,
"loss": 0.359,
"step": 228
},
{
"epoch": 0.9196787148594378,
"grad_norm": 1.1371561288833618,
"learning_rate": 4.379599518697444e-05,
"loss": 0.3495,
"step": 229
},
{
"epoch": 0.9236947791164659,
"grad_norm": 0.7997297048568726,
"learning_rate": 4.3718729207960586e-05,
"loss": 0.3186,
"step": 230
},
{
"epoch": 0.927710843373494,
"grad_norm": 0.7108686566352844,
"learning_rate": 4.364105412207914e-05,
"loss": 0.4086,
"step": 231
},
{
"epoch": 0.9317269076305221,
"grad_norm": 0.7407398223876953,
"learning_rate": 4.356297162695662e-05,
"loss": 0.3657,
"step": 232
},
{
"epoch": 0.9357429718875502,
"grad_norm": 0.7209028005599976,
"learning_rate": 4.3484483429123656e-05,
"loss": 0.3305,
"step": 233
},
{
"epoch": 0.9397590361445783,
"grad_norm": 0.841200590133667,
"learning_rate": 4.3405591243977736e-05,
"loss": 0.4038,
"step": 234
},
{
"epoch": 0.9437751004016064,
"grad_norm": 0.6993167400360107,
"learning_rate": 4.332629679574566e-05,
"loss": 0.3341,
"step": 235
},
{
"epoch": 0.9477911646586346,
"grad_norm": 0.6530373096466064,
"learning_rate": 4.324660181744589e-05,
"loss": 0.3926,
"step": 236
},
{
"epoch": 0.9518072289156626,
"grad_norm": 0.6918975710868835,
"learning_rate": 4.316650805085068e-05,
"loss": 0.3806,
"step": 237
},
{
"epoch": 0.9558232931726908,
"grad_norm": 0.6316202282905579,
"learning_rate": 4.308601724644797e-05,
"loss": 0.3622,
"step": 238
},
{
"epoch": 0.9598393574297188,
"grad_norm": 0.6180090308189392,
"learning_rate": 4.300513116340317e-05,
"loss": 0.3748,
"step": 239
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.6538357138633728,
"learning_rate": 4.2923851569520685e-05,
"loss": 0.2862,
"step": 240
},
{
"epoch": 0.9678714859437751,
"grad_norm": 0.774797797203064,
"learning_rate": 4.28421802412053e-05,
"loss": 0.3484,
"step": 241
},
{
"epoch": 0.9718875502008032,
"grad_norm": 0.6641711592674255,
"learning_rate": 4.276011896342336e-05,
"loss": 0.3897,
"step": 242
},
{
"epoch": 0.9759036144578314,
"grad_norm": 0.5238835215568542,
"learning_rate": 4.267766952966369e-05,
"loss": 0.3045,
"step": 243
},
{
"epoch": 0.9799196787148594,
"grad_norm": 0.6551547646522522,
"learning_rate": 4.259483374189851e-05,
"loss": 0.2997,
"step": 244
},
{
"epoch": 0.9839357429718876,
"grad_norm": 0.7510992288589478,
"learning_rate": 4.251161341054396e-05,
"loss": 0.3996,
"step": 245
},
{
"epoch": 0.9879518072289156,
"grad_norm": 0.6663261651992798,
"learning_rate": 4.242801035442058e-05,
"loss": 0.3949,
"step": 246
},
{
"epoch": 0.9919678714859438,
"grad_norm": 0.632393479347229,
"learning_rate": 4.2344026400713545e-05,
"loss": 0.3046,
"step": 247
},
{
"epoch": 0.9959839357429718,
"grad_norm": 0.6645919680595398,
"learning_rate": 4.225966338493271e-05,
"loss": 0.2913,
"step": 248
},
{
"epoch": 1.0,
"grad_norm": 0.622279942035675,
"learning_rate": 4.2174923150872544e-05,
"loss": 0.3214,
"step": 249
},
{
"epoch": 1.0040160642570282,
"grad_norm": 0.6085844039916992,
"learning_rate": 4.208980755057178e-05,
"loss": 0.398,
"step": 250
},
{
"epoch": 1.0080321285140563,
"grad_norm": 0.5199077129364014,
"learning_rate": 4.2004318444272985e-05,
"loss": 0.3322,
"step": 251
},
{
"epoch": 1.0120481927710843,
"grad_norm": 0.7024986743927002,
"learning_rate": 4.1918457700381855e-05,
"loss": 0.3984,
"step": 252
},
{
"epoch": 1.0160642570281124,
"grad_norm": 0.6403225660324097,
"learning_rate": 4.183222719542643e-05,
"loss": 0.3169,
"step": 253
},
{
"epoch": 1.0200803212851406,
"grad_norm": 0.7067918181419373,
"learning_rate": 4.174562881401602e-05,
"loss": 0.2851,
"step": 254
},
{
"epoch": 1.0240963855421688,
"grad_norm": 0.8557071685791016,
"learning_rate": 4.16586644488001e-05,
"loss": 0.2954,
"step": 255
},
{
"epoch": 1.0281124497991967,
"grad_norm": 0.6150246858596802,
"learning_rate": 4.157133600042686e-05,
"loss": 0.3925,
"step": 256
},
{
"epoch": 1.0321285140562249,
"grad_norm": 0.629190981388092,
"learning_rate": 4.148364537750172e-05,
"loss": 0.351,
"step": 257
},
{
"epoch": 1.036144578313253,
"grad_norm": 0.6059038639068604,
"learning_rate": 4.139559449654561e-05,
"loss": 0.3343,
"step": 258
},
{
"epoch": 1.0401606425702812,
"grad_norm": 0.6396567225456238,
"learning_rate": 4.130718528195303e-05,
"loss": 0.3432,
"step": 259
},
{
"epoch": 1.0441767068273093,
"grad_norm": 0.6217969059944153,
"learning_rate": 4.1218419665950094e-05,
"loss": 0.3487,
"step": 260
},
{
"epoch": 1.0481927710843373,
"grad_norm": 0.736483633518219,
"learning_rate": 4.1129299588552193e-05,
"loss": 0.3325,
"step": 261
},
{
"epoch": 1.0522088353413654,
"grad_norm": 0.6012124419212341,
"learning_rate": 4.103982699752167e-05,
"loss": 0.3232,
"step": 262
},
{
"epoch": 1.0562248995983936,
"grad_norm": 0.5797888040542603,
"learning_rate": 4.095000384832522e-05,
"loss": 0.3664,
"step": 263
},
{
"epoch": 1.0602409638554218,
"grad_norm": 0.6586582064628601,
"learning_rate": 4.085983210409114e-05,
"loss": 0.3973,
"step": 264
},
{
"epoch": 1.0642570281124497,
"grad_norm": 0.7206650972366333,
"learning_rate": 4.076931373556646e-05,
"loss": 0.3004,
"step": 265
},
{
"epoch": 1.0682730923694779,
"grad_norm": 0.8033329248428345,
"learning_rate": 4.067845072107383e-05,
"loss": 0.3916,
"step": 266
},
{
"epoch": 1.072289156626506,
"grad_norm": 0.8563597202301025,
"learning_rate": 4.058724504646834e-05,
"loss": 0.351,
"step": 267
},
{
"epoch": 1.0763052208835342,
"grad_norm": 0.7244225740432739,
"learning_rate": 4.0495698705094044e-05,
"loss": 0.3782,
"step": 268
},
{
"epoch": 1.0803212851405624,
"grad_norm": 0.704351007938385,
"learning_rate": 4.0403813697740446e-05,
"loss": 0.3864,
"step": 269
},
{
"epoch": 1.0843373493975903,
"grad_norm": 0.6686497926712036,
"learning_rate": 4.0311592032598754e-05,
"loss": 0.3588,
"step": 270
},
{
"epoch": 1.0883534136546185,
"grad_norm": 0.7728563547134399,
"learning_rate": 4.021903572521802e-05,
"loss": 0.354,
"step": 271
},
{
"epoch": 1.0923694779116466,
"grad_norm": 0.6716745495796204,
"learning_rate": 4.012614679846103e-05,
"loss": 0.3481,
"step": 272
},
{
"epoch": 1.0963855421686748,
"grad_norm": 0.6699212193489075,
"learning_rate": 4.0032927282460146e-05,
"loss": 0.3038,
"step": 273
},
{
"epoch": 1.1004016064257027,
"grad_norm": 0.5542762875556946,
"learning_rate": 3.9939379214572915e-05,
"loss": 0.3026,
"step": 274
},
{
"epoch": 1.104417670682731,
"grad_norm": 0.6108137369155884,
"learning_rate": 3.9845504639337535e-05,
"loss": 0.2906,
"step": 275
},
{
"epoch": 1.108433734939759,
"grad_norm": 0.6392081379890442,
"learning_rate": 3.9751305608428205e-05,
"loss": 0.3387,
"step": 276
},
{
"epoch": 1.1124497991967872,
"grad_norm": 0.6696359515190125,
"learning_rate": 3.965678418061023e-05,
"loss": 0.2961,
"step": 277
},
{
"epoch": 1.1164658634538154,
"grad_norm": 0.6447118520736694,
"learning_rate": 3.956194242169506e-05,
"loss": 0.2949,
"step": 278
},
{
"epoch": 1.1204819277108433,
"grad_norm": 1.0393074750900269,
"learning_rate": 3.946678240449515e-05,
"loss": 0.3566,
"step": 279
},
{
"epoch": 1.1244979919678715,
"grad_norm": 0.630760908126831,
"learning_rate": 3.937130620877863e-05,
"loss": 0.3576,
"step": 280
},
{
"epoch": 1.1285140562248996,
"grad_norm": 0.8362942934036255,
"learning_rate": 3.927551592122388e-05,
"loss": 0.4025,
"step": 281
},
{
"epoch": 1.1325301204819278,
"grad_norm": 0.7339195013046265,
"learning_rate": 3.9179413635373897e-05,
"loss": 0.3454,
"step": 282
},
{
"epoch": 1.1365461847389557,
"grad_norm": 0.8447798490524292,
"learning_rate": 3.908300145159055e-05,
"loss": 0.2539,
"step": 283
},
{
"epoch": 1.140562248995984,
"grad_norm": 0.8059999942779541,
"learning_rate": 3.898628147700868e-05,
"loss": 0.3638,
"step": 284
},
{
"epoch": 1.144578313253012,
"grad_norm": 0.7167591452598572,
"learning_rate": 3.888925582549006e-05,
"loss": 0.3427,
"step": 285
},
{
"epoch": 1.1485943775100402,
"grad_norm": 0.6503918170928955,
"learning_rate": 3.8791926617577144e-05,
"loss": 0.3232,
"step": 286
},
{
"epoch": 1.1526104417670684,
"grad_norm": 0.9456886649131775,
"learning_rate": 3.869429598044679e-05,
"loss": 0.3642,
"step": 287
},
{
"epoch": 1.1566265060240963,
"grad_norm": 0.6594911813735962,
"learning_rate": 3.859636604786372e-05,
"loss": 0.2961,
"step": 288
},
{
"epoch": 1.1606425702811245,
"grad_norm": 0.6953052282333374,
"learning_rate": 3.8498138960133917e-05,
"loss": 0.2412,
"step": 289
},
{
"epoch": 1.1646586345381527,
"grad_norm": 0.641533374786377,
"learning_rate": 3.8399616864057816e-05,
"loss": 0.3229,
"step": 290
},
{
"epoch": 1.1686746987951806,
"grad_norm": 0.6257909536361694,
"learning_rate": 3.830080191288342e-05,
"loss": 0.3686,
"step": 291
},
{
"epoch": 1.1726907630522088,
"grad_norm": 0.675899088382721,
"learning_rate": 3.82016962662592e-05,
"loss": 0.3846,
"step": 292
},
{
"epoch": 1.176706827309237,
"grad_norm": 0.6206428408622742,
"learning_rate": 3.810230209018694e-05,
"loss": 0.2751,
"step": 293
},
{
