model_qvlm7b-codegen-expert / trainer_state.json
anasmahasin's picture
Upload trainer_state.json with huggingface_hub
b13e6f4 verified
Raw
History Blame Contribute Delete
277 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 6.0,
"eval_steps": 500,
"global_step": 1590,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003780718336483932,
"grad_norm": 26.49860954284668,
"learning_rate": 0.0,
"loss": 8.6314,
"step": 1
},
{
"epoch": 0.007561436672967864,
"grad_norm": 26.446386337280273,
"learning_rate": 3.1446540880503146e-08,
"loss": 8.5661,
"step": 2
},
{
"epoch": 0.011342155009451797,
"grad_norm": 26.49188804626465,
"learning_rate": 6.289308176100629e-08,
"loss": 8.6126,
"step": 3
},
{
"epoch": 0.015122873345935728,
"grad_norm": 26.449148178100586,
"learning_rate": 9.433962264150944e-08,
"loss": 8.6061,
"step": 4
},
{
"epoch": 0.01890359168241966,
"grad_norm": 26.4529972076416,
"learning_rate": 1.2578616352201258e-07,
"loss": 8.564,
"step": 5
},
{
"epoch": 0.022684310018903593,
"grad_norm": 26.17426109313965,
"learning_rate": 1.5723270440251575e-07,
"loss": 8.5168,
"step": 6
},
{
"epoch": 0.026465028355387523,
"grad_norm": 26.439720153808594,
"learning_rate": 1.886792452830189e-07,
"loss": 8.4829,
"step": 7
},
{
"epoch": 0.030245746691871456,
"grad_norm": 26.469741821289062,
"learning_rate": 2.2012578616352203e-07,
"loss": 8.5575,
"step": 8
},
{
"epoch": 0.034026465028355386,
"grad_norm": 26.483909606933594,
"learning_rate": 2.5157232704402517e-07,
"loss": 8.6731,
"step": 9
},
{
"epoch": 0.03780718336483932,
"grad_norm": 26.617839813232422,
"learning_rate": 2.8301886792452833e-07,
"loss": 8.5927,
"step": 10
},
{
"epoch": 0.04158790170132325,
"grad_norm": 26.199474334716797,
"learning_rate": 3.144654088050315e-07,
"loss": 8.581,
"step": 11
},
{
"epoch": 0.045368620037807186,
"grad_norm": 26.53860092163086,
"learning_rate": 3.459119496855346e-07,
"loss": 8.6535,
"step": 12
},
{
"epoch": 0.04914933837429111,
"grad_norm": 26.27257537841797,
"learning_rate": 3.773584905660378e-07,
"loss": 8.4668,
"step": 13
},
{
"epoch": 0.052930056710775046,
"grad_norm": 26.675107955932617,
"learning_rate": 4.088050314465409e-07,
"loss": 8.6081,
"step": 14
},
{
"epoch": 0.05671077504725898,
"grad_norm": 26.27010154724121,
"learning_rate": 4.4025157232704405e-07,
"loss": 8.5503,
"step": 15
},
{
"epoch": 0.06049149338374291,
"grad_norm": 26.363964080810547,
"learning_rate": 4.716981132075472e-07,
"loss": 8.581,
"step": 16
},
{
"epoch": 0.06427221172022685,
"grad_norm": 26.43025016784668,
"learning_rate": 5.031446540880503e-07,
"loss": 8.5613,
"step": 17
},
{
"epoch": 0.06805293005671077,
"grad_norm": 26.831392288208008,
"learning_rate": 5.345911949685535e-07,
"loss": 8.4854,
"step": 18
},
{
"epoch": 0.07183364839319471,
"grad_norm": 26.333005905151367,
"learning_rate": 5.660377358490567e-07,
"loss": 8.4525,
"step": 19
},
{
"epoch": 0.07561436672967864,
"grad_norm": 26.117008209228516,
"learning_rate": 5.974842767295598e-07,
"loss": 8.4525,
"step": 20
},
{
"epoch": 0.07939508506616257,
"grad_norm": 26.460399627685547,
"learning_rate": 6.28930817610063e-07,
"loss": 8.5734,
"step": 21
},
{
"epoch": 0.0831758034026465,
"grad_norm": 26.33306884765625,
"learning_rate": 6.603773584905661e-07,
"loss": 8.4768,
"step": 22
},
{
"epoch": 0.08695652173913043,
"grad_norm": 25.97040557861328,
"learning_rate": 6.918238993710692e-07,
"loss": 8.4153,
"step": 23
},
{
"epoch": 0.09073724007561437,
"grad_norm": 25.835500717163086,
"learning_rate": 7.232704402515723e-07,
"loss": 8.3122,
"step": 24
},
{
"epoch": 0.0945179584120983,
"grad_norm": 25.97608757019043,
"learning_rate": 7.547169811320755e-07,
"loss": 8.4553,
"step": 25
},
{
"epoch": 0.09829867674858223,
"grad_norm": 26.065210342407227,
"learning_rate": 7.861635220125787e-07,
"loss": 8.2699,
"step": 26
},
{
"epoch": 0.10207939508506617,
"grad_norm": 25.830997467041016,
"learning_rate": 8.176100628930818e-07,
"loss": 8.2455,
"step": 27
},
{
"epoch": 0.10586011342155009,
"grad_norm": 25.63951873779297,
"learning_rate": 8.490566037735849e-07,
"loss": 8.2783,
"step": 28
},
{
"epoch": 0.10964083175803403,
"grad_norm": 25.445558547973633,
"learning_rate": 8.805031446540881e-07,
"loss": 8.2718,
"step": 29
},
{
"epoch": 0.11342155009451796,
"grad_norm": 25.129175186157227,
"learning_rate": 9.119496855345912e-07,
"loss": 8.1379,
"step": 30
},
{
"epoch": 0.11720226843100189,
"grad_norm": 25.028966903686523,
"learning_rate": 9.433962264150944e-07,
"loss": 8.1035,
"step": 31
},
{
"epoch": 0.12098298676748583,
"grad_norm": 25.174604415893555,
"learning_rate": 9.748427672955975e-07,
"loss": 8.0904,
"step": 32
},
{
"epoch": 0.12476370510396975,
"grad_norm": 24.973167419433594,
"learning_rate": 1.0062893081761007e-06,
"loss": 7.9715,
"step": 33
},
{
"epoch": 0.1285444234404537,
"grad_norm": 24.82118034362793,
"learning_rate": 1.037735849056604e-06,
"loss": 7.9507,
"step": 34
},
{
"epoch": 0.1323251417769376,
"grad_norm": 24.09725570678711,
"learning_rate": 1.069182389937107e-06,
"loss": 7.8026,
"step": 35
},
{
"epoch": 0.13610586011342155,
"grad_norm": 24.269426345825195,
"learning_rate": 1.1006289308176102e-06,
"loss": 7.7298,
"step": 36
},
{
"epoch": 0.13988657844990549,
"grad_norm": 23.643186569213867,
"learning_rate": 1.1320754716981133e-06,
"loss": 7.6304,
"step": 37
},
{
"epoch": 0.14366729678638943,
"grad_norm": 23.674152374267578,
"learning_rate": 1.1635220125786164e-06,
"loss": 7.6305,
"step": 38
},
{
"epoch": 0.14744801512287334,
"grad_norm": 22.583824157714844,
"learning_rate": 1.1949685534591195e-06,
"loss": 7.4051,
"step": 39
},
{
"epoch": 0.15122873345935728,
"grad_norm": 22.695629119873047,
"learning_rate": 1.2264150943396227e-06,
"loss": 7.4661,
"step": 40
},
{
"epoch": 0.15500945179584122,
"grad_norm": 21.805404663085938,
"learning_rate": 1.257861635220126e-06,
"loss": 7.3498,
"step": 41
},
{
"epoch": 0.15879017013232513,
"grad_norm": 21.137290954589844,
"learning_rate": 1.289308176100629e-06,
"loss": 7.0845,
"step": 42
},
{
"epoch": 0.16257088846880907,
"grad_norm": 20.846675872802734,
"learning_rate": 1.3207547169811322e-06,