"epoch": 1.180722891566265,
"grad_norm": 0.7517561912536621,
"learning_rate": 3.8002621556974367e-05,
"loss": 0.3071,
"step": 294
},
{
"epoch": 1.1847389558232932,
"grad_norm": 0.7836295962333679,
"learning_rate": 3.790265684518767e-05,
"loss": 0.3599,
"step": 295
},
{
"epoch": 1.1887550200803212,
"grad_norm": 0.6151822209358215,
"learning_rate": 3.780241013960391e-05,
"loss": 0.3623,
"step": 296
},
{
"epoch": 1.1927710843373494,
"grad_norm": 0.623931348323822,
"learning_rate": 3.770188363116324e-05,
"loss": 0.2892,
"step": 297
},
{
"epoch": 1.1967871485943775,
"grad_norm": 0.9910692572593689,
"learning_rate": 3.760107951692107e-05,
"loss": 0.336,
"step": 298
},
{
"epoch": 1.2008032128514057,
"grad_norm": 0.7002805471420288,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.2656,
"step": 299
},
{
"epoch": 1.2048192771084336,
"grad_norm": 0.7750109434127808,
"learning_rate": 3.73986472895417e-05,
"loss": 0.3896,
"step": 300
},
{
"epoch": 1.2088353413654618,
"grad_norm": 0.7624642252922058,
"learning_rate": 3.729702360065859e-05,
"loss": 0.36,
"step": 301
},
{
"epoch": 1.21285140562249,
"grad_norm": 0.629544734954834,
"learning_rate": 3.7195131154385474e-05,
"loss": 0.2891,
"step": 302
},
{
"epoch": 1.216867469879518,
"grad_norm": 0.692724883556366,
"learning_rate": 3.7092972177631e-05,
"loss": 0.4024,
"step": 303
},
{
"epoch": 1.2208835341365463,
"grad_norm": 0.7353823184967041,
"learning_rate": 3.699054890312892e-05,
"loss": 0.4004,
"step": 304
},
{
"epoch": 1.2248995983935742,
"grad_norm": 0.6432971358299255,
"learning_rate": 3.6887863569389386e-05,
"loss": 0.2805,
"step": 305
},
{
"epoch": 1.2289156626506024,
"grad_norm": 0.7083505392074585,
"learning_rate": 3.678491842064995e-05,
"loss": 0.3511,
"step": 306
},
{
"epoch": 1.2329317269076305,
"grad_norm": 0.6677691340446472,
"learning_rate": 3.668171570682655e-05,
"loss": 0.2609,
"step": 307
},
{
"epoch": 1.2369477911646587,
"grad_norm": 0.6491400599479675,
"learning_rate": 3.6578257683464364e-05,
"loss": 0.2552,
"step": 308
},
{
"epoch": 1.2409638554216866,
"grad_norm": 0.7234424948692322,
"learning_rate": 3.6474546611688445e-05,
"loss": 0.3698,
"step": 309
},
{
"epoch": 1.2449799196787148,
"grad_norm": 0.6780205368995667,
"learning_rate": 3.6370584758154366e-05,
"loss": 0.3884,
"step": 310
},
{
"epoch": 1.248995983935743,
"grad_norm": 0.8064125180244446,
"learning_rate": 3.6266374394998634e-05,
"loss": 0.3976,
"step": 311
},
{
"epoch": 1.2530120481927711,
"grad_norm": 0.9440551400184631,
"learning_rate": 3.616191779978907e-05,
"loss": 0.4309,
"step": 312
},
{
"epoch": 1.2570281124497993,
"grad_norm": 0.6932246685028076,
"learning_rate": 3.6057217255475034e-05,
"loss": 0.3404,
"step": 313
},
{
"epoch": 1.2610441767068274,
"grad_norm": 0.7167385816574097,
"learning_rate": 3.595227505033747e-05,
"loss": 0.367,
"step": 314
},
{
"epoch": 1.2650602409638554,
"grad_norm": 0.7061241865158081,
"learning_rate": 3.5847093477938956e-05,
"loss": 0.3075,
"step": 315
},
{
"epoch": 1.2690763052208835,
"grad_norm": 0.7282212376594543,
"learning_rate": 3.574167483707356e-05,
"loss": 0.3986,
"step": 316
},
{
"epoch": 1.2730923694779117,
"grad_norm": 0.6596242189407349,
"learning_rate": 3.5636021431716606e-05,
"loss": 0.3331,
"step": 317
},
{
"epoch": 1.2771084337349397,
"grad_norm": 0.6918690204620361,
"learning_rate": 3.5530135570974274e-05,
"loss": 0.3008,
"step": 318
},
{
"epoch": 1.2811244979919678,
"grad_norm": 0.8608424067497253,
"learning_rate": 3.542401956903321e-05,
"loss": 0.3083,
"step": 319
},
{
"epoch": 1.285140562248996,
"grad_norm": 0.665269136428833,
"learning_rate": 3.5317675745109866e-05,
"loss": 0.3158,
"step": 320
},
{
"epoch": 1.2891566265060241,
"grad_norm": 0.6643700003623962,
"learning_rate": 3.521110642339991e-05,
"loss": 0.3511,
"step": 321
},
{
"epoch": 1.2931726907630523,
"grad_norm": 0.6341935992240906,
"learning_rate": 3.51043139330273e-05,
"loss": 0.2854,
"step": 322
},
{
"epoch": 1.2971887550200802,
"grad_norm": 0.6715558171272278,
"learning_rate": 3.499730060799352e-05,
"loss": 0.2628,
"step": 323
},
{
"epoch": 1.3012048192771084,
"grad_norm": 0.7047196626663208,
"learning_rate": 3.489006878712647e-05,
"loss": 0.3705,
"step": 324
},
{
"epoch": 1.3052208835341366,
"grad_norm": 0.7242834568023682,
"learning_rate": 3.4782620814029375e-05,
"loss": 0.3081,
"step": 325
},
{
"epoch": 1.3092369477911647,
"grad_norm": 0.7659518718719482,
"learning_rate": 3.4674959037029594e-05,
"loss": 0.3186,
"step": 326
},
{
"epoch": 1.3132530120481927,
"grad_norm": 0.836984395980835,
"learning_rate": 3.456708580912725e-05,
"loss": 0.3965,
"step": 327
},
{
"epoch": 1.3172690763052208,
"grad_norm": 0.6630202531814575,
"learning_rate": 3.4459003487943844e-05,
"loss": 0.2702,
"step": 328
},
{
"epoch": 1.321285140562249,
"grad_norm": 0.7526226043701172,
"learning_rate": 3.4350714435670707e-05,
"loss": 0.3367,
"step": 329
},
{
"epoch": 1.3253012048192772,
"grad_norm": 0.6993024349212646,
"learning_rate": 3.424222101901738e-05,
"loss": 0.3486,
"step": 330
},
{
"epoch": 1.3293172690763053,
"grad_norm": 0.8149450421333313,
"learning_rate": 3.413352560915988e-05,
"loss": 0.3661,
"step": 331
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.6562392115592957,
"learning_rate": 3.4024630581688894e-05,
"loss": 0.2749,
"step": 332
},
{
"epoch": 1.3373493975903614,
"grad_norm": 0.7134260535240173,
"learning_rate": 3.391553831655782e-05,
"loss": 0.3177,
"step": 333
},
{
"epoch": 1.3413654618473896,
"grad_norm": 0.7992382049560547,
"learning_rate": 3.380625119803084e-05,
"loss": 0.3613,
"step": 334
},
{
"epoch": 1.3453815261044177,
"grad_norm": 0.7649348378181458,
"learning_rate": 3.369677161463068e-05,
"loss": 0.3374,
"step": 335
},
{
"epoch": 1.3493975903614457,
"grad_norm": 0.8444297313690186,
"learning_rate": 3.358710195908653e-05,
"loss": 0.3972,
"step": 336
},
{
"epoch": 1.3534136546184738,
"grad_norm": 0.6742205619812012,
"learning_rate": 3.3477244628281666e-05,
"loss": 0.3344,
"step": 337
},
{
"epoch": 1.357429718875502,
"grad_norm": 0.6053270697593689,
"learning_rate": 3.336720202320113e-05,
"loss": 0.3483,
"step": 338
},
{
"epoch": 1.3614457831325302,
"grad_norm": 0.6525555849075317,
"learning_rate": 3.3256976548879184e-05,
"loss": 0.2982,
"step": 339
},
{
"epoch": 1.3654618473895583,
"grad_norm": 0.6057913899421692,
"learning_rate": 3.314657061434681e-05,
"loss": 0.2809,
"step": 340
},
{
"epoch": 1.3694779116465863,
"grad_norm": 0.6161317229270935,
"learning_rate": 3.303598663257904e-05,
"loss": 0.3059,
"step": 341
},
{
"epoch": 1.3734939759036144,
"grad_norm": 0.6134277582168579,
"learning_rate": 3.292522702044221e-05,
"loss": 0.2933,
"step": 342
},
{
"epoch": 1.3775100401606426,
"grad_norm": 0.7496150135993958,
"learning_rate": 3.281429419864112e-05,
"loss": 0.3587,
"step": 343
},
{
"epoch": 1.3815261044176708,
"grad_norm": 0.7023509740829468,
"learning_rate": 3.270319059166617e-05,
"loss": 0.2981,
"step": 344
},
{
"epoch": 1.3855421686746987,
"grad_norm": 0.6817592978477478,
"learning_rate": 3.259191862774037e-05,
"loss": 0.3579,
"step": 345
},
{
"epoch": 1.3895582329317269,
"grad_norm": 0.6259456276893616,
"learning_rate": 3.248048073876622e-05,
"loss": 0.2917,
"step": 346
},
{
"epoch": 1.393574297188755,
"grad_norm": 0.6996810436248779,
"learning_rate": 3.2368879360272606e-05,
"loss": 0.2643,
"step": 347
},
{
"epoch": 1.3975903614457832,
"grad_norm": 0.7394677996635437,
"learning_rate": 3.225711693136156e-05,
"loss": 0.3286,
"step": 348
},
{
"epoch": 1.4016064257028114,
"grad_norm": 0.7352170944213867,
"learning_rate": 3.214519589465495e-05,
"loss": 0.2706,
"step": 349
},
{
"epoch": 1.4056224899598393,
"grad_norm": 0.7182479500770569,
"learning_rate": 3.203311869624107e-05,
"loss": 0.3704,
"step": 350
},
{
"epoch": 1.4096385542168675,
"grad_norm": 0.7015247344970703,
"learning_rate": 3.1920887785621235e-05,
"loss": 0.3309,
"step": 351
},
{
"epoch": 1.4136546184738956,
"grad_norm": 0.6726226806640625,
"learning_rate": 3.180850561565621e-05,
"loss": 0.2826,
"step": 352
},
{
"epoch": 1.4176706827309236,
"grad_norm": 0.7643373608589172,
"learning_rate": 3.1695974642512585e-05,
"loss": 0.3184,
"step": 353
},
{
"epoch": 1.4216867469879517,
"grad_norm": 0.851716935634613,
"learning_rate": 3.158329732560912e-05,
"loss": 0.3393,
"step": 354
},
{
"epoch": 1.4257028112449799,
"grad_norm": 0.8587582111358643,
"learning_rate": 3.147047612756302e-05,
"loss": 0.2996,
"step": 355
},
{
"epoch": 1.429718875502008,
"grad_norm": 0.8089635372161865,
"learning_rate": 3.1357513514136044e-05,
"loss": 0.3699,
"step": 356
},
{
"epoch": 1.4337349397590362,
"grad_norm": 0.7529250383377075,
"learning_rate": 3.1244411954180676e-05,
"loss": 0.3509,
"step": 357
},
{
"epoch": 1.4377510040160644,
"grad_norm": 0.7800310254096985,
"learning_rate": 3.113117391958612e-05,
"loss": 0.341,
"step": 358
},
{