"loss": 6.9925,
"step": 43
},
{
"epoch": 0.166351606805293,
"grad_norm": 19.387805938720703,
"learning_rate": 1.3522012578616353e-06,
"loss": 6.8302,
"step": 44
},
{
"epoch": 0.17013232514177692,
"grad_norm": 19.525684356689453,
"learning_rate": 1.3836477987421384e-06,
"loss": 6.8045,
"step": 45
},
{
"epoch": 0.17391304347826086,
"grad_norm": 18.30414390563965,
"learning_rate": 1.4150943396226415e-06,
"loss": 6.6591,
"step": 46
},
{
"epoch": 0.1776937618147448,
"grad_norm": 17.67142105102539,
"learning_rate": 1.4465408805031447e-06,
"loss": 6.5074,
"step": 47
},
{
"epoch": 0.18147448015122875,
"grad_norm": 16.708816528320312,
"learning_rate": 1.4779874213836478e-06,
"loss": 6.3829,
"step": 48
},
{
"epoch": 0.18525519848771266,
"grad_norm": 15.584643363952637,
"learning_rate": 1.509433962264151e-06,
"loss": 6.2258,
"step": 49
},
{
"epoch": 0.1890359168241966,
"grad_norm": 14.791690826416016,
"learning_rate": 1.5408805031446542e-06,
"loss": 6.1332,
"step": 50
},
{
"epoch": 0.19281663516068054,
"grad_norm": 14.2457275390625,
"learning_rate": 1.5723270440251573e-06,
"loss": 5.9686,
"step": 51
},
{
"epoch": 0.19659735349716445,
"grad_norm": 12.846230506896973,
"learning_rate": 1.6037735849056604e-06,
"loss": 5.8422,
"step": 52
},
{
"epoch": 0.2003780718336484,
"grad_norm": 12.019917488098145,
"learning_rate": 1.6352201257861635e-06,
"loss": 5.7827,
"step": 53
},
{
"epoch": 0.20415879017013233,
"grad_norm": 11.415548324584961,
"learning_rate": 1.6666666666666667e-06,
"loss": 5.6459,
"step": 54
},
{
"epoch": 0.20793950850661624,
"grad_norm": 10.511237144470215,
"learning_rate": 1.6981132075471698e-06,
"loss": 5.5174,
"step": 55
},
{
"epoch": 0.21172022684310018,
"grad_norm": 9.58881950378418,
"learning_rate": 1.7295597484276729e-06,
"loss": 5.3711,
"step": 56
},
{
"epoch": 0.21550094517958412,
"grad_norm": 9.30813980102539,
"learning_rate": 1.7610062893081762e-06,
"loss": 5.2917,
"step": 57
},
{
"epoch": 0.21928166351606806,
"grad_norm": 8.832740783691406,
"learning_rate": 1.7924528301886793e-06,
"loss": 5.1584,
"step": 58
},
{
"epoch": 0.22306238185255198,
"grad_norm": 8.407164573669434,
"learning_rate": 1.8238993710691824e-06,
"loss": 5.063,
"step": 59
},
{
"epoch": 0.22684310018903592,
"grad_norm": 7.9452314376831055,
"learning_rate": 1.8553459119496855e-06,
"loss": 4.9789,
"step": 60
},
{
"epoch": 0.23062381852551986,
"grad_norm": 7.413273811340332,
"learning_rate": 1.8867924528301889e-06,
"loss": 4.8478,
"step": 61
},
{
"epoch": 0.23440453686200377,
"grad_norm": 6.946573734283447,
"learning_rate": 1.918238993710692e-06,
"loss": 4.7507,
"step": 62
},
{
"epoch": 0.2381852551984877,
"grad_norm": 6.1533098220825195,
"learning_rate": 1.949685534591195e-06,
"loss": 4.6498,
"step": 63
},
{
"epoch": 0.24196597353497165,
"grad_norm": 6.1265459060668945,
"learning_rate": 1.981132075471698e-06,
"loss": 4.5882,
"step": 64
},
{
"epoch": 0.24574669187145556,
"grad_norm": 5.499799728393555,
"learning_rate": 2.0125786163522013e-06,
"loss": 4.4678,
"step": 65
},
{
"epoch": 0.2495274102079395,
"grad_norm": 5.957705974578857,
"learning_rate": 2.044025157232705e-06,
"loss": 4.4099,
"step": 66
},
{
"epoch": 0.2533081285444234,
"grad_norm": 4.965778827667236,
"learning_rate": 2.075471698113208e-06,
"loss": 4.2544,
"step": 67
},
{
"epoch": 0.2570888468809074,
"grad_norm": 4.846767902374268,
"learning_rate": 2.106918238993711e-06,
"loss": 4.1501,
"step": 68
},
{
"epoch": 0.2608695652173913,
"grad_norm": 4.912769794464111,
"learning_rate": 2.138364779874214e-06,
"loss": 4.0971,
"step": 69
},
{
"epoch": 0.2646502835538752,
"grad_norm": 4.732649326324463,
"learning_rate": 2.1698113207547173e-06,
"loss": 4.0242,
"step": 70
},
{
"epoch": 0.2684310018903592,
"grad_norm": 4.664409160614014,
"learning_rate": 2.2012578616352204e-06,
"loss": 3.8954,
"step": 71
},
{
"epoch": 0.2722117202268431,
"grad_norm": 4.453545570373535,
"learning_rate": 2.2327044025157235e-06,
"loss": 3.826,
"step": 72
},
{
"epoch": 0.27599243856332706,
"grad_norm": 4.640097618103027,
"learning_rate": 2.2641509433962266e-06,
"loss": 3.7638,
"step": 73
},
{
"epoch": 0.27977315689981097,
"grad_norm": 4.071688652038574,
"learning_rate": 2.2955974842767298e-06,
"loss": 3.6744,
"step": 74
},
{
"epoch": 0.2835538752362949,
"grad_norm": 3.917768716812134,
"learning_rate": 2.327044025157233e-06,
"loss": 3.6072,
"step": 75
},
{
"epoch": 0.28733459357277885,
"grad_norm": 3.8766708374023438,
"learning_rate": 2.358490566037736e-06,
"loss": 3.5294,
"step": 76
},
{
"epoch": 0.29111531190926276,
"grad_norm": 3.7344677448272705,
"learning_rate": 2.389937106918239e-06,
"loss": 3.4474,
"step": 77
},
{
"epoch": 0.2948960302457467,
"grad_norm": 3.67256498336792,
"learning_rate": 2.421383647798742e-06,
"loss": 3.3529,
"step": 78
},
{
"epoch": 0.29867674858223064,
"grad_norm": 3.6902360916137695,
"learning_rate": 2.4528301886792453e-06,
"loss": 3.2858,
"step": 79
},
{
"epoch": 0.30245746691871456,
"grad_norm": 3.6304261684417725,
"learning_rate": 2.4842767295597484e-06,
"loss": 3.1951,
"step": 80
},
{
"epoch": 0.30623818525519847,
"grad_norm": 3.5764212608337402,
"learning_rate": 2.515723270440252e-06,
"loss": 3.1112,
"step": 81
},
{
"epoch": 0.31001890359168244,
"grad_norm": 3.480290651321411,
"learning_rate": 2.547169811320755e-06,
"loss": 3.0123,
"step": 82
},
{
"epoch": 0.31379962192816635,
"grad_norm": 3.4299983978271484,
"learning_rate": 2.578616352201258e-06,
"loss": 2.9359,
"step": 83
},
{
"epoch": 0.31758034026465026,
"grad_norm": 3.398749589920044,
"learning_rate": 2.6100628930817613e-06,
"loss": 2.8677,
"step": 84
},
{
"epoch": 0.32136105860113423,
"grad_norm": 3.494180679321289,
"learning_rate": 2.6415094339622644e-06,
"loss": 2.7776,
"step": 85
},
{
"epoch": 0.32514177693761814,
"grad_norm": 3.3304502964019775,
"learning_rate": 2.6729559748427675e-06,
"loss": 2.7106,
"step": 86
},
{
"epoch": 0.32892249527410206,
"grad_norm": 3.269944429397583,
"learning_rate": 2.7044025157232706e-06,
"loss": 2.6133,
"step": 87
},
{
"epoch": 0.332703213610586,
"grad_norm": 3.377781867980957,