"epoch": 1.4417670682730923,
"grad_norm": 0.6380038261413574,
"learning_rate": 3.101780188522433e-05,
"loss": 0.3674,
"step": 359
},
{
"epoch": 1.4457831325301205,
"grad_norm": 0.827573835849762,
"learning_rate": 3.090429832889586e-05,
"loss": 0.3459,
"step": 360
},
{
"epoch": 1.4497991967871486,
"grad_norm": 0.8122319579124451,
"learning_rate": 3.0790665731275765e-05,
"loss": 0.3204,
"step": 361
},
{
"epoch": 1.4538152610441766,
"grad_norm": 0.8184553384780884,
"learning_rate": 3.0676906575859334e-05,
"loss": 0.3107,
"step": 362
},
{
"epoch": 1.4578313253012047,
"grad_norm": 0.8002389073371887,
"learning_rate": 3.056302334890786e-05,
"loss": 0.3823,
"step": 363
},
{
"epoch": 1.461847389558233,
"grad_norm": 0.6957512497901917,
"learning_rate": 3.0449018539394276e-05,
"loss": 0.3028,
"step": 364
},
{
"epoch": 1.465863453815261,
"grad_norm": 0.7060897350311279,
"learning_rate": 3.0334894638948753e-05,
"loss": 0.356,
"step": 365
},
{
"epoch": 1.4698795180722892,
"grad_norm": 0.6910706162452698,
"learning_rate": 3.022065414180425e-05,
"loss": 0.3894,
"step": 366
},
{
"epoch": 1.4738955823293174,
"grad_norm": 0.634735643863678,
"learning_rate": 3.010629954474201e-05,
"loss": 0.3114,
"step": 367
},
{
"epoch": 1.4779116465863453,
"grad_norm": 0.7240673303604126,
"learning_rate": 2.9991833347036985e-05,
"loss": 0.3103,
"step": 368
},
{
"epoch": 1.4819277108433735,
"grad_norm": 0.7436959743499756,
"learning_rate": 2.9877258050403212e-05,
"loss": 0.3444,
"step": 369
},
{
"epoch": 1.4859437751004017,
"grad_norm": 0.736075758934021,
"learning_rate": 2.9762576158939125e-05,
"loss": 0.3869,
"step": 370
},
{
"epoch": 1.4899598393574296,
"grad_norm": 0.7735042572021484,
"learning_rate": 2.9647790179072872e-05,
"loss": 0.3084,
"step": 371
},
{
"epoch": 1.4939759036144578,
"grad_norm": 0.7524950504302979,
"learning_rate": 2.9532902619507462e-05,
"loss": 0.35,
"step": 372
},
{
"epoch": 1.497991967871486,
"grad_norm": 0.6790709495544434,
"learning_rate": 2.9417915991166008e-05,
"loss": 0.3098,
"step": 373
},
{
"epoch": 1.502008032128514,
"grad_norm": 0.6772611141204834,
"learning_rate": 2.93028328071368e-05,
"loss": 0.294,
"step": 374
},
{
"epoch": 1.5060240963855422,
"grad_norm": 0.6951244473457336,
"learning_rate": 2.918765558261841e-05,
"loss": 0.2844,
"step": 375
},
{
"epoch": 1.5100401606425704,
"grad_norm": 0.6909642219543457,
"learning_rate": 2.9072386834864724e-05,
"loss": 0.3082,
"step": 376
},
{
"epoch": 1.5140562248995983,
"grad_norm": 0.6457985639572144,
"learning_rate": 2.8957029083129873e-05,
"loss": 0.2628,
"step": 377
},
{
"epoch": 1.5180722891566265,
"grad_norm": 0.7581838369369507,
"learning_rate": 2.884158484861325e-05,
"loss": 0.3804,
"step": 378
},
{
"epoch": 1.5220883534136547,
"grad_norm": 0.7615995407104492,
"learning_rate": 2.872605665440436e-05,
"loss": 0.3667,
"step": 379
},
{
"epoch": 1.5261044176706826,
"grad_norm": 0.7376864552497864,
"learning_rate": 2.8610447025427683e-05,
"loss": 0.302,
"step": 380
},
{
"epoch": 1.5301204819277108,
"grad_norm": 0.6898524165153503,
"learning_rate": 2.849475848838749e-05,
"loss": 0.2746,
"step": 381
},
{
"epoch": 1.534136546184739,
"grad_norm": 0.6715502142906189,
"learning_rate": 2.8378993571712638e-05,
"loss": 0.3599,
"step": 382
},
{
"epoch": 1.538152610441767,
"grad_norm": 0.7485578060150146,
"learning_rate": 2.8263154805501297e-05,
"loss": 0.3692,
"step": 383
},
{
"epoch": 1.5421686746987953,
"grad_norm": 0.6707048416137695,
"learning_rate": 2.8147244721465636e-05,
"loss": 0.2836,
"step": 384
},
{
"epoch": 1.5461847389558234,
"grad_norm": 0.7567557096481323,
"learning_rate": 2.8031265852876537e-05,
"loss": 0.3734,
"step": 385
},
{
"epoch": 1.5502008032128514,
"grad_norm": 0.7292672991752625,
"learning_rate": 2.791522073450819e-05,
"loss": 0.3703,
"step": 386
},
{
"epoch": 1.5542168674698795,
"grad_norm": 0.7452452778816223,
"learning_rate": 2.7799111902582696e-05,
"loss": 0.3482,
"step": 387
},
{
"epoch": 1.5582329317269075,
"grad_norm": 0.7260645627975464,
"learning_rate": 2.7682941894714666e-05,
"loss": 0.345,
"step": 388
},
{
"epoch": 1.5622489959839356,
"grad_norm": 0.7072299718856812,
"learning_rate": 2.7566713249855712e-05,
"loss": 0.3538,
"step": 389
},
{
"epoch": 1.5662650602409638,
"grad_norm": 0.7503701448440552,
"learning_rate": 2.7450428508239024e-05,
"loss": 0.3245,
"step": 390
},
{
"epoch": 1.570281124497992,
"grad_norm": 0.671421468257904,
"learning_rate": 2.7334090211323766e-05,
"loss": 0.3616,
"step": 391
},
{
"epoch": 1.5742971887550201,
"grad_norm": 0.7954780459403992,
"learning_rate": 2.7217700901739636e-05,
"loss": 0.3726,
"step": 392
},
{
"epoch": 1.5783132530120483,
"grad_norm": 0.7115846872329712,
"learning_rate": 2.710126312323119e-05,
"loss": 0.3879,
"step": 393
},
{
"epoch": 1.5823293172690764,
"grad_norm": 0.8727709650993347,
"learning_rate": 2.6984779420602323e-05,
"loss": 0.374,
"step": 394
},
{
"epoch": 1.5863453815261044,
"grad_norm": 0.6634027361869812,
"learning_rate": 2.686825233966061e-05,
"loss": 0.272,
"step": 395
},
{
"epoch": 1.5903614457831325,
"grad_norm": 0.7339637279510498,
"learning_rate": 2.6751684427161683e-05,
"loss": 0.3792,
"step": 396
},
{
"epoch": 1.5943775100401605,
"grad_norm": 0.6629202961921692,
"learning_rate": 2.663507823075358e-05,
"loss": 0.3044,
"step": 397
},
{
"epoch": 1.5983935742971886,
"grad_norm": 0.7362025380134583,
"learning_rate": 2.651843629892103e-05,
"loss": 0.3503,
"step": 398
},
{
"epoch": 1.6024096385542168,
"grad_norm": 0.7173884510993958,
"learning_rate": 2.6401761180929797e-05,
"loss": 0.2917,
"step": 399
},
{
"epoch": 1.606425702811245,
"grad_norm": 0.7019863128662109,
"learning_rate": 2.6285055426770934e-05,
"loss": 0.3217,
"step": 400
},
{
"epoch": 1.6104417670682731,
"grad_norm": 0.668793261051178,
"learning_rate": 2.616832158710506e-05,
"loss": 0.3452,
"step": 401
},
{
"epoch": 1.6144578313253013,
"grad_norm": 0.6839523315429688,
"learning_rate": 2.6051562213206632e-05,
"loss": 0.2853,
"step": 402
},
{
"epoch": 1.6184738955823295,
"grad_norm": 0.7479641437530518,
"learning_rate": 2.593477985690815e-05,
"loss": 0.4029,
"step": 403
},
{
"epoch": 1.6224899598393574,
"grad_norm": 0.7546053528785706,
"learning_rate": 2.5817977070544407e-05,
"loss": 0.3556,
"step": 404
},
{
"epoch": 1.6265060240963856,
"grad_norm": 0.8448750972747803,
"learning_rate": 2.5701156406896725e-05,
"loss": 0.399,
"step": 405
},
{
"epoch": 1.6305220883534135,
"grad_norm": 0.77425616979599,
"learning_rate": 2.5584320419137127e-05,
"loss": 0.301,
"step": 406
},
{
"epoch": 1.6345381526104417,
"grad_norm": 0.9346187114715576,
"learning_rate": 2.5467471660772557e-05,
"loss": 0.3723,
"step": 407
},
{
"epoch": 1.6385542168674698,
"grad_norm": 0.768231213092804,
"learning_rate": 2.535061268558906e-05,
"loss": 0.3502,
"step": 408
},
{
"epoch": 1.642570281124498,
"grad_norm": 0.673270583152771,
"learning_rate": 2.5233746047595985e-05,
"loss": 0.3066,
"step": 409
},
{
"epoch": 1.6465863453815262,
"grad_norm": 0.6739712357521057,
"learning_rate": 2.5116874300970138e-05,
"loss": 0.3378,
"step": 410
},
{
"epoch": 1.6506024096385543,
"grad_norm": 0.6984811425209045,
"learning_rate": 2.5e-05,
"loss": 0.3083,
"step": 411
},
{
"epoch": 1.6546184738955825,
"grad_norm": 0.6985664963722229,
"learning_rate": 2.4883125699029865e-05,
"loss": 0.3625,
"step": 412
},
{
"epoch": 1.6586345381526104,
"grad_norm": 0.6901569962501526,
"learning_rate": 2.4766253952404024e-05,
"loss": 0.3392,
"step": 413
},
{
"epoch": 1.6626506024096386,
"grad_norm": 0.7855417132377625,
"learning_rate": 2.4649387314410945e-05,
"loss": 0.3255,
"step": 414
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.8108392953872681,
"learning_rate": 2.4532528339227452e-05,
"loss": 0.3242,
"step": 415
},
{
"epoch": 1.6706827309236947,
"grad_norm": 0.8146324753761292,
"learning_rate": 2.441567958086288e-05,
"loss": 0.3173,
"step": 416
},
{
"epoch": 1.6746987951807228,
"grad_norm": 0.7759451270103455,
"learning_rate": 2.429884359310328e-05,
"loss": 0.3513,
"step": 417
},
{
"epoch": 1.678714859437751,
"grad_norm": 0.7257838845252991,
"learning_rate": 2.41820229294556e-05,
"loss": 0.3018,
"step": 418
},
{
"epoch": 1.6827309236947792,
"grad_norm": 0.7915422916412354,
"learning_rate": 2.4065220143091864e-05,
"loss": 0.39,
"step": 419
},
{
"epoch": 1.6867469879518073,
"grad_norm": 0.6713058352470398,
"learning_rate": 2.3948437786793377e-05,
"loss": 0.3026,
"step": 420
},
{
"epoch": 1.6907630522088355,
"grad_norm": 0.6408040523529053,
"learning_rate": 2.383167841289495e-05,
"loss": 0.2606,
"step": 421
},
{
"epoch": 1.6947791164658634,
"grad_norm": 0.7138453125953674,
"learning_rate": 2.371494457322907e-05,
"loss": 0.3099,
"step": 422
},
{
"epoch": 1.6987951807228916,
"grad_norm": 0.6884890198707581,
"learning_rate": 2.3598238819070202e-05,
"loss": 0.2958,
"step": 423
},
{
"epoch": 1.7028112449799195,
"grad_norm": 0.7783042788505554,
"learning_rate": 2.3481563701078966e-05,
"loss": 0.3311,
"step": 424
},
{
"epoch": 1.7068273092369477,
"grad_norm": 0.7458612322807312,
"learning_rate": 2.3364921769246423e-05,
"loss": 0.3376,
"step": 425
},
{
"epoch": 1.7108433734939759,
"grad_norm": 0.8598488569259644,
"learning_rate": 2.3248315572838316e-05,
"loss": 0.3199,
"step": 426
},
{
"epoch": 1.714859437751004,
"grad_norm": 0.7067093253135681,
"learning_rate": 2.3131747660339394e-05,
"loss": 0.3161,
"step": 427
},
{
"epoch": 1.7188755020080322,
"grad_norm": 0.7883020639419556,
"learning_rate": 2.3015220579397682e-05,
"loss": 0.3633,
"step": 428
},
{
"epoch": 1.7228915662650603,
"grad_norm": 0.7497788667678833,
"learning_rate": 2.2898736876768815e-05,
"loss": 0.3441,
"step": 429
},
{
"epoch": 1.7269076305220885,
"grad_norm": 0.6594833135604858,
"learning_rate": 2.278229909826037e-05,
"loss": 0.2846,
"step": 430
},
{
"epoch": 1.7309236947791165,
"grad_norm": 0.6981406807899475,
"learning_rate": 2.2665909788676237e-05,
"loss": 0.336,
"step": 431
},
{
"epoch": 1.7349397590361446,
"grad_norm": 0.7191076874732971,
"learning_rate": 2.2549571491760986e-05,
"loss": 0.3304,
"step": 432
},
{
"epoch": 1.7389558232931726,
"grad_norm": 0.8120121359825134,
"learning_rate": 2.2433286750144293e-05,
"loss": 0.3102,
"step": 433
},
{
"epoch": 1.7429718875502007,
"grad_norm": 0.7788616418838501,
"learning_rate": 2.231705810528534e-05,
"loss": 0.3253,
"step": 434
},
{
"epoch": 1.7469879518072289,
"grad_norm": 0.6789368391036987,
"learning_rate": 2.2200888097417307e-05,
"loss": 0.3341,
"step": 435
},
{
"epoch": 1.751004016064257,
"grad_norm": 0.730965256690979,
"learning_rate": 2.2084779265491813e-05,
"loss": 0.3263,
"step": 436
},
{
"epoch": 1.7550200803212852,
"grad_norm": 0.7907600402832031,
"learning_rate": 2.1968734147123465e-05,
"loss": 0.3053,
"step": 437
},
{
"epoch": 1.7590361445783134,
"grad_norm": 0.7607869505882263,
"learning_rate": 2.185275527853437e-05,
"loss": 0.3683,
"step": 438
},
{
"epoch": 1.7630522088353415,
"grad_norm": 0.7206976413726807,
"learning_rate": 2.173684519449872e-05,
"loss": 0.3087,
"step": 439
},
{
"epoch": 1.7670682730923695,
"grad_norm": 0.7315901517868042,
"learning_rate": 2.162100642828737e-05,
"loss": 0.2987,
"step": 440
},
{
"epoch": 1.7710843373493976,
"grad_norm": 0.7159738540649414,
"learning_rate": 2.1505241511612522e-05,
"loss": 0.2849,
"step": 441
},
{
"epoch": 1.7751004016064256,
"grad_norm": 0.7728487849235535,
"learning_rate": 2.138955297457233e-05,
"loss": 0.3888,
"step": 442
},
{
"epoch": 1.7791164658634537,
"grad_norm": 0.829146683216095,
"learning_rate": 2.1273943345595637e-05,
"loss": 0.333,
"step": 443
},
{
"epoch": 1.783132530120482,
"grad_norm": 0.7567833662033081,
"learning_rate": 2.1158415151386744e-05,
"loss": 0.3216,
"step": 444
},
{
"epoch": 1.78714859437751,
"grad_norm": 0.676815390586853,
"learning_rate": 2.104297091687013e-05,
"loss": 0.2618,
"step": 445
},
{
"epoch": 1.7911646586345382,
"grad_norm": 0.8237574100494385,
"learning_rate": 2.0927613165135285e-05,
"loss": 0.3332,
"step": 446
},
{
"epoch": 1.7951807228915664,
"grad_norm": 0.8036498427391052,
"learning_rate": 2.0812344417381595e-05,
"loss": 0.3136,
"step": 447
},
{
"epoch": 1.7991967871485943,
"grad_norm": 0.8381817936897278,
"learning_rate": 2.0697167192863204e-05,
"loss": 0.3241,
"step": 448
},
{
"epoch": 1.8032128514056225,
"grad_norm": 0.8362058401107788,
"learning_rate": 2.0582084008834e-05,
"loss": 0.2887,
"step": 449
},
{
"epoch": 1.8072289156626506,
"grad_norm": 0.7610471248626709,
"learning_rate": 2.0467097380492544e-05,
"loss": 0.2986,
"step": 450
},
{
"epoch": 1.8112449799196786,
"grad_norm": 0.7676022052764893,
"learning_rate": 2.0352209820927137e-05,
"loss": 0.3148,
"step": 451
},
{
"epoch": 1.8152610441767068,
"grad_norm": 0.7500276565551758,
"learning_rate": 2.0237423841060877e-05,
"loss": 0.2663,
"step": 452
},
{
"epoch": 1.819277108433735,
"grad_norm": 0.7033796906471252,
"learning_rate": 2.0122741949596797e-05,
"loss": 0.3025,
"step": 453
},
{
"epoch": 1.823293172690763,
"grad_norm": 0.7410452365875244,
"learning_rate": 2.000816665296302e-05,
"loss": 0.3209,
"step": 454
},
{
"epoch": 1.8273092369477912,
"grad_norm": 0.7933539748191833,
"learning_rate": 1.9893700455257996e-05,
"loss": 0.329,
"step": 455
},
{
"epoch": 1.8313253012048194,
"grad_norm": 0.7989787459373474,
"learning_rate": 1.977934585819576e-05,
"loss": 0.3232,
"step": 456
},
{
"epoch": 1.8353413654618473,
"grad_norm": 0.7234666347503662,
"learning_rate": 1.9665105361051256e-05,
"loss": 0.3061,
"step": 457
},
{
"epoch": 1.8393574297188755,
"grad_norm": 0.7008742094039917,
"learning_rate": 1.9550981460605733e-05,
"loss": 0.2821,
"step": 458
},
{
"epoch": 1.8433734939759037,
"grad_norm": 0.7315691709518433,
"learning_rate": 1.9436976651092144e-05,
"loss": 0.3464,
"step": 459
},
{
"epoch": 1.8473895582329316,
"grad_norm": 0.7527895569801331,
"learning_rate": 1.932309342414067e-05,
"loss": 0.3522,
"step": 460
},
{
"epoch": 1.8514056224899598,
"grad_norm": 0.6792200207710266,
"learning_rate": 1.9209334268724247e-05,
"loss": 0.252,
"step": 461
},
{
"epoch": 1.855421686746988,
"grad_norm": 0.7846032977104187,
"learning_rate": 1.909570167110415e-05,
"loss": 0.3317,
"step": 462
},
{
"epoch": 1.859437751004016,
"grad_norm": 0.7925510406494141,
"learning_rate": 1.8982198114775682e-05,
"loss": 0.3658,
"step": 463
},
{
"epoch": 1.8634538152610443,
"grad_norm": 0.7116525173187256,
"learning_rate": 1.8868826080413875e-05,
"loss": 0.3278,
"step": 464
},
{
"epoch": 1.8674698795180724,
"grad_norm": 0.8022361993789673,
"learning_rate": 1.8755588045819327e-05,
"loss": 0.3357,
"step": 465
},
{
"epoch": 1.8714859437751004,
"grad_norm": 0.7105912566184998,
"learning_rate": 1.864248648586395e-05,
"loss": 0.3216,
"step": 466
},
{
"epoch": 1.8755020080321285,
"grad_norm": 0.7227326035499573,
"learning_rate": 1.852952387243698e-05,
"loss": 0.2626,
"step": 467
},
{
"epoch": 1.8795180722891565,
"grad_norm": 0.7557914853096008,
"learning_rate": 1.841670267439088e-05,
"loss": 0.3039,
"step": 468
},
{
"epoch": 1.8835341365461846,
"grad_norm": 0.719526469707489,
"learning_rate": 1.8304025357487424e-05,
"loss": 0.3395,
"step": 469
},
{
"epoch": 1.8875502008032128,
"grad_norm": 0.7930237650871277,
"learning_rate": 1.81914943843438e-05,
"loss": 0.3219,
"step": 470
},
{
"epoch": 1.891566265060241,
"grad_norm": 0.7459167838096619,
"learning_rate": 1.8079112214378768e-05,
"loss": 0.3521,
"step": 471
},
{
"epoch": 1.895582329317269,
"grad_norm": 0.6873487830162048,
"learning_rate": 1.7966881303758938e-05,
"loss": 0.2815,
"step": 472
},
{
"epoch": 1.8995983935742973,
"grad_norm": 0.8124189972877502,
"learning_rate": 1.7854804105345062e-05,
"loss": 0.29,
"step": 473
},
{
"epoch": 1.9036144578313254,
"grad_norm": 0.8539310097694397,
"learning_rate": 1.7742883068638447e-05,
"loss": 0.323,
"step": 474
},
{
"epoch": 1.9076305220883534,
"grad_norm": 0.7768955230712891,
"learning_rate": 1.7631120639727393e-05,
"loss": 0.3408,
"step": 475
},
{
"epoch": 1.9116465863453815,
"grad_norm": 0.7574753165245056,
"learning_rate": 1.7519519261233786e-05,
"loss": 0.3565,
"step": 476
},
{
"epoch": 1.9156626506024095,
"grad_norm": 0.7839083671569824,
"learning_rate": 1.7408081372259632e-05,
"loss": 0.3402,
"step": 477
},
{
"epoch": 1.9196787148594376,
"grad_norm": 0.826919674873352,
"learning_rate": 1.729680940833383e-05,
"loss": 0.3277,
"step": 478
},
{
"epoch": 1.9236947791164658,
"grad_norm": 0.6887242197990417,
"learning_rate": 1.7185705801358892e-05,
"loss": 0.28,
"step": 479
},
{
"epoch": 1.927710843373494,
"grad_norm": 0.8087524771690369,
"learning_rate": 1.7074772979557802e-05,
"loss": 0.3387,
"step": 480
},
{
"epoch": 1.9317269076305221,
"grad_norm": 0.775202751159668,
"learning_rate": 1.6964013367420966e-05,
"loss": 0.2977,
"step": 481
},
{
"epoch": 1.9357429718875503,
"grad_norm": 0.8086408376693726,
"learning_rate": 1.6853429385653196e-05,
"loss": 0.3207,
"step": 482
},
{
"epoch": 1.9397590361445785,
"grad_norm": 0.8384583592414856,
"learning_rate": 1.6743023451120832e-05,
"loss": 0.3649,
"step": 483
},
{
"epoch": 1.9437751004016064,
"grad_norm": 0.8741695284843445,
"learning_rate": 1.6632797976798886e-05,
"loss": 0.2905,
"step": 484
},
{
"epoch": 1.9477911646586346,
"grad_norm": 0.7894666194915771,
"learning_rate": 1.6522755371718333e-05,
"loss": 0.2825,
"step": 485
},
{
"epoch": 1.9518072289156625,
"grad_norm": 0.8331698179244995,
"learning_rate": 1.641289804091347e-05,
"loss": 0.2996,
"step": 486
},
{
"epoch": 1.9558232931726907,
"grad_norm": 0.8386313915252686,
"learning_rate": 1.6303228385369318e-05,
"loss": 0.397,
"step": 487
},
{
"epoch": 1.9598393574297188,
"grad_norm": 0.8523121476173401,
"learning_rate": 1.6193748801969163e-05,
"loss": 0.3357,
"step": 488
},
{
"epoch": 1.963855421686747,