"learning_rate": 2.7358490566037738e-06,
"loss": 2.555,
"step": 88
},
{
"epoch": 0.33648393194706994,
"grad_norm": 3.3409223556518555,
"learning_rate": 2.767295597484277e-06,
"loss": 2.4417,
"step": 89
},
{
"epoch": 0.34026465028355385,
"grad_norm": 3.4132015705108643,
"learning_rate": 2.79874213836478e-06,
"loss": 2.3684,
"step": 90
},
{
"epoch": 0.3440453686200378,
"grad_norm": 3.278552770614624,
"learning_rate": 2.830188679245283e-06,
"loss": 2.2846,
"step": 91
},
{
"epoch": 0.34782608695652173,
"grad_norm": 3.3326900005340576,
"learning_rate": 2.861635220125786e-06,
"loss": 2.2018,
"step": 92
},
{
"epoch": 0.3516068052930057,
"grad_norm": 3.2643816471099854,
"learning_rate": 2.8930817610062893e-06,
"loss": 2.1349,
"step": 93
},
{
"epoch": 0.3553875236294896,
"grad_norm": 3.1792874336242676,
"learning_rate": 2.9245283018867924e-06,
"loss": 2.0337,
"step": 94
},
{
"epoch": 0.3591682419659735,
"grad_norm": 3.205648183822632,
"learning_rate": 2.9559748427672955e-06,
"loss": 1.9311,
"step": 95
},
{
"epoch": 0.3629489603024575,
"grad_norm": 3.1705102920532227,
"learning_rate": 2.987421383647799e-06,
"loss": 1.8769,
"step": 96
},
{
"epoch": 0.3667296786389414,
"grad_norm": 3.181612491607666,
"learning_rate": 3.018867924528302e-06,
"loss": 1.77,
"step": 97
},
{
"epoch": 0.3705103969754253,
"grad_norm": 3.1688497066497803,
"learning_rate": 3.0503144654088053e-06,
"loss": 1.6855,
"step": 98
},
{
"epoch": 0.3742911153119093,
"grad_norm": 3.070916175842285,
"learning_rate": 3.0817610062893084e-06,
"loss": 1.5951,
"step": 99
},
{
"epoch": 0.3780718336483932,
"grad_norm": 3.0769195556640625,
"learning_rate": 3.1132075471698115e-06,
"loss": 1.5228,
"step": 100
},
{
"epoch": 0.3818525519848771,
"grad_norm": 3.046743392944336,
"learning_rate": 3.1446540880503146e-06,
"loss": 1.4414,
"step": 101
},
{
"epoch": 0.3856332703213611,
"grad_norm": 3.2690372467041016,
"learning_rate": 3.1761006289308178e-06,
"loss": 1.3764,
"step": 102
},
{
"epoch": 0.389413988657845,
"grad_norm": 3.578078031539917,
"learning_rate": 3.207547169811321e-06,
"loss": 1.2999,
"step": 103
},
{
"epoch": 0.3931947069943289,
"grad_norm": 2.790632486343384,
"learning_rate": 3.238993710691824e-06,
"loss": 1.2404,
"step": 104
},
{
"epoch": 0.39697542533081287,
"grad_norm": 3.452138662338257,
"learning_rate": 3.270440251572327e-06,
"loss": 1.1618,
"step": 105
},
{
"epoch": 0.4007561436672968,
"grad_norm": 2.630812883377075,
"learning_rate": 3.30188679245283e-06,
"loss": 1.0705,
"step": 106
},
{
"epoch": 0.4045368620037807,
"grad_norm": 2.906174421310425,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.0143,
"step": 107
},
{
"epoch": 0.40831758034026466,
"grad_norm": 2.599710702896118,
"learning_rate": 3.3647798742138364e-06,
"loss": 0.9305,
"step": 108
},
{
"epoch": 0.4120982986767486,
"grad_norm": 2.5494208335876465,
"learning_rate": 3.3962264150943395e-06,
"loss": 0.8758,
"step": 109
},
{
"epoch": 0.4158790170132325,
"grad_norm": 2.4350202083587646,
"learning_rate": 3.4276729559748427e-06,
"loss": 0.8038,
"step": 110
},
{
"epoch": 0.41965973534971646,
"grad_norm": 2.3131184577941895,
"learning_rate": 3.4591194968553458e-06,
"loss": 0.7518,
"step": 111
},
{
"epoch": 0.42344045368620037,
"grad_norm": 2.2873146533966064,
"learning_rate": 3.4905660377358493e-06,
"loss": 0.6975,
"step": 112
},
{
"epoch": 0.42722117202268434,
"grad_norm": 2.0021121501922607,
"learning_rate": 3.5220125786163524e-06,
"loss": 0.6393,
"step": 113
},
{
"epoch": 0.43100189035916825,
"grad_norm": 1.9343374967575073,
"learning_rate": 3.5534591194968555e-06,
"loss": 0.6132,
"step": 114
},
{
"epoch": 0.43478260869565216,
"grad_norm": 2.566892147064209,
"learning_rate": 3.5849056603773586e-06,
"loss": 0.5614,
"step": 115
},
{
"epoch": 0.43856332703213613,
"grad_norm": 1.834770679473877,
"learning_rate": 3.6163522012578618e-06,
"loss": 0.5186,
"step": 116
},
{
"epoch": 0.44234404536862004,
"grad_norm": 1.8110265731811523,
"learning_rate": 3.647798742138365e-06,
"loss": 0.4833,
"step": 117
},
{
"epoch": 0.44612476370510395,
"grad_norm": 1.7480475902557373,
"learning_rate": 3.679245283018868e-06,
"loss": 0.4449,
"step": 118
},
{
"epoch": 0.4499054820415879,
"grad_norm": 1.978847861289978,
"learning_rate": 3.710691823899371e-06,
"loss": 0.4241,
"step": 119
},
{
"epoch": 0.45368620037807184,
"grad_norm": 1.6734317541122437,
"learning_rate": 3.742138364779874e-06,
"loss": 0.3884,
"step": 120
},
{
"epoch": 0.45746691871455575,
"grad_norm": 1.6196964979171753,
"learning_rate": 3.7735849056603777e-06,
"loss": 0.3537,
"step": 121
},
{
"epoch": 0.4612476370510397,
"grad_norm": 1.6024487018585205,
"learning_rate": 3.8050314465408813e-06,
"loss": 0.3275,
"step": 122
},
{
"epoch": 0.46502835538752363,
"grad_norm": 1.6706016063690186,
"learning_rate": 3.836477987421384e-06,
"loss": 0.2987,
"step": 123
},
{
"epoch": 0.46880907372400754,
"grad_norm": 1.461807370185852,
"learning_rate": 3.8679245283018875e-06,
"loss": 0.2756,
"step": 124
},
{
"epoch": 0.4725897920604915,
"grad_norm": 1.795247197151184,
"learning_rate": 3.89937106918239e-06,
"loss": 0.2272,
"step": 125
},
{
"epoch": 0.4763705103969754,
"grad_norm": 1.149410367012024,
"learning_rate": 3.930817610062894e-06,
"loss": 0.2106,
"step": 126
},
{
"epoch": 0.48015122873345933,
"grad_norm": 1.4200841188430786,
"learning_rate": 3.962264150943396e-06,
"loss": 0.1888,
"step": 127
},
{
"epoch": 0.4839319470699433,
"grad_norm": 1.2156733274459839,
"learning_rate": 3.9937106918239e-06,
"loss": 0.1695,
"step": 128
},
{
"epoch": 0.4877126654064272,
"grad_norm": 0.9911368489265442,
"learning_rate": 4.025157232704403e-06,
"loss": 0.1501,
"step": 129
},
{
"epoch": 0.4914933837429111,
"grad_norm": 0.7694626450538635,
"learning_rate": 4.056603773584906e-06,
"loss": 0.1289,
"step": 130
},
{
"epoch": 0.4952741020793951,
"grad_norm": 0.7689205408096313,
"learning_rate": 4.08805031446541e-06,
"loss": 0.1184,
"step": 131
},
{
"epoch": 0.499054820415879,
"grad_norm": 0.9590215086936951,
"learning_rate": 4.119496855345912e-06,
"loss": 0.0999,
"step": 132
},
{
"epoch": 0.502835538752363,