"grad_norm": 0.7356472015380859,
"learning_rate": 1.6084461683442176e-05,
"loss": 0.3044,
"step": 489
},
{
"epoch": 1.9678714859437751,
"grad_norm": 0.7336781620979309,
"learning_rate": 1.5975369418311112e-05,
"loss": 0.3055,
"step": 490
},
{
"epoch": 1.9718875502008033,
"grad_norm": 0.8148435354232788,
"learning_rate": 1.5866474390840125e-05,
"loss": 0.385,
"step": 491
},
{
"epoch": 1.9759036144578315,
"grad_norm": 0.7665354609489441,
"learning_rate": 1.5757778980982626e-05,
"loss": 0.2889,
"step": 492
},
{
"epoch": 1.9799196787148594,
"grad_norm": 0.7822697162628174,
"learning_rate": 1.5649285564329296e-05,
"loss": 0.3019,
"step": 493
},
{
"epoch": 1.9839357429718876,
"grad_norm": 0.8199531435966492,
"learning_rate": 1.5540996512056165e-05,
"loss": 0.36,
"step": 494
},
{
"epoch": 1.9879518072289155,
"grad_norm": 0.7746401429176331,
"learning_rate": 1.5432914190872757e-05,
"loss": 0.3167,
"step": 495
},
{
"epoch": 1.9919678714859437,
"grad_norm": 0.8390997052192688,
"learning_rate": 1.5325040962970415e-05,
"loss": 0.3859,
"step": 496
},
{
"epoch": 1.9959839357429718,
"grad_norm": 0.8079429864883423,
"learning_rate": 1.5217379185970629e-05,
"loss": 0.3322,
"step": 497
},
{
"epoch": 2.0,
"grad_norm": 0.7987474203109741,
"learning_rate": 1.5109931212873534e-05,
"loss": 0.2731,
"step": 498
},
{
"epoch": 2.004016064257028,
"grad_norm": 0.7697616219520569,
"learning_rate": 1.500269939200648e-05,
"loss": 0.3519,
"step": 499
},
{
"epoch": 2.0080321285140563,
"grad_norm": 0.7250295281410217,
"learning_rate": 1.4895686066972703e-05,
"loss": 0.2922,
"step": 500
},
{
"epoch": 2.0120481927710845,
"grad_norm": 0.6913548111915588,
"learning_rate": 1.4788893576600099e-05,
"loss": 0.2943,
"step": 501
},
{
"epoch": 2.0160642570281126,
"grad_norm": 0.7827958464622498,
"learning_rate": 1.4682324254890134e-05,
"loss": 0.3059,
"step": 502
},
{
"epoch": 2.0200803212851404,
"grad_norm": 0.8085004091262817,
"learning_rate": 1.4575980430966807e-05,
"loss": 0.3039,
"step": 503
},
{
"epoch": 2.0240963855421685,
"grad_norm": 0.8477670550346375,
"learning_rate": 1.446986442902574e-05,
"loss": 0.2839,
"step": 504
},
{
"epoch": 2.0281124497991967,
"grad_norm": 0.7724645137786865,
"learning_rate": 1.4363978568283412e-05,
"loss": 0.2767,
"step": 505
},
{
"epoch": 2.032128514056225,
"grad_norm": 0.6548305749893188,
"learning_rate": 1.4258325162926442e-05,
"loss": 0.2194,
"step": 506
},
{
"epoch": 2.036144578313253,
"grad_norm": 0.7934015393257141,
"learning_rate": 1.4152906522061048e-05,
"loss": 0.312,
"step": 507
},
{
"epoch": 2.040160642570281,
"grad_norm": 0.8331040143966675,
"learning_rate": 1.4047724949662539e-05,
"loss": 0.3619,
"step": 508
},
{
"epoch": 2.0441767068273093,
"grad_norm": 0.8048452138900757,
"learning_rate": 1.3942782744524973e-05,
"loss": 0.3362,
"step": 509
},
{
"epoch": 2.0481927710843375,
"grad_norm": 0.7336853742599487,
"learning_rate": 1.3838082200210931e-05,
"loss": 0.2897,
"step": 510
},
{
"epoch": 2.0522088353413657,
"grad_norm": 0.7978460192680359,
"learning_rate": 1.3733625605001365e-05,
"loss": 0.2325,
"step": 511
},
{
"epoch": 2.0562248995983934,
"grad_norm": 0.7312659025192261,
"learning_rate": 1.362941524184564e-05,
"loss": 0.2831,
"step": 512
},
{
"epoch": 2.0602409638554215,
"grad_norm": 0.8522686958312988,
"learning_rate": 1.3525453388311554e-05,
"loss": 0.2837,
"step": 513
},
{
"epoch": 2.0642570281124497,
"grad_norm": 0.8868270516395569,
"learning_rate": 1.3421742316535637e-05,
"loss": 0.2896,
"step": 514
},
{
"epoch": 2.068273092369478,
"grad_norm": 0.8070749044418335,
"learning_rate": 1.331828429317345e-05,
"loss": 0.271,
"step": 515
},
{
"epoch": 2.072289156626506,
"grad_norm": 0.7994053363800049,
"learning_rate": 1.3215081579350058e-05,
"loss": 0.2622,
"step": 516
},
{
"epoch": 2.076305220883534,
"grad_norm": 0.7788709402084351,
"learning_rate": 1.3112136430610621e-05,
"loss": 0.3015,
"step": 517
},
{
"epoch": 2.0803212851405624,
"grad_norm": 0.8416688442230225,
"learning_rate": 1.3009451096871083e-05,
"loss": 0.25,
"step": 518
},
{
"epoch": 2.0843373493975905,
"grad_norm": 1.1160537004470825,
"learning_rate": 1.2907027822369005e-05,
"loss": 0.268,
"step": 519
},
{
"epoch": 2.0883534136546187,
"grad_norm": 0.8505333065986633,
"learning_rate": 1.2804868845614525e-05,
"loss": 0.3321,
"step": 520
},
{
"epoch": 2.0923694779116464,
"grad_norm": 0.8287850022315979,
"learning_rate": 1.2702976399341421e-05,
"loss": 0.2448,
"step": 521
},
{
"epoch": 2.0963855421686746,
"grad_norm": 0.8349308967590332,
"learning_rate": 1.2601352710458313e-05,
"loss": 0.2749,
"step": 522
},
{
"epoch": 2.1004016064257027,
"grad_norm": 0.7752396464347839,
"learning_rate": 1.2500000000000006e-05,
"loss": 0.2392,
"step": 523
},
{
"epoch": 2.104417670682731,
"grad_norm": 0.7974147796630859,
"learning_rate": 1.2398920483078938e-05,
"loss": 0.255,
"step": 524
},
{
"epoch": 2.108433734939759,
"grad_norm": 0.8653110861778259,
"learning_rate": 1.229811636883677e-05,
"loss": 0.2972,
"step": 525
},
{
"epoch": 2.112449799196787,
"grad_norm": 0.820702075958252,
"learning_rate": 1.2197589860396108e-05,
"loss": 0.2817,
"step": 526
},
{
"epoch": 2.1164658634538154,
"grad_norm": 0.8137445449829102,
"learning_rate": 1.2097343154812332e-05,
"loss": 0.2726,
"step": 527
},
{
"epoch": 2.1204819277108435,
"grad_norm": 0.9908772706985474,
"learning_rate": 1.1997378443025634e-05,
"loss": 0.2772,
"step": 528
},
{
"epoch": 2.1244979919678713,
"grad_norm": 0.9932394027709961,
"learning_rate": 1.1897697909813057e-05,
"loss": 0.3138,
"step": 529
},
{
"epoch": 2.1285140562248994,
"grad_norm": 0.8923698663711548,
"learning_rate": 1.1798303733740802e-05,
"loss": 0.253,
"step": 530
},
{
"epoch": 2.1325301204819276,
"grad_norm": 0.9020155668258667,
"learning_rate": 1.1699198087116589e-05,
"loss": 0.2609,
"step": 531
},
{
"epoch": 2.1365461847389557,
"grad_norm": 0.9627537131309509,
"learning_rate": 1.1600383135942191e-05,
"loss": 0.2872,
"step": 532
},
{
"epoch": 2.140562248995984,
"grad_norm": 0.8959113955497742,
"learning_rate": 1.1501861039866094e-05,
"loss": 0.2812,
"step": 533
},
{
"epoch": 2.144578313253012,
"grad_norm": 0.9380473494529724,
"learning_rate": 1.140363395213629e-05,
"loss": 0.3253,
"step": 534
},
{
"epoch": 2.1485943775100402,
"grad_norm": 0.8573694825172424,
"learning_rate": 1.130570401955322e-05,
"loss": 0.2901,
"step": 535
},
{
"epoch": 2.1526104417670684,
"grad_norm": 0.9344317317008972,
"learning_rate": 1.1208073382422865e-05,
"loss": 0.3408,
"step": 536
},
{
"epoch": 2.1566265060240966,
"grad_norm": 0.8469679355621338,
"learning_rate": 1.1110744174509952e-05,
"loss": 0.2857,
"step": 537
},
{
"epoch": 2.1606425702811247,
"grad_norm": 0.9033740162849426,
"learning_rate": 1.1013718522991315e-05,
"loss": 0.3118,
"step": 538
},
{
"epoch": 2.1646586345381524,
"grad_norm": 0.817452073097229,
"learning_rate": 1.0916998548409449e-05,
"loss": 0.2363,
"step": 539
},
{
"epoch": 2.1686746987951806,
"grad_norm": 0.8782771825790405,
"learning_rate": 1.0820586364626104e-05,
"loss": 0.2205,
"step": 540
},
{
"epoch": 2.1726907630522088,
"grad_norm": 0.7995834946632385,
"learning_rate": 1.072448407877612e-05,
"loss": 0.2646,
"step": 541
},
{
"epoch": 2.176706827309237,
"grad_norm": 0.8504219055175781,
"learning_rate": 1.0628693791221372e-05,
"loss": 0.2827,
"step": 542
},
{
"epoch": 2.180722891566265,
"grad_norm": 0.9514392614364624,
"learning_rate": 1.0533217595504858e-05,
"loss": 0.2945,
"step": 543
},
{
"epoch": 2.1847389558232932,
"grad_norm": 0.9158867001533508,
"learning_rate": 1.043805757830495e-05,
"loss": 0.2777,
"step": 544
},
{
"epoch": 2.1887550200803214,
"grad_norm": 0.825770914554596,
"learning_rate": 1.0343215819389782e-05,
"loss": 0.2978,
"step": 545
},
{
"epoch": 2.1927710843373496,
"grad_norm": 1.0489541292190552,
"learning_rate": 1.0248694391571801e-05,
"loss": 0.295,
"step": 546
},
{
"epoch": 2.1967871485943773,
"grad_norm": 0.8876386880874634,
"learning_rate": 1.0154495360662464e-05,
"loss": 0.2816,
"step": 547
},
{
"epoch": 2.2008032128514055,
"grad_norm": 1.0437474250793457,
"learning_rate": 1.0060620785427084e-05,
"loss": 0.2969,
"step": 548
},
{
"epoch": 2.2048192771084336,
"grad_norm": 0.9109717607498169,
"learning_rate": 9.967072717539851e-06,
"loss": 0.2674,
"step": 549
},
{
"epoch": 2.208835341365462,
"grad_norm": 0.9593303799629211,
"learning_rate": 9.873853201538971e-06,
"loss": 0.3017,
"step": 550
},
{
"epoch": 2.21285140562249,
"grad_norm": 0.9704967737197876,
"learning_rate": 9.780964274781984e-06,
"loss": 0.3028,
"step": 551
},
{
"epoch": 2.216867469879518,
"grad_norm": 1.0326952934265137,
"learning_rate": 9.688407967401248e-06,
"loss": 0.2795,
"step": 552
},
{
"epoch": 2.2208835341365463,
"grad_norm": 0.9703342914581299,
"learning_rate": 9.596186302259563e-06,
"loss": 0.3243,
"step": 553
},
{
"epoch": 2.2248995983935744,
"grad_norm": 0.8894467949867249,
"learning_rate": 9.504301294905965e-06,
"loss": 0.2599,
"step": 554
},
{
"epoch": 2.2289156626506026,
"grad_norm": 0.8972454071044922,
"learning_rate": 9.412754953531663e-06,
"loss": 0.2648,
"step": 555
},
{
"epoch": 2.2329317269076308,
"grad_norm": 0.9155896306037903,
"learning_rate": 9.32154927892617e-06,
"loss": 0.2739,
"step": 556
},
{
"epoch": 2.2369477911646585,
"grad_norm": 0.8431274890899658,
"learning_rate": 9.230686264433547e-06,
"loss": 0.2307,
"step": 557
},
{
"epoch": 2.2409638554216866,
"grad_norm": 0.9339072108268738,
"learning_rate": 9.140167895908867e-06,
"loss": 0.2638,
"step": 558
},
{
"epoch": 2.244979919678715,
"grad_norm": 0.8334931135177612,
"learning_rate": 9.049996151674789e-06,
"loss": 0.2273,
"step": 559
},
{
"epoch": 2.248995983935743,
"grad_norm": 0.9012922048568726,
"learning_rate": 8.960173002478336e-06,
"loss": 0.3319,
"step": 560
},
{
"epoch": 2.253012048192771,
"grad_norm": 0.8430778980255127,
"learning_rate": 8.870700411447816e-06,
"loss": 0.3088,
"step": 561
},
{
"epoch": 2.2570281124497993,
"grad_norm": 0.9216142892837524,
"learning_rate": 8.781580334049919e-06,
"loss": 0.3163,
"step": 562
},
{
"epoch": 2.2610441767068274,
"grad_norm": 0.9573366641998291,
"learning_rate": 8.69281471804698e-06,
"loss": 0.3146,
"step": 563
},
{
"epoch": 2.2650602409638556,
"grad_norm": 0.8481350541114807,
"learning_rate": 8.604405503454399e-06,
"loss": 0.2425,
"step": 564
},
{
"epoch": 2.2690763052208833,
"grad_norm": 0.9956733584403992,
"learning_rate": 8.51635462249828e-06,
"loss": 0.3545,
"step": 565
},
{
"epoch": 2.2730923694779115,
"grad_norm": 0.9506481885910034,
"learning_rate": 8.428663999573142e-06,
"loss": 0.3288,
"step": 566
},
{
"epoch": 2.2771084337349397,
"grad_norm": 0.8302280306816101,
"learning_rate": 8.341335551199902e-06,
"loss": 0.2435,
"step": 567
},
{
"epoch": 2.281124497991968,
"grad_norm": 0.9323450922966003,
"learning_rate": 8.25437118598398e-06,
"loss": 0.2971,
"step": 568
},
{
"epoch": 2.285140562248996,
"grad_norm": 0.9358099699020386,
"learning_rate": 8.167772804573571e-06,
"loss": 0.3021,
"step": 569
},
{
"epoch": 2.289156626506024,
"grad_norm": 0.9353638291358948,
"learning_rate": 8.081542299618139e-06,
"loss": 0.2378,
"step": 570
},
{
"epoch": 2.2931726907630523,
"grad_norm": 0.8503626585006714,
"learning_rate": 7.99568155572701e-06,
"loss": 0.2587,
"step": 571
},
{
"epoch": 2.2971887550200805,
"grad_norm": 0.9960978627204895,
"learning_rate": 7.910192449428217e-06,
"loss": 0.291,
"step": 572
},
{
"epoch": 2.3012048192771086,
"grad_norm": 1.006942629814148,
"learning_rate": 7.825076849127458e-06,
"loss": 0.2936,
"step": 573
},
{
"epoch": 2.305220883534137,
"grad_norm": 0.9417559504508972,
"learning_rate": 7.740336615067292e-06,
"loss": 0.3128,
"step": 574
},
{
"epoch": 2.3092369477911645,
"grad_norm": 0.8731968998908997,
"learning_rate": 7.65597359928646e-06,
"loss": 0.2541,
"step": 575
},
{
"epoch": 2.3132530120481927,
"grad_norm": 1.1042046546936035,
"learning_rate": 7.571989645579419e-06,
"loss": 0.3464,
"step": 576
},
{
"epoch": 2.317269076305221,
"grad_norm": 0.9111700057983398,
"learning_rate": 7.488386589456042e-06,
"loss": 0.2939,
"step": 577
},
{
"epoch": 2.321285140562249,
"grad_norm": 1.125536322593689,
"learning_rate": 7.405166258101495e-06,
"loss": 0.2891,
"step": 578
},
{
"epoch": 2.325301204819277,
"grad_norm": 1.0490670204162598,
"learning_rate": 7.3223304703363135e-06,
"loss": 0.3382,
"step": 579
},
{
"epoch": 2.3293172690763053,
"grad_norm": 1.0409460067749023,
"learning_rate": 7.239881036576651e-06,
"loss": 0.2961,
"step": 580
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.9009960889816284,
"learning_rate": 7.157819758794698e-06,
"loss": 0.2534,
"step": 581
},
{
"epoch": 2.337349397590361,
"grad_norm": 0.9325215220451355,
"learning_rate": 7.076148430479321e-06,
"loss": 0.2809,
"step": 582
},
{
"epoch": 2.3413654618473894,
"grad_norm": 0.8957793712615967,
"learning_rate": 6.99486883659684e-06,
"loss": 0.231,
"step": 583
},
{
"epoch": 2.3453815261044175,
"grad_norm": 0.9242108464241028,
"learning_rate": 6.91398275355204e-06,
"loss": 0.2859,
"step": 584
},
{
"epoch": 2.3493975903614457,
"grad_norm": 0.9579724073410034,
"learning_rate": 6.833491949149329e-06,
"loss": 0.2741,
"step": 585
},
{
"epoch": 2.353413654618474,
"grad_norm": 0.9978570938110352,
"learning_rate": 6.753398182554116e-06,
"loss": 0.3263,
"step": 586
},
{
"epoch": 2.357429718875502,
"grad_norm": 0.9748632907867432,
"learning_rate": 6.673703204254347e-06,
"loss": 0.289,
"step": 587
},
{
"epoch": 2.36144578313253,
"grad_norm": 0.9801468253135681,
"learning_rate": 6.594408756022272e-06,
"loss": 0.2379,
"step": 588
},
{
"epoch": 2.3654618473895583,
"grad_norm": 0.9527619481086731,
"learning_rate": 6.515516570876351e-06,
"loss": 0.3387,
"step": 589
},
{
"epoch": 2.3694779116465865,
"grad_norm": 0.9124782681465149,
"learning_rate": 6.437028373043386e-06,
"loss": 0.2344,
"step": 590
},
{
"epoch": 2.3734939759036147,
"grad_norm": 0.9202039837837219,
"learning_rate": 6.358945877920861e-06,
"loss": 0.2787,
"step": 591
},
{
"epoch": 2.3775100401606424,
"grad_norm": 1.1043528318405151,
"learning_rate": 6.281270792039418e-06,
"loss": 0.2872,
"step": 592
},
{
"epoch": 2.3815261044176705,
"grad_norm": 0.981726884841919,
"learning_rate": 6.204004813025568e-06,
"loss": 0.3403,
"step": 593
},
{
"epoch": 2.3855421686746987,
"grad_norm": 0.997560441493988,
"learning_rate": 6.127149629564605e-06,
"loss": 0.2656,
"step": 594
},
{
"epoch": 2.389558232931727,
"grad_norm": 1.06382155418396,
"learning_rate": 6.050706921363672e-06,
"loss": 0.313,
"step": 595
},
{
"epoch": 2.393574297188755,
"grad_norm": 0.9605445861816406,
"learning_rate": 5.974678359115094e-06,
"loss": 0.2862,
"step": 596
},
{
"epoch": 2.397590361445783,
"grad_norm": 1.005161166191101,
"learning_rate": 5.899065604459814e-06,
"loss": 0.3033,
"step": 597
},
{
"epoch": 2.4016064257028114,
"grad_norm": 0.9669124484062195,
"learning_rate": 5.823870309951096e-06,
"loss": 0.2675,
"step": 598
},
{
"epoch": 2.4056224899598395,
"grad_norm": 0.9362532496452332,
"learning_rate": 5.749094119018431e-06,
"loss": 0.2854,
"step": 599
},
{
"epoch": 2.4096385542168672,
"grad_norm": 0.8828224539756775,
"learning_rate": 5.674738665931575e-06,
"loss": 0.2155,
"step": 600
},
{
"epoch": 2.4136546184738954,
"grad_norm": 0.8983003497123718,
"learning_rate": 5.600805575764878e-06,
"loss": 0.2563,
"step": 601
},
{
"epoch": 2.4176706827309236,
"grad_norm": 0.9212509989738464,
"learning_rate": 5.527296464361726e-06,
"loss": 0.2815,
"step": 602
},
{
"epoch": 2.4216867469879517,
"grad_norm": 0.9113696217536926,
"learning_rate": 5.454212938299255e-06,
"loss": 0.2933,
"step": 603
},
{
"epoch": 2.42570281124498,
"grad_norm": 1.1318951845169067,
"learning_rate": 5.381556594853227e-06,
"loss": 0.3307,
"step": 604
},
{
"epoch": 2.429718875502008,
"grad_norm": 0.9144623875617981,
"learning_rate": 5.3093290219631165e-06,
"loss": 0.2928,
"step": 605
},
{
"epoch": 2.433734939759036,
"grad_norm": 1.0059735774993896,
"learning_rate": 5.237531798197415e-06,
"loss": 0.33,
"step": 606
},
{
"epoch": 2.4377510040160644,
"grad_norm": 1.103473424911499,
"learning_rate": 5.166166492719124e-06,
"loss": 0.2933,
"step": 607
},
{
"epoch": 2.4417670682730925,
"grad_norm": 1.002747893333435,
"learning_rate": 5.0952346652514635e-06,
"loss": 0.2399,
"step": 608
},
{
"epoch": 2.4457831325301207,
"grad_norm": 0.8700054883956909,
"learning_rate": 5.02473786604378e-06,
"loss": 0.2665,
"step": 609
},
{
"epoch": 2.4497991967871484,
"grad_norm": 0.9539564847946167,
"learning_rate": 4.954677635837668e-06,
"loss": 0.2912,
"step": 610
},
{
"epoch": 2.4538152610441766,
"grad_norm": 1.2253509759902954,
"learning_rate": 4.885055505833291e-06,
"loss": 0.3908,
"step": 611
},
{
"epoch": 2.4578313253012047,
"grad_norm": 0.9646692276000977,
"learning_rate": 4.81587299765594e-06,
"loss": 0.2735,
"step": 612
},
{
"epoch": 2.461847389558233,
"grad_norm": 1.0701141357421875,
"learning_rate": 4.747131623322737e-06,
"loss": 0.3467,
"step": 613
},
{
"epoch": 2.465863453815261,
"grad_norm": 1.003066897392273,
"learning_rate": 4.678832885209622e-06,
"loss": 0.2982,
"step": 614
},
{
"epoch": 2.4698795180722892,
"grad_norm": 1.0548231601715088,
"learning_rate": 4.610978276018496e-06,
"loss": 0.3342,
"step": 615
},
{
"epoch": 2.4738955823293174,
"grad_norm": 0.9853499531745911,
"learning_rate": 4.543569278744625e-06,
"loss": 0.2735,
"step": 616
},
{
"epoch": 2.4779116465863456,
"grad_norm": 1.1168359518051147,
"learning_rate": 4.476607366644192e-06,
"loss": 0.3019,
"step": 617
},
{
"epoch": 2.4819277108433733,
"grad_norm": 1.0065770149230957,
"learning_rate": 4.410094003202134e-06,
"loss": 0.3089,
"step": 618
},
{
"epoch": 2.4859437751004014,
"grad_norm": 1.005988597869873,
"learning_rate": 4.344030642100133e-06,
"loss": 0.2691,