"grad_norm": 0.6190182566642761,
"learning_rate": 4.150943396226416e-06,
"loss": 0.091,
"step": 133
},
{
"epoch": 0.5066162570888468,
"grad_norm": 0.8857131004333496,
"learning_rate": 4.182389937106919e-06,
"loss": 0.08,
"step": 134
},
{
"epoch": 0.5103969754253308,
"grad_norm": 0.6522263884544373,
"learning_rate": 4.213836477987422e-06,
"loss": 0.0709,
"step": 135
},
{
"epoch": 0.5141776937618148,
"grad_norm": 0.6781020760536194,
"learning_rate": 4.245283018867925e-06,
"loss": 0.079,
"step": 136
},
{
"epoch": 0.5179584120982986,
"grad_norm": 0.6817317008972168,
"learning_rate": 4.276729559748428e-06,
"loss": 0.056,
"step": 137
},
{
"epoch": 0.5217391304347826,
"grad_norm": 0.5343799591064453,
"learning_rate": 4.308176100628931e-06,
"loss": 0.0518,
"step": 138
},
{
"epoch": 0.5255198487712666,
"grad_norm": 0.4243936836719513,
"learning_rate": 4.339622641509435e-06,
"loss": 0.0312,
"step": 139
},
{
"epoch": 0.5293005671077504,
"grad_norm": 0.4387669265270233,
"learning_rate": 4.371069182389937e-06,
"loss": 0.0305,
"step": 140
},
{
"epoch": 0.5330812854442344,
"grad_norm": 0.48877736926078796,
"learning_rate": 4.402515723270441e-06,
"loss": 0.0287,
"step": 141
},
{
"epoch": 0.5368620037807184,
"grad_norm": 0.3683076500892639,
"learning_rate": 4.4339622641509435e-06,
"loss": 0.0227,
"step": 142
},
{
"epoch": 0.5406427221172023,
"grad_norm": 0.2861018478870392,
"learning_rate": 4.465408805031447e-06,
"loss": 0.0222,
"step": 143
},
{
"epoch": 0.5444234404536862,
"grad_norm": 0.22285304963588715,
"learning_rate": 4.49685534591195e-06,
"loss": 0.0205,
"step": 144
},
{
"epoch": 0.5482041587901701,
"grad_norm": 0.2912003695964813,
"learning_rate": 4.528301886792453e-06,
"loss": 0.0183,
"step": 145
},
{
"epoch": 0.5519848771266541,
"grad_norm": 0.2376680076122284,
"learning_rate": 4.559748427672957e-06,
"loss": 0.0178,
"step": 146
},
{
"epoch": 0.555765595463138,
"grad_norm": 0.19624432921409607,
"learning_rate": 4.5911949685534595e-06,
"loss": 0.0187,
"step": 147
},
{
"epoch": 0.5595463137996219,
"grad_norm": 0.2549561560153961,
"learning_rate": 4.622641509433963e-06,
"loss": 0.017,
"step": 148
},
{
"epoch": 0.5633270321361059,
"grad_norm": 0.18909113109111786,
"learning_rate": 4.654088050314466e-06,
"loss": 0.0203,
"step": 149
},
{
"epoch": 0.5671077504725898,
"grad_norm": 0.39210933446884155,
"learning_rate": 4.685534591194969e-06,
"loss": 0.0177,
"step": 150
},
{
"epoch": 0.5708884688090737,
"grad_norm": 0.29460176825523376,
"learning_rate": 4.716981132075472e-06,
"loss": 0.0174,
"step": 151
},
{
"epoch": 0.5746691871455577,
"grad_norm": 0.15810728073120117,
"learning_rate": 4.7484276729559755e-06,
"loss": 0.0158,
"step": 152
},
{
"epoch": 0.5784499054820416,
"grad_norm": 0.19970139861106873,
"learning_rate": 4.779874213836478e-06,
"loss": 0.0187,
"step": 153
},
{
"epoch": 0.5822306238185255,
"grad_norm": 0.20323728024959564,
"learning_rate": 4.811320754716982e-06,
"loss": 0.0195,
"step": 154
},
{
"epoch": 0.5860113421550095,
"grad_norm": 0.20774324238300323,
"learning_rate": 4.842767295597484e-06,
"loss": 0.0153,
"step": 155
},
{
"epoch": 0.5897920604914934,
"grad_norm": 0.5454590320587158,
"learning_rate": 4.874213836477988e-06,
"loss": 0.0161,
"step": 156
},
{
"epoch": 0.5935727788279773,
"grad_norm": 0.19222363829612732,
"learning_rate": 4.905660377358491e-06,
"loss": 0.0139,
"step": 157
},
{
"epoch": 0.5973534971644613,
"grad_norm": 0.1607956439256668,
"learning_rate": 4.937106918238994e-06,
"loss": 0.0139,
"step": 158
},
{
"epoch": 0.6011342155009451,
"grad_norm": 0.19767163693904877,
"learning_rate": 4.968553459119497e-06,
"loss": 0.0132,
"step": 159
},
{
"epoch": 0.6049149338374291,
"grad_norm": 0.3366517424583435,
"learning_rate": 5e-06,
"loss": 0.0193,
"step": 160
},
{
"epoch": 0.6086956521739131,
"grad_norm": 0.14002861082553864,
"learning_rate": 4.999993975370883e-06,
"loss": 0.0127,
"step": 161
},
{
"epoch": 0.6124763705103969,
"grad_norm": 0.17500659823417664,
"learning_rate": 4.999975901512567e-06,
"loss": 0.0141,
"step": 162
},
{
"epoch": 0.6162570888468809,
"grad_norm": 0.1498129516839981,
"learning_rate": 4.999945778512164e-06,
"loss": 0.0142,
"step": 163
},
{
"epoch": 0.6200378071833649,
"grad_norm": 0.1581406444311142,
"learning_rate": 4.999903606514858e-06,
"loss": 0.0143,
"step": 164
},
{
"epoch": 0.6238185255198487,
"grad_norm": 0.30295559763908386,
"learning_rate": 4.999849385723904e-06,
"loss": 0.0132,
"step": 165
},
{
"epoch": 0.6275992438563327,
"grad_norm": 0.3571516275405884,
"learning_rate": 4.9997831164006315e-06,
"loss": 0.0144,
"step": 166
},
{
"epoch": 0.6313799621928167,
"grad_norm": 0.3662567138671875,
"learning_rate": 4.999704798864439e-06,
"loss": 0.0138,
"step": 167
},
{
"epoch": 0.6351606805293005,
"grad_norm": 0.16797330975532532,
"learning_rate": 4.9996144334927924e-06,
"loss": 0.0143,
"step": 168
},
{
"epoch": 0.6389413988657845,
"grad_norm": 0.14023132622241974,
"learning_rate": 4.9995120207212275e-06,
"loss": 0.013,
"step": 169
},
{
"epoch": 0.6427221172022685,
"grad_norm": 0.20948678255081177,
"learning_rate": 4.999397561043343e-06,
"loss": 0.0124,
"step": 170
},
{
"epoch": 0.6465028355387523,
"grad_norm": 0.2124815732240677,
"learning_rate": 4.9992710550108e-06,
"loss": 0.0137,
"step": 171
},
{
"epoch": 0.6502835538752363,
"grad_norm": 0.2177082598209381,
"learning_rate": 4.999132503233321e-06,
"loss": 0.0151,
"step": 172
},
{
"epoch": 0.6540642722117203,
"grad_norm": 0.271586149930954,
"learning_rate": 4.998981906378684e-06,
"loss": 0.0127,
"step": 173
},
{
"epoch": 0.6578449905482041,
"grad_norm": 0.36368614435195923,
"learning_rate": 4.998819265172721e-06,
"loss": 0.0139,
"step": 174
},
{
"epoch": 0.6616257088846881,
"grad_norm": 0.14393746852874756,
"learning_rate": 4.9986445803993144e-06,
"loss": 0.0122,
"step": 175
},
{
"epoch": 0.665406427221172,
"grad_norm": 0.21265114843845367,
"learning_rate": 4.998457852900393e-06,
"loss": 0.0173,
"step": 176
},
{
"epoch": 0.6691871455576559,
"grad_norm": 0.3905700743198395,
"learning_rate": 4.998259083575929e-06,
"loss": 0.0165,
"step": 177
},
{
"epoch": 0.6729678638941399,
"grad_norm": 0.240561380982399,
"learning_rate": 4.99804827338393e-06,
"loss": 0.0115,
"step": 178
},
{
"epoch": 0.6767485822306238,
"grad_norm": 0.13536234200000763,
"learning_rate": 4.997825423340439e-06,
"loss": 0.0106,
"step": 179
},
{
"epoch": 0.6805293005671077,
"grad_norm": 0.5081303119659424,
"learning_rate": 4.997590534519528e-06,
"loss": 0.012,
"step": 180
},
{
"epoch": 0.6843100189035917,
"grad_norm": 0.15104596316814423,
"learning_rate": 4.99734360805329e-06,
"loss": 0.013,
"step": 181
},
{
"epoch": 0.6880907372400756,
"grad_norm": 0.20405662059783936,
"learning_rate": 4.997084645131837e-06,
"loss": 0.0116,
"step": 182
},
{
"epoch": 0.6918714555765595,
"grad_norm": 0.2502865493297577,
"learning_rate": 4.996813647003296e-06,
"loss": 0.0113,
"step": 183
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.20884762704372406,
"learning_rate": 4.9965306149737955e-06,
"loss": 0.0107,
"step": 184
},
{
"epoch": 0.6994328922495274,
"grad_norm": 0.20507586002349854,
"learning_rate": 4.996235550407467e-06,
"loss": 0.0144,
"step": 185
},
{
"epoch": 0.7032136105860114,
"grad_norm": 0.29745373129844666,
"learning_rate": 4.995928454726433e-06,
"loss": 0.0142,
"step": 186
},
{
"epoch": 0.7069943289224953,
"grad_norm": 0.2321498990058899,
"learning_rate": 4.9956093294108046e-06,
"loss": 0.0134,
"step": 187
},
{
"epoch": 0.7107750472589792,
"grad_norm": 0.35312822461128235,
"learning_rate": 4.995278175998669e-06,
"loss": 0.0117,
"step": 188
},
{
"epoch": 0.7145557655954632,
"grad_norm": 0.3376014530658722,
"learning_rate": 4.994934996086091e-06,
"loss": 0.0139,
"step": 189
},
{
"epoch": 0.718336483931947,
"grad_norm": 0.7203236818313599,
"learning_rate": 4.994579791327093e-06,
"loss": 0.0133,
"step": 190
},
{
"epoch": 0.722117202268431,
"grad_norm": 0.15841040015220642,
"learning_rate": 4.994212563433658e-06,
"loss": 0.0138,
"step": 191
},
{
"epoch": 0.725897920604915,
"grad_norm": 0.27243638038635254,
"learning_rate": 4.993833314175716e-06,
"loss": 0.0123,
"step": 192
},
{
"epoch": 0.7296786389413988,
"grad_norm": 0.20597811043262482,
"learning_rate": 4.993442045381133e-06,
"loss": 0.0108,
"step": 193
},
{
"epoch": 0.7334593572778828,
"grad_norm": 0.18340808153152466,
"learning_rate": 4.993038758935712e-06,
"loss": 0.0105,
"step": 194
},
{
"epoch": 0.7372400756143668,
"grad_norm": 0.11817918717861176,
"learning_rate": 4.9926234567831716e-06,
"loss": 0.0095,
"step": 195
},
{
"epoch": 0.7410207939508506,
"grad_norm": 0.2828313708305359,
"learning_rate": 4.9921961409251465e-06,
"loss": 0.0149,
"step": 196
},
{
"epoch": 0.7448015122873346,
"grad_norm": 0.28721073269844055,
"learning_rate": 4.991756813421171e-06,
"loss": 0.0102,
"step": 197
},
{
"epoch": 0.7485822306238186,
"grad_norm": 0.20243827998638153,
"learning_rate": 4.991305476388674e-06,
"loss": 0.0114,
"step": 198
},
{
"epoch": 0.7523629489603024,
"grad_norm": 0.2657279074192047,
"learning_rate": 4.990842132002967e-06,
"loss": 0.009,
"step": 199
},
{
"epoch": 0.7561436672967864,
"grad_norm": 0.16293185949325562,
"learning_rate": 4.99036678249723e-06,
"loss": 0.0108,
"step": 200
},
{
"epoch": 0.7599243856332704,
"grad_norm": 0.24433821439743042,
"learning_rate": 4.98987943016251e-06,
"loss": 0.0115,
"step": 201
},
{
"epoch": 0.7637051039697542,
"grad_norm": 0.13836057484149933,
"learning_rate": 4.9893800773476965e-06,
"loss": 0.0095,
"step": 202
},
{
"epoch": 0.7674858223062382,
"grad_norm": 0.25805094838142395,
"learning_rate": 4.988868726459526e-06,
"loss": 0.0142,
"step": 203
},
{
"epoch": 0.7712665406427222,
"grad_norm": 0.1681707352399826,
"learning_rate": 4.988345379962556e-06,
"loss": 0.0094,
"step": 204
},
{
"epoch": 0.775047258979206,
"grad_norm": 0.5535834431648254,
"learning_rate": 4.987810040379161e-06,
"loss": 0.0093,
"step": 205
},
{
"epoch": 0.77882797731569,
"grad_norm": 0.2519840896129608,
"learning_rate": 4.987262710289519e-06,
"loss": 0.0131,
"step": 206
},
{
"epoch": 0.782608695652174,
"grad_norm": 0.18629945814609528,
"learning_rate": 4.9867033923316e-06,
"loss": 0.0086,
"step": 207
},
{
"epoch": 0.7863894139886578,
"grad_norm": 0.40064531564712524,
"learning_rate": 4.98613208920115e-06,
"loss": 0.0126,
"step": 208
},
{
"epoch": 0.7901701323251418,
"grad_norm": 0.5355131030082703,
"learning_rate": 4.98554880365168e-06,
"loss": 0.0105,
"step": 209
},
{
"epoch": 0.7939508506616257,
"grad_norm": 0.2643525302410126,
"learning_rate": 4.984953538494453e-06,
"loss": 0.0113,
"step": 210
},
{
"epoch": 0.7977315689981096,
"grad_norm": 0.21430149674415588,
"learning_rate": 4.984346296598472e-06,
"loss": 0.0092,
"step": 211
},
{
"epoch": 0.8015122873345936,
"grad_norm": 0.15202465653419495,
"learning_rate": 4.983727080890461e-06,
"loss": 0.0082,
"step": 212
},
{
"epoch": 0.8052930056710775,
"grad_norm": 0.12687154114246368,
"learning_rate": 4.983095894354858e-06,
"loss": 0.009,
"step": 213
},
{
"epoch": 0.8090737240075614,
"grad_norm": 0.3270910978317261,
"learning_rate": 4.982452740033793e-06,
"loss": 0.0136,
"step": 214
},
{
"epoch": 0.8128544423440454,
"grad_norm": 0.20315034687519073,
"learning_rate": 4.981797621027079e-06,
"loss": 0.011,
"step": 215
},
{
"epoch": 0.8166351606805293,
"grad_norm": 0.31483131647109985,
"learning_rate": 4.981130540492197e-06,
"loss": 0.012,
"step": 216
},
{
"epoch": 0.8204158790170132,
"grad_norm": 0.22836464643478394,
"learning_rate": 4.9804515016442754e-06,
"loss": 0.0094,
"step": 217
},
{
"epoch": 0.8241965973534972,
"grad_norm": 0.24444779753684998,
"learning_rate": 4.97976050775608e-06,
"loss": 0.0092,
"step": 218
},
{
"epoch": 0.8279773156899811,
"grad_norm": 0.29489508271217346,
"learning_rate": 4.979057562157998e-06,
"loss": 0.0086,
"step": 219
},
{
"epoch": 0.831758034026465,
"grad_norm": 0.40103089809417725,
"learning_rate": 4.978342668238018e-06,
"loss": 0.0106,
"step": 220
},
{
"epoch": 0.8355387523629489,
"grad_norm": 0.3467403054237366,
"learning_rate": 4.977615829441716e-06,
"loss": 0.0097,
"step": 221
},
{
"epoch": 0.8393194706994329,
"grad_norm": 0.3331691324710846,
"learning_rate": 4.97687704927224e-06,