"step": 619
},
{
"epoch": 2.4899598393574296,
"grad_norm": 0.9778465628623962,
"learning_rate": 4.278418727184849e-06,
"loss": 0.2609,
"step": 620
},
{
"epoch": 2.4939759036144578,
"grad_norm": 0.9522945880889893,
"learning_rate": 4.213259692436367e-06,
"loss": 0.2752,
"step": 621
},
{
"epoch": 2.497991967871486,
"grad_norm": 1.0316386222839355,
"learning_rate": 4.148554961936876e-06,
"loss": 0.259,
"step": 622
},
{
"epoch": 2.502008032128514,
"grad_norm": 0.9173260927200317,
"learning_rate": 4.0843059498395065e-06,
"loss": 0.2776,
"step": 623
},
{
"epoch": 2.5060240963855422,
"grad_norm": 0.8724738359451294,
"learning_rate": 4.0205140603374465e-06,
"loss": 0.252,
"step": 624
},
{
"epoch": 2.5100401606425704,
"grad_norm": 1.024414300918579,
"learning_rate": 3.957180687633252e-06,
"loss": 0.312,
"step": 625
},
{
"epoch": 2.5140562248995986,
"grad_norm": 0.9675637483596802,
"learning_rate": 3.894307215908371e-06,
"loss": 0.296,
"step": 626
},
{
"epoch": 2.5180722891566267,
"grad_norm": 1.0301586389541626,
"learning_rate": 3.831895019292897e-06,
"loss": 0.2816,
"step": 627
},
{
"epoch": 2.522088353413655,
"grad_norm": 0.9780864715576172,
"learning_rate": 3.769945461835531e-06,
"loss": 0.2968,
"step": 628
},
{
"epoch": 2.5261044176706826,
"grad_norm": 1.0341154336929321,
"learning_rate": 3.708459897473768e-06,
"loss": 0.2843,
"step": 629
},
{
"epoch": 2.5301204819277108,
"grad_norm": 1.1320409774780273,
"learning_rate": 3.6474396700043158e-06,
"loss": 0.306,
"step": 630
},
{
"epoch": 2.534136546184739,
"grad_norm": 1.0244733095169067,
"learning_rate": 3.586886113053717e-06,
"loss": 0.2872,
"step": 631
},
{
"epoch": 2.538152610441767,
"grad_norm": 1.066190242767334,
"learning_rate": 3.5268005500492003e-06,
"loss": 0.3342,
"step": 632
},
{
"epoch": 2.5421686746987953,
"grad_norm": 1.1290909051895142,
"learning_rate": 3.4671842941897765e-06,
"loss": 0.3326,
"step": 633
},
{
"epoch": 2.5461847389558234,
"grad_norm": 1.0926635265350342,
"learning_rate": 3.4080386484175024e-06,
"loss": 0.3094,
"step": 634
},
{
"epoch": 2.550200803212851,
"grad_norm": 1.0564324855804443,
"learning_rate": 3.3493649053890326e-06,
"loss": 0.2822,
"step": 635
},
{
"epoch": 2.5542168674698793,
"grad_norm": 1.0619338750839233,
"learning_rate": 3.2911643474473646e-06,
"loss": 0.2945,
"step": 636
},
{
"epoch": 2.5582329317269075,
"grad_norm": 0.9817492961883545,
"learning_rate": 3.233438246593795e-06,
"loss": 0.2891,
"step": 637
},
{
"epoch": 2.5622489959839356,
"grad_norm": 0.9481024146080017,
"learning_rate": 3.1761878644601424e-06,
"loss": 0.2383,
"step": 638
},
{
"epoch": 2.566265060240964,
"grad_norm": 0.9262716174125671,
"learning_rate": 3.119414452281158e-06,
"loss": 0.287,
"step": 639
},
{
"epoch": 2.570281124497992,
"grad_norm": 1.0571141242980957,
"learning_rate": 3.0631192508671857e-06,
"loss": 0.2832,
"step": 640
},
{
"epoch": 2.57429718875502,
"grad_norm": 1.0114316940307617,
"learning_rate": 3.007303490577043e-06,
"loss": 0.2715,
"step": 641
},
{
"epoch": 2.5783132530120483,
"grad_norm": 0.9859365224838257,
"learning_rate": 2.9519683912911266e-06,
"loss": 0.2556,
"step": 642
},
{
"epoch": 2.5823293172690764,
"grad_norm": 0.9159937500953674,
"learning_rate": 2.8971151623847587e-06,
"loss": 0.2602,
"step": 643
},
{
"epoch": 2.5863453815261046,
"grad_norm": 1.0229847431182861,
"learning_rate": 2.8427450027017493e-06,
"loss": 0.254,
"step": 644
},
{
"epoch": 2.5903614457831328,
"grad_norm": 1.029196858406067,
"learning_rate": 2.788859100528196e-06,
"loss": 0.2929,
"step": 645
},
{
"epoch": 2.5943775100401605,
"grad_norm": 1.0137423276901245,
"learning_rate": 2.7354586335665207e-06,
"loss": 0.258,
"step": 646
},
{
"epoch": 2.5983935742971886,
"grad_norm": 0.9362167119979858,
"learning_rate": 2.682544768909717e-06,
"loss": 0.258,
"step": 647
},
{
"epoch": 2.602409638554217,
"grad_norm": 0.940181314945221,
"learning_rate": 2.6301186630158485e-06,
"loss": 0.2775,
"step": 648
},
{
"epoch": 2.606425702811245,
"grad_norm": 1.0691094398498535,
"learning_rate": 2.578181461682794e-06,
"loss": 0.3158,
"step": 649
},
{
"epoch": 2.610441767068273,
"grad_norm": 0.8856833577156067,
"learning_rate": 2.526734300023162e-06,
"loss": 0.2632,
"step": 650
},
{
"epoch": 2.6144578313253013,
"grad_norm": 0.9836992025375366,
"learning_rate": 2.475778302439524e-06,
"loss": 0.2691,
"step": 651
},
{
"epoch": 2.6184738955823295,
"grad_norm": 0.9964855909347534,
"learning_rate": 2.4253145825998137e-06,
"loss": 0.2473,
"step": 652
},
{
"epoch": 2.622489959839357,
"grad_norm": 0.9155222177505493,
"learning_rate": 2.3753442434129997e-06,
"loss": 0.2215,
"step": 653
},
{
"epoch": 2.6265060240963853,
"grad_norm": 1.017866611480713,
"learning_rate": 2.325868377004986e-06,
"loss": 0.2686,
"step": 654
},
{
"epoch": 2.6305220883534135,
"grad_norm": 1.0167937278747559,
"learning_rate": 2.2768880646947268e-06,
"loss": 0.2424,
"step": 655
},
{
"epoch": 2.6345381526104417,
"grad_norm": 1.1613540649414062,
"learning_rate": 2.2284043769706027e-06,
"loss": 0.2903,
"step": 656
},
{
"epoch": 2.63855421686747,
"grad_norm": 1.0084424018859863,
"learning_rate": 2.1804183734670277e-06,
"loss": 0.2861,
"step": 657
},
{
"epoch": 2.642570281124498,
"grad_norm": 0.96149080991745,
"learning_rate": 2.1329311029412797e-06,
"loss": 0.2732,
"step": 658
},
{
"epoch": 2.646586345381526,
"grad_norm": 0.9497535824775696,
"learning_rate": 2.0859436032505952e-06,
"loss": 0.2528,
"step": 659
},
{
"epoch": 2.6506024096385543,
"grad_norm": 1.0602991580963135,
"learning_rate": 2.039456901329473e-06,
"loss": 0.2795,
"step": 660
},
{
"epoch": 2.6546184738955825,
"grad_norm": 0.9473398327827454,
"learning_rate": 1.9934720131672357e-06,
"loss": 0.2551,
"step": 661
},
{
"epoch": 2.6586345381526106,
"grad_norm": 1.0119514465332031,
"learning_rate": 1.9479899437858245e-06,
"loss": 0.2472,
"step": 662
},
{
"epoch": 2.662650602409639,
"grad_norm": 1.049772024154663,
"learning_rate": 1.9030116872178316e-06,
"loss": 0.2846,
"step": 663
},
{
"epoch": 2.6666666666666665,
"grad_norm": 1.0074094533920288,
"learning_rate": 1.8585382264847794e-06,
"loss": 0.256,
"step": 664
},
{
"epoch": 2.6706827309236947,
"grad_norm": 1.084901213645935,
"learning_rate": 1.8145705335756297e-06,
"loss": 0.3454,
"step": 665
},
{
"epoch": 2.674698795180723,
"grad_norm": 1.0208693742752075,
"learning_rate": 1.771109569425547e-06,
"loss": 0.2574,
"step": 666
},
{
"epoch": 2.678714859437751,
"grad_norm": 0.9760468602180481,
"learning_rate": 1.7281562838948966e-06,
"loss": 0.2695,
"step": 667
},
{
"epoch": 2.682730923694779,
"grad_norm": 1.043257474899292,
"learning_rate": 1.6857116157484753e-06,
"loss": 0.2763,
"step": 668
},
{
"epoch": 2.6867469879518073,
"grad_norm": 1.030243992805481,
"learning_rate": 1.6437764926350074e-06,
"loss": 0.2936,
"step": 669
},
{
"epoch": 2.6907630522088355,
"grad_norm": 1.028914451599121,
"learning_rate": 1.6023518310668618e-06,
"loss": 0.2742,
"step": 670
},
{
"epoch": 2.694779116465863,
"grad_norm": 0.9653047919273376,
"learning_rate": 1.5614385364000228e-06,
"loss": 0.2326,
"step": 671
},
{
"epoch": 2.6987951807228914,
"grad_norm": 1.0667688846588135,
"learning_rate": 1.5210375028143097e-06,
"loss": 0.3188,
"step": 672
},
{
"epoch": 2.7028112449799195,
"grad_norm": 1.0015482902526855,
"learning_rate": 1.4811496132938196e-06,
"loss": 0.2942,
"step": 673
},
{
"epoch": 2.7068273092369477,
"grad_norm": 0.9504645466804504,
"learning_rate": 1.4417757396076425e-06,
"loss": 0.252,
"step": 674
},
{
"epoch": 2.710843373493976,
"grad_norm": 0.9457132816314697,
"learning_rate": 1.4029167422908107e-06,
"loss": 0.2785,
"step": 675
},
{
"epoch": 2.714859437751004,
"grad_norm": 0.9735773801803589,
"learning_rate": 1.364573470625477e-06,
"loss": 0.2757,
"step": 676
},
{
"epoch": 2.718875502008032,
"grad_norm": 1.1076792478561401,
"learning_rate": 1.3267467626223606e-06,
"loss": 0.2984,
"step": 677
},
{
"epoch": 2.7228915662650603,
"grad_norm": 1.0301073789596558,
"learning_rate": 1.2894374450024338e-06,
"loss": 0.2856,
"step": 678
},
{
"epoch": 2.7269076305220885,
"grad_norm": 1.1026225090026855,
"learning_rate": 1.25264633317885e-06,
"loss": 0.2769,
"step": 679
},
{
"epoch": 2.7309236947791167,
"grad_norm": 1.3219835758209229,
"learning_rate": 1.216374231239134e-06,
"loss": 0.2904,
"step": 680
},
{
"epoch": 2.734939759036145,
"grad_norm": 0.9455382823944092,
"learning_rate": 1.180621931927592e-06,
"loss": 0.2344,
"step": 681
},
{
"epoch": 2.7389558232931726,
"grad_norm": 0.9791133999824524,
"learning_rate": 1.1453902166279895e-06,
"loss": 0.2742,
"step": 682
},
{
"epoch": 2.7429718875502007,
"grad_norm": 0.9991709589958191,
"learning_rate": 1.1106798553464804e-06,
"loss": 0.2547,
"step": 683
},
{
"epoch": 2.746987951807229,
"grad_norm": 0.9072936773300171,
"learning_rate": 1.0764916066947794e-06,
"loss": 0.2326,
"step": 684
},
{