"loss": 0.0114,
"step": 222
},
{
"epoch": 0.8431001890359168,
"grad_norm": 0.16955935955047607,
"learning_rate": 4.97612633129029e-06,
"loss": 0.008,
"step": 223
},
{
"epoch": 0.8468809073724007,
"grad_norm": 0.29774612188339233,
"learning_rate": 4.975363679114104e-06,
"loss": 0.0093,
"step": 224
},
{
"epoch": 0.8506616257088847,
"grad_norm": 0.1812770813703537,
"learning_rate": 4.974589096419441e-06,
"loss": 0.0078,
"step": 225
},
{
"epoch": 0.8544423440453687,
"grad_norm": 0.27868548035621643,
"learning_rate": 4.973802586939558e-06,
"loss": 0.0119,
"step": 226
},
{
"epoch": 0.8582230623818525,
"grad_norm": 1.3429784774780273,
"learning_rate": 4.973004154465198e-06,
"loss": 0.0104,
"step": 227
},
{
"epoch": 0.8620037807183365,
"grad_norm": 0.18361057341098785,
"learning_rate": 4.9721938028445695e-06,
"loss": 0.0081,
"step": 228
},
{
"epoch": 0.8657844990548205,
"grad_norm": 0.1629960983991623,
"learning_rate": 4.971371535983325e-06,
"loss": 0.0095,
"step": 229
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.21906869113445282,
"learning_rate": 4.970537357844549e-06,
"loss": 0.0118,
"step": 230
},
{
"epoch": 0.8733459357277883,
"grad_norm": 0.32470542192459106,
"learning_rate": 4.96969127244873e-06,
"loss": 0.0075,
"step": 231
},
{
"epoch": 0.8771266540642723,
"grad_norm": 0.19773821532726288,
"learning_rate": 4.96883328387375e-06,
"loss": 0.008,
"step": 232
},
{
"epoch": 0.8809073724007561,
"grad_norm": 0.764606237411499,
"learning_rate": 4.9679633962548615e-06,
"loss": 0.0098,
"step": 233
},
{
"epoch": 0.8846880907372401,
"grad_norm": 0.35291630029678345,
"learning_rate": 4.96708161378466e-06,
"loss": 0.0092,
"step": 234
},
{
"epoch": 0.888468809073724,
"grad_norm": 0.3802095055580139,
"learning_rate": 4.96618794071308e-06,
"loss": 0.009,
"step": 235
},
{
"epoch": 0.8922495274102079,
"grad_norm": 0.22711920738220215,
"learning_rate": 4.965282381347357e-06,
"loss": 0.0102,
"step": 236
},
{
"epoch": 0.8960302457466919,
"grad_norm": 0.29683762788772583,
"learning_rate": 4.964364940052021e-06,
"loss": 0.0086,
"step": 237
},
{
"epoch": 0.8998109640831758,
"grad_norm": 0.3059028685092926,
"learning_rate": 4.963435621248865e-06,
"loss": 0.0139,
"step": 238
},
{
"epoch": 0.9035916824196597,
"grad_norm": 0.29774123430252075,
"learning_rate": 4.962494429416931e-06,
"loss": 0.009,
"step": 239
},
{
"epoch": 0.9073724007561437,
"grad_norm": 0.409982830286026,
"learning_rate": 4.9615413690924844e-06,
"loss": 0.0095,
"step": 240
},
{
"epoch": 0.9111531190926276,
"grad_norm": 0.18289801478385925,
"learning_rate": 4.960576444868992e-06,
"loss": 0.0073,
"step": 241
},
{
"epoch": 0.9149338374291115,
"grad_norm": 0.8128044605255127,
"learning_rate": 4.9595996613971046e-06,
"loss": 0.0159,
"step": 242
},
{
"epoch": 0.9187145557655955,
"grad_norm": 0.2674903869628906,
"learning_rate": 4.958611023384626e-06,
"loss": 0.0098,
"step": 243
},
{
"epoch": 0.9224952741020794,
"grad_norm": 0.18680743873119354,
"learning_rate": 4.9576105355965e-06,
"loss": 0.0084,
"step": 244
},
{
"epoch": 0.9262759924385633,
"grad_norm": 0.3209806978702545,
"learning_rate": 4.956598202854778e-06,
"loss": 0.0087,
"step": 245
},
{
"epoch": 0.9300567107750473,
"grad_norm": 0.20674099028110504,
"learning_rate": 4.955574030038608e-06,
"loss": 0.0062,
"step": 246
},
{
"epoch": 0.9338374291115312,
"grad_norm": 0.2949323356151581,
"learning_rate": 4.954538022084197e-06,
"loss": 0.0077,
"step": 247
},
{
"epoch": 0.9376181474480151,
"grad_norm": 0.16267140209674835,
"learning_rate": 4.953490183984795e-06,
"loss": 0.0064,
"step": 248
},
{
"epoch": 0.941398865784499,
"grad_norm": 0.30251970887184143,
"learning_rate": 4.9524305207906705e-06,
"loss": 0.0071,
"step": 249
},
{
"epoch": 0.945179584120983,
"grad_norm": 0.2351951152086258,
"learning_rate": 4.951359037609088e-06,
"loss": 0.008,
"step": 250
},
{
"epoch": 0.9489603024574669,
"grad_norm": 0.32132813334465027,
"learning_rate": 4.950275739604277e-06,
"loss": 0.0076,
"step": 251
},
{
"epoch": 0.9527410207939508,
"grad_norm": 0.2724415957927704,
"learning_rate": 4.9491806319974115e-06,
"loss": 0.0074,
"step": 252
},
{
"epoch": 0.9565217391304348,
"grad_norm": 0.30280518531799316,
"learning_rate": 4.948073720066587e-06,
"loss": 0.0078,
"step": 253
},
{
"epoch": 0.9603024574669187,
"grad_norm": 0.3232327401638031,
"learning_rate": 4.94695500914679e-06,
"loss": 0.0085,
"step": 254
},
{
"epoch": 0.9640831758034026,
"grad_norm": 0.2186642438173294,
"learning_rate": 4.945824504629875e-06,
"loss": 0.0067,
"step": 255
},
{
"epoch": 0.9678638941398866,
"grad_norm": 0.3736853003501892,
"learning_rate": 4.944682211964538e-06,
"loss": 0.0134,
"step": 256
},
{
"epoch": 0.9716446124763705,
"grad_norm": 0.12058785557746887,
"learning_rate": 4.943528136656292e-06,
"loss": 0.0046,
"step": 257
},
{
"epoch": 0.9754253308128544,
"grad_norm": 0.2973642647266388,
"learning_rate": 4.9423622842674366e-06,
"loss": 0.0051,
"step": 258
},
{
"epoch": 0.9792060491493384,
"grad_norm": 0.48048704862594604,
"learning_rate": 4.941184660417034e-06,
"loss": 0.0093,
"step": 259
},
{
"epoch": 0.9829867674858223,
"grad_norm": 0.2897495627403259,
"learning_rate": 4.939995270780883e-06,
"loss": 0.0099,
"step": 260
},
{
"epoch": 0.9867674858223062,
"grad_norm": 0.3390437662601471,
"learning_rate": 4.938794121091488e-06,
"loss": 0.0127,
"step": 261
},
{
"epoch": 0.9905482041587902,
"grad_norm": 0.18855001032352448,
"learning_rate": 4.937581217138033e-06,
"loss": 0.0058,
"step": 262
},
{
"epoch": 0.994328922495274,
"grad_norm": 0.2813539206981659,
"learning_rate": 4.936356564766358e-06,
"loss": 0.0056,
"step": 263
},
{
"epoch": 0.998109640831758,
"grad_norm": 0.21155141294002533,
"learning_rate": 4.935120169878922e-06,
"loss": 0.0062,
"step": 264
},
{
"epoch": 1.0,
"grad_norm": 0.3886541426181793,
"learning_rate": 4.933872038434782e-06,
"loss": 0.0085,
"step": 265
},
{
"epoch": 1.003780718336484,
"grad_norm": 0.35919421911239624,
"learning_rate": 4.93261217644956e-06,
"loss": 0.0071,
"step": 266
},
{
"epoch": 1.007561436672968,
"grad_norm": 0.1540549099445343,
"learning_rate": 4.931340589995419e-06,
"loss": 0.0064,
"step": 267
},
{
"epoch": 1.011342155009452,
"grad_norm": 0.2220161110162735,
"learning_rate": 4.930057285201028e-06,
"loss": 0.0064,
"step": 268
},
{
"epoch": 1.0151228733459357,
"grad_norm": 0.3442586362361908,
"learning_rate": 4.928762268251533e-06,
"loss": 0.0106,
"step": 269
},
{
"epoch": 1.0189035916824196,
"grad_norm": 0.19892996549606323,
"learning_rate": 4.9274555453885355e-06,
"loss": 0.0052,
"step": 270
},
{
"epoch": 1.0226843100189036,
"grad_norm": 0.27407875657081604,
"learning_rate": 4.926137122910049e-06,
"loss": 0.01,
"step": 271
},
{
"epoch": 1.0264650283553876,
"grad_norm": 0.5205274224281311,
"learning_rate": 4.924807007170479e-06,
"loss": 0.0053,
"step": 272
},
{
"epoch": 1.0302457466918715,
"grad_norm": 0.1610601395368576,
"learning_rate": 4.923465204580589e-06,
"loss": 0.0047,
"step": 273
},
{
"epoch": 1.0340264650283555,
"grad_norm": 0.33532628417015076,
"learning_rate": 4.922111721607471e-06,
"loss": 0.0083,
"step": 274
},
{
"epoch": 1.0378071833648392,
"grad_norm": 0.19273239374160767,
"learning_rate": 4.920746564774509e-06,
"loss": 0.0078,
"step": 275
},
{
"epoch": 1.0415879017013232,
"grad_norm": 0.5411931276321411,
"learning_rate": 4.9193697406613555e-06,
"loss": 0.0081,
"step": 276
},
{
"epoch": 1.0453686200378072,
"grad_norm": 0.40823793411254883,
"learning_rate": 4.917981255903893e-06,
"loss": 0.006,
"step": 277
},
{
"epoch": 1.0491493383742911,
"grad_norm": 0.20618820190429688,
"learning_rate": 4.9165811171942065e-06,
"loss": 0.0061,
"step": 278
},
{
"epoch": 1.052930056710775,
"grad_norm": 0.22818367183208466,
"learning_rate": 4.915169331280549e-06,
"loss": 0.007,
"step": 279
},
{
"epoch": 1.056710775047259,
"grad_norm": 1.1232956647872925,
"learning_rate": 4.9137459049673105e-06,
"loss": 0.0119,
"step": 280
},
{
"epoch": 1.0604914933837428,
"grad_norm": 0.24216459691524506,
"learning_rate": 4.912310845114983e-06,
"loss": 0.0058,
"step": 281
},
{
"epoch": 1.0642722117202268,
"grad_norm": 0.2712610960006714,
"learning_rate": 4.910864158640129e-06,
"loss": 0.0055,
"step": 282
},
{
"epoch": 1.0680529300567108,
"grad_norm": 0.2083127498626709,
"learning_rate": 4.909405852515348e-06,
"loss": 0.0041,
"step": 283
},
{
"epoch": 1.0718336483931947,
"grad_norm": 0.24266625940799713,
"learning_rate": 4.907935933769243e-06,
"loss": 0.0045,
"step": 284
},
{
"epoch": 1.0756143667296787,
"grad_norm": 0.3229181468486786,
"learning_rate": 4.9064544094863854e-06,
"loss": 0.0037,
"step": 285
},
{
"epoch": 1.0793950850661627,
"grad_norm": 0.3141697347164154,
"learning_rate": 4.904961286807285e-06,
"loss": 0.0061,
"step": 286
},
{
"epoch": 1.0831758034026464,
"grad_norm": 0.4243822693824768,
"learning_rate": 4.903456572928348e-06,
"loss": 0.0088,
"step": 287
},
{
"epoch": 1.0869565217391304,
"grad_norm": 0.3038384020328522,
"learning_rate": 4.90194027510185e-06,
"loss": 0.0043,
"step": 288
},
{
"epoch": 1.0907372400756143,
"grad_norm": 0.3872905373573303,
"learning_rate": 4.900412400635895e-06,
"loss": 0.0084,
"step": 289
},
{
"epoch": 1.0945179584120983,
"grad_norm": 0.2568505108356476,
"learning_rate": 4.898872956894386e-06,
"loss": 0.0048,
"step": 290
},
{
"epoch": 1.0982986767485823,
"grad_norm": 0.32108235359191895,
"learning_rate": 4.897321951296985e-06,
"loss": 0.0109,
"step": 291
},
{
"epoch": 1.1020793950850662,
"grad_norm": 0.224013552069664,
"learning_rate": 4.8957593913190785e-06,
"loss": 0.0052,
"step": 292
},
{
"epoch": 1.10586011342155,
"grad_norm": 0.1713561862707138,
"learning_rate": 4.894185284491742e-06,
"loss": 0.0056,
"step": 293
},
{
"epoch": 1.109640831758034,
"grad_norm": 0.19997638463974,
"learning_rate": 4.892599638401704e-06,
"loss": 0.0037,
"step": 294
},
{
"epoch": 1.113421550094518,
"grad_norm": 0.3231566250324249,
"learning_rate": 4.891002460691306e-06,
"loss": 0.0087,
"step": 295
},
{
"epoch": 1.117202268431002,
"grad_norm": 0.4332091808319092,
"learning_rate": 4.889393759058474e-06,
"loss": 0.0121,
"step": 296
},
{
"epoch": 1.1209829867674859,
"grad_norm": 0.30564525723457336,
"learning_rate": 4.887773541256669e-06,
"loss": 0.0103,
"step": 297
},
{
"epoch": 1.1247637051039698,
"grad_norm": 0.5109544992446899,
"learning_rate": 4.886141815094864e-06,
"loss": 0.0048,
"step": 298
},
{
"epoch": 1.1285444234404536,
"grad_norm": 0.27209943532943726,
"learning_rate": 4.884498588437491e-06,
"loss": 0.0054,
"step": 299
},
{
"epoch": 1.1323251417769375,
"grad_norm": 0.21001404523849487,
"learning_rate": 4.882843869204417e-06,
"loss": 0.0031,
"step": 300
},
{
"epoch": 1.1361058601134215,
"grad_norm": 0.16830381751060486,
"learning_rate": 4.8811776653708985e-06,
"loss": 0.0044,
"step": 301
},
{
"epoch": 1.1398865784499055,
"grad_norm": 18.487056732177734,
"learning_rate": 4.8794999849675416e-06,
"loss": 0.0052,
"step": 302
},
{
"epoch": 1.1436672967863895,
"grad_norm": 0.19815954566001892,
"learning_rate": 4.877810836080269e-06,
"loss": 0.0044,
"step": 303
},
{
"epoch": 1.1474480151228734,
"grad_norm": 0.2645842134952545,
"learning_rate": 4.876110226850278e-06,
"loss": 0.0055,
"step": 304
},
{
"epoch": 1.1512287334593574,
"grad_norm": 0.20674429833889008,
"learning_rate": 4.874398165473999e-06,
"loss": 0.0052,
"step": 305
},
{
"epoch": 1.1550094517958411,
"grad_norm": 0.15041235089302063,
"learning_rate": 4.872674660203062e-06,
"loss": 0.0027,
"step": 306
},
{
"epoch": 1.158790170132325,
"grad_norm": 0.31751132011413574,
"learning_rate": 4.870939719344248e-06,
"loss": 0.0049,
"step": 307
},
{
"epoch": 1.162570888468809,
"grad_norm": 0.24077922105789185,
"learning_rate": 4.869193351259459e-06,
"loss": 0.0061,
"step": 308
},
{
"epoch": 1.166351606805293,
"grad_norm": 1.1419602632522583,
"learning_rate": 4.867435564365671e-06,
"loss": 0.0037,
"step": 309
},
{
"epoch": 1.170132325141777,
"grad_norm": 0.19489724934101105,
"learning_rate": 4.8656663671348954e-06,
"loss": 0.0028,
"step": 310
},
{
"epoch": 1.1739130434782608,
"grad_norm": 0.41008174419403076,
"learning_rate": 4.863885768094138e-06,
"loss": 0.0046,
"step": 311
},
{
"epoch": 1.1776937618147447,