"epoch": 2.751004016064257,
"grad_norm": 1.0011482238769531,
"learning_rate": 1.042826217873566e-06,
"loss": 0.2756,
"step": 685
},
{
"epoch": 2.755020080321285,
"grad_norm": 1.135215163230896,
"learning_rate": 1.0096844246561794e-06,
"loss": 0.3407,
"step": 686
},
{
"epoch": 2.7590361445783134,
"grad_norm": 0.9893269538879395,
"learning_rate": 9.770669513725128e-07,
"loss": 0.2826,
"step": 687
},
{
"epoch": 2.7630522088353415,
"grad_norm": 1.1436357498168945,
"learning_rate": 9.449745108931985e-07,
"loss": 0.3692,
"step": 688
},
{
"epoch": 2.7670682730923692,
"grad_norm": 1.0183185338974,
"learning_rate": 9.134078046140248e-07,
"loss": 0.2917,
"step": 689
},
{
"epoch": 2.7710843373493974,
"grad_norm": 1.0603082180023193,
"learning_rate": 8.823675224406053e-07,
"loss": 0.2667,
"step": 690
},
{
"epoch": 2.7751004016064256,
"grad_norm": 1.0653798580169678,
"learning_rate": 8.51854342773295e-07,
"loss": 0.2404,
"step": 691
},
{
"epoch": 2.7791164658634537,
"grad_norm": 1.0163860321044922,
"learning_rate": 8.218689324923728e-07,
"loss": 0.3033,
"step": 692
},
{
"epoch": 2.783132530120482,
"grad_norm": 1.1382652521133423,
"learning_rate": 7.924119469434665e-07,
"loss": 0.2795,
"step": 693
},
{
"epoch": 2.78714859437751,
"grad_norm": 1.0515531301498413,
"learning_rate": 7.634840299232171e-07,
"loss": 0.3137,
"step": 694
},
{
"epoch": 2.791164658634538,
"grad_norm": 1.016628384590149,
"learning_rate": 7.350858136652261e-07,
"loss": 0.2685,
"step": 695
},
{
"epoch": 2.7951807228915664,
"grad_norm": 0.9027230739593506,
"learning_rate": 7.072179188262251e-07,
"loss": 0.2717,
"step": 696
},
{
"epoch": 2.7991967871485945,
"grad_norm": 0.9216163754463196,
"learning_rate": 6.798809544725171e-07,
"loss": 0.2571,
"step": 697
},
{
"epoch": 2.8032128514056227,
"grad_norm": 1.0517877340316772,
"learning_rate": 6.530755180666592e-07,
"loss": 0.2658,
"step": 698
},
{
"epoch": 2.807228915662651,
"grad_norm": 1.0404746532440186,
"learning_rate": 6.268021954544096e-07,
"loss": 0.3251,
"step": 699
},
{
"epoch": 2.8112449799196786,
"grad_norm": 1.021146297454834,
"learning_rate": 6.01061560851926e-07,
"loss": 0.288,
"step": 700
},
{
"epoch": 2.8152610441767068,
"grad_norm": 0.9593784213066101,
"learning_rate": 5.758541768332071e-07,
"loss": 0.2732,
"step": 701
},
{
"epoch": 2.819277108433735,
"grad_norm": 1.1057608127593994,
"learning_rate": 5.5118059431781e-07,
"loss": 0.3118,
"step": 702
},
{
"epoch": 2.823293172690763,
"grad_norm": 0.915035605430603,
"learning_rate": 5.270413525587909e-07,
"loss": 0.2067,
"step": 703
},
{
"epoch": 2.8273092369477912,
"grad_norm": 1.1384711265563965,
"learning_rate": 5.03436979130939e-07,
"loss": 0.3264,
"step": 704
},
{
"epoch": 2.8313253012048194,
"grad_norm": 1.0579373836517334,
"learning_rate": 4.803679899192392e-07,
"loss": 0.3069,
"step": 705
},
{
"epoch": 2.835341365461847,
"grad_norm": 1.0465303659439087,
"learning_rate": 4.57834889107589e-07,
"loss": 0.2865,
"step": 706
},
{
"epoch": 2.8393574297188753,
"grad_norm": 0.9909630417823792,
"learning_rate": 4.358381691677932e-07,
"loss": 0.2333,
"step": 707
},
{
"epoch": 2.8433734939759034,
"grad_norm": 0.9599091410636902,
"learning_rate": 4.1437831084878974e-07,
"loss": 0.2484,
"step": 708
},
{
"epoch": 2.8473895582329316,
"grad_norm": 0.9916269779205322,
"learning_rate": 3.9345578316614396e-07,
"loss": 0.287,
"step": 709
},
{
"epoch": 2.8514056224899598,
"grad_norm": 1.1012687683105469,
"learning_rate": 3.730710433918039e-07,
"loss": 0.2694,
"step": 710
},
{
"epoch": 2.855421686746988,
"grad_norm": 1.0106234550476074,
"learning_rate": 3.5322453704410286e-07,
"loss": 0.2871,
"step": 711
},
{
"epoch": 2.859437751004016,
"grad_norm": 1.0441107749938965,
"learning_rate": 3.339166978780256e-07,
"loss": 0.2951,
"step": 712
},
{
"epoch": 2.8634538152610443,
"grad_norm": 1.1031628847122192,
"learning_rate": 3.1514794787571854e-07,
"loss": 0.3026,
"step": 713
},
{
"epoch": 2.8674698795180724,
"grad_norm": 1.0138731002807617,
"learning_rate": 2.969186972372806e-07,
"loss": 0.2397,
"step": 714
},
{
"epoch": 2.8714859437751006,
"grad_norm": 1.1411104202270508,
"learning_rate": 2.7922934437178695e-07,
"loss": 0.3456,
"step": 715
},
{
"epoch": 2.8755020080321287,
"grad_norm": 1.181161642074585,
"learning_rate": 2.620802758885876e-07,
"loss": 0.3124,
"step": 716
},
{
"epoch": 2.8795180722891565,
"grad_norm": 1.0166224241256714,
"learning_rate": 2.454718665888589e-07,
"loss": 0.3142,
"step": 717
},
{
"epoch": 2.8835341365461846,
"grad_norm": 1.0102145671844482,
"learning_rate": 2.2940447945740962e-07,
"loss": 0.3187,
"step": 718
},
{
"epoch": 2.887550200803213,
"grad_norm": 1.0226353406906128,
"learning_rate": 2.1387846565474045e-07,
"loss": 0.2711,
"step": 719
},
{
"epoch": 2.891566265060241,
"grad_norm": 1.0354499816894531,
"learning_rate": 1.9889416450938337e-07,
"loss": 0.2537,
"step": 720
},
{
"epoch": 2.895582329317269,
"grad_norm": 0.8995523452758789,
"learning_rate": 1.844519035104797e-07,
"loss": 0.218,
"step": 721
},
{
"epoch": 2.8995983935742973,
"grad_norm": 0.9724816679954529,
"learning_rate": 1.7055199830061652e-07,
"loss": 0.2262,
"step": 722
},
{
"epoch": 2.9036144578313254,
"grad_norm": 0.9942877888679504,
"learning_rate": 1.571947526689349e-07,
"loss": 0.2676,
"step": 723
},
{
"epoch": 2.907630522088353,
"grad_norm": 1.0389326810836792,
"learning_rate": 1.4438045854449355e-07,
"loss": 0.2532,
"step": 724
},
{
"epoch": 2.9116465863453813,
"grad_norm": 0.9621029496192932,
"learning_rate": 1.3210939598987392e-07,
"loss": 0.243,
"step": 725
},
{
"epoch": 2.9156626506024095,
"grad_norm": 1.1710126399993896,
"learning_rate": 1.2038183319507955e-07,
"loss": 0.2951,
"step": 726
},
{
"epoch": 2.9196787148594376,
"grad_norm": 1.0926920175552368,
"learning_rate": 1.0919802647165466e-07,
"loss": 0.3208,
"step": 727
},
{
"epoch": 2.923694779116466,
"grad_norm": 0.9979294538497925,
"learning_rate": 9.855822024710248e-08,
"loss": 0.2534,
"step": 728
},
{
"epoch": 2.927710843373494,
"grad_norm": 1.0968432426452637,
"learning_rate": 8.846264705952289e-08,
"loss": 0.3354,
"step": 729
},
{
"epoch": 2.931726907630522,
"grad_norm": 0.9580751061439514,
"learning_rate": 7.891152755254428e-08,
"loss": 0.2418,
"step": 730
},
{
"epoch": 2.9357429718875503,
"grad_norm": 1.0959455966949463,
"learning_rate": 6.990507047049676e-08,
"loss": 0.2974,
"step": 731
},
{
"epoch": 2.9397590361445785,
"grad_norm": 1.103136658668518,
"learning_rate": 6.14434726538493e-08,
"loss": 0.3499,
"step": 732
},
{
"epoch": 2.9437751004016066,
"grad_norm": 1.0734608173370361,
"learning_rate": 5.352691903491303e-08,
"loss": 0.2999,
"step": 733
},
{
"epoch": 2.9477911646586348,
"grad_norm": 0.9409282803535461,
"learning_rate": 4.615558263378617e-08,
"loss": 0.2473,
"step": 734
},
{
"epoch": 2.9518072289156625,
"grad_norm": 0.918748140335083,
"learning_rate": 3.9329624554584884e-08,
"loss": 0.2743,
"step": 735
},
{
"epoch": 2.9558232931726907,
"grad_norm": 0.9397270083427429,
"learning_rate": 3.304919398192663e-08,
"loss": 0.2784,
"step": 736
},
{
"epoch": 2.959839357429719,
"grad_norm": 1.139475703239441,
"learning_rate": 2.731442817764662e-08,
"loss": 0.2644,
"step": 737
},
{
"epoch": 2.963855421686747,
"grad_norm": 1.0574469566345215,
"learning_rate": 2.2125452477828047e-08,
"loss": 0.3031,
"step": 738
},
{
"epoch": 2.967871485943775,
"grad_norm": 0.9595881104469299,
"learning_rate": 1.7482380290034794e-08,
"loss": 0.2359,
"step": 739
},
{
"epoch": 2.9718875502008033,
"grad_norm": 1.00413179397583,
"learning_rate": 1.3385313090857887e-08,
"loss": 0.3167,
"step": 740
},
{
"epoch": 2.9759036144578315,
"grad_norm": 0.921945333480835,
"learning_rate": 9.834340423678368e-09,
"loss": 0.2687,
"step": 741
},
{
"epoch": 2.979919678714859,
"grad_norm": 1.0744749307632446,
"learning_rate": 6.829539896716086e-09,
"loss": 0.2881,
"step": 742
},
{
"epoch": 2.9839357429718874,
"grad_norm": 0.913150429725647,
"learning_rate": 4.370977181339386e-09,
"loss": 0.2319,
"step": 743
},
{
"epoch": 2.9879518072289155,
"grad_norm": 1.1232759952545166,
"learning_rate": 2.4587060106245895e-09,
"loss": 0.2681,
"step": 744
},
{
"epoch": 2.9919678714859437,
"grad_norm": 1.0994449853897095,
"learning_rate": 1.0927681781819355e-09,
"loss": 0.2662,
"step": 745
},
{
"epoch": 2.995983935742972,
"grad_norm": 0.8616254925727844,
"learning_rate": 2.7319353723964656e-10,
"loss": 0.2212,
"step": 746
},
{
"epoch": 3.0,
"grad_norm": 1.0538721084594727,
"learning_rate": 0.0,
"loss": 0.2558,
"step": 747
},
{
"epoch": 3.0,
"step": 747,
"total_flos": 8.987881602831483e+17,
"train_loss": 0.363180737478666,
"train_runtime": 7967.6952,
"train_samples_per_second": 11.999,
"train_steps_per_second": 0.094
}
],
"logging_steps": 1,
"max_steps": 747,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8.987881602831483e+17,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}