"grad_norm": 0.2505297064781189,
"learning_rate": 4.862093775825356e-06,
"loss": 0.0039,
"step": 312
},
{
"epoch": 1.1814744801512287,
"grad_norm": 0.38411465287208557,
"learning_rate": 4.860290398965423e-06,
"loss": 0.0043,
"step": 313
},
{
"epoch": 1.1852551984877127,
"grad_norm": 0.12815220654010773,
"learning_rate": 4.858475646206079e-06,
"loss": 0.0048,
"step": 314
},
{
"epoch": 1.1890359168241966,
"grad_norm": 0.2467423975467682,
"learning_rate": 4.856649526293894e-06,
"loss": 0.0056,
"step": 315
},
{
"epoch": 1.1928166351606806,
"grad_norm": 0.24193055927753448,
"learning_rate": 4.854812048030225e-06,
"loss": 0.0089,
"step": 316
},
{
"epoch": 1.1965973534971646,
"grad_norm": 0.2023049145936966,
"learning_rate": 4.852963220271171e-06,
"loss": 0.0026,
"step": 317
},
{
"epoch": 1.2003780718336483,
"grad_norm": 0.35156261920928955,
"learning_rate": 4.851103051927533e-06,
"loss": 0.0081,
"step": 318
},
{
"epoch": 1.2041587901701323,
"grad_norm": 0.20046557486057281,
"learning_rate": 4.849231551964771e-06,
"loss": 0.0061,
"step": 319
},
{
"epoch": 1.2079395085066162,
"grad_norm": 0.30065521597862244,
"learning_rate": 4.84734872940296e-06,
"loss": 0.0068,
"step": 320
},
{
"epoch": 1.2117202268431002,
"grad_norm": 0.2166394144296646,
"learning_rate": 4.845454593316747e-06,
"loss": 0.0056,
"step": 321
},
{
"epoch": 1.2155009451795842,
"grad_norm": 0.6717196106910706,
"learning_rate": 4.843549152835303e-06,
"loss": 0.0045,
"step": 322
},
{
"epoch": 1.2192816635160681,
"grad_norm": 1.3720797300338745,
"learning_rate": 4.841632417142288e-06,
"loss": 0.0024,
"step": 323
},
{
"epoch": 1.223062381852552,
"grad_norm": 0.15353788435459137,
"learning_rate": 4.839704395475798e-06,
"loss": 0.0037,
"step": 324
},
{
"epoch": 1.2268431001890359,
"grad_norm": 1.0019330978393555,
"learning_rate": 4.837765097128326e-06,
"loss": 0.0076,
"step": 325
},
{
"epoch": 1.2306238185255198,
"grad_norm": 0.4266430735588074,
"learning_rate": 4.8358145314467165e-06,
"loss": 0.0051,
"step": 326
},
{
"epoch": 1.2344045368620038,
"grad_norm": 0.3188490569591522,
"learning_rate": 4.833852707832114e-06,
"loss": 0.0047,
"step": 327
},
{
"epoch": 1.2381852551984878,
"grad_norm": 0.18008936941623688,
"learning_rate": 4.831879635739929e-06,
"loss": 0.0039,
"step": 328
},
{
"epoch": 1.2419659735349717,
"grad_norm": 0.20714521408081055,
"learning_rate": 4.829895324679782e-06,
"loss": 0.0047,
"step": 329
},
{
"epoch": 1.2457466918714555,
"grad_norm": 0.2258540391921997,
"learning_rate": 4.827899784215463e-06,
"loss": 0.003,
"step": 330
},
{
"epoch": 1.2495274102079394,
"grad_norm": 0.3024104833602905,
"learning_rate": 4.825893023964886e-06,
"loss": 0.0031,
"step": 331
},
{
"epoch": 1.2533081285444234,
"grad_norm": 0.10315516591072083,
"learning_rate": 4.823875053600041e-06,
"loss": 0.0019,
"step": 332
},
{
"epoch": 1.2570888468809074,
"grad_norm": 0.43588098883628845,
"learning_rate": 4.821845882846945e-06,
"loss": 0.007,
"step": 333
},
{
"epoch": 1.2608695652173914,
"grad_norm": 0.23908646404743195,
"learning_rate": 4.819805521485599e-06,
"loss": 0.0025,
"step": 334
},
{
"epoch": 1.264650283553875,
"grad_norm": 0.28753727674484253,
"learning_rate": 4.81775397934994e-06,
"loss": 0.002,
"step": 335
},
{
"epoch": 1.268431001890359,
"grad_norm": 0.15248548984527588,
"learning_rate": 4.815691266327791e-06,
"loss": 0.0041,
"step": 336
},
{
"epoch": 1.272211720226843,
"grad_norm": 0.17409035563468933,
"learning_rate": 4.813617392360819e-06,
"loss": 0.0032,
"step": 337
},
{
"epoch": 1.275992438563327,
"grad_norm": 0.21849989891052246,
"learning_rate": 4.8115323674444795e-06,
"loss": 0.0027,
"step": 338
},
{
"epoch": 1.279773156899811,
"grad_norm": 0.26738497614860535,
"learning_rate": 4.809436201627974e-06,
"loss": 0.0035,
"step": 339
},
{
"epoch": 1.283553875236295,
"grad_norm": 0.23720847070217133,
"learning_rate": 4.807328905014201e-06,
"loss": 0.0037,
"step": 340
},
{
"epoch": 1.287334593572779,
"grad_norm": 0.25866371393203735,
"learning_rate": 4.805210487759704e-06,
"loss": 0.0066,
"step": 341
},
{
"epoch": 1.2911153119092629,
"grad_norm": 0.2647152245044708,
"learning_rate": 4.8030809600746255e-06,
"loss": 0.0064,
"step": 342
},
{
"epoch": 1.2948960302457466,
"grad_norm": 0.30110201239585876,
"learning_rate": 4.800940332222656e-06,
"loss": 0.007,
"step": 343
},
{
"epoch": 1.2986767485822306,
"grad_norm": 0.29565685987472534,
"learning_rate": 4.798788614520991e-06,
"loss": 0.0027,
"step": 344
},
{
"epoch": 1.3024574669187146,
"grad_norm": 0.3720064163208008,
"learning_rate": 4.796625817340266e-06,
"loss": 0.0075,
"step": 345
},
{
"epoch": 1.3062381852551985,
"grad_norm": 0.2937348186969757,
"learning_rate": 4.794451951104525e-06,
"loss": 0.0025,
"step": 346
},
{
"epoch": 1.3100189035916825,
"grad_norm": 0.24226990342140198,
"learning_rate": 4.792267026291157e-06,
"loss": 0.0026,
"step": 347
},
{
"epoch": 1.3137996219281662,
"grad_norm": 0.24391494691371918,
"learning_rate": 4.790071053430852e-06,
"loss": 0.0036,
"step": 348
},
{
"epoch": 1.3175803402646502,
"grad_norm": 0.204819455742836,
"learning_rate": 4.7878640431075466e-06,
"loss": 0.0057,
"step": 349
},
{
"epoch": 1.3213610586011342,
"grad_norm": 0.303836852312088,
"learning_rate": 4.785646005958377e-06,
"loss": 0.0042,
"step": 350
},
{
"epoch": 1.3251417769376181,
"grad_norm": 0.3792591989040375,
"learning_rate": 4.783416952673623e-06,
"loss": 0.0048,
"step": 351
},
{
"epoch": 1.3289224952741021,
"grad_norm": 1.29289972782135,
"learning_rate": 4.7811768939966605e-06,
"loss": 0.0071,
"step": 352
},
{
"epoch": 1.332703213610586,
"grad_norm": 0.3882276117801666,
"learning_rate": 4.778925840723909e-06,
"loss": 0.0047,
"step": 353
},
{
"epoch": 1.33648393194707,
"grad_norm": 0.38745951652526855,
"learning_rate": 4.776663803704775e-06,
"loss": 0.0066,
"step": 354
},
{
"epoch": 1.3402646502835538,
"grad_norm": 0.2202010154724121,
"learning_rate": 4.774390793841607e-06,
"loss": 0.0044,
"step": 355
},
{
"epoch": 1.3440453686200378,
"grad_norm": 0.1906670331954956,
"learning_rate": 4.772106822089639e-06,
"loss": 0.0043,
"step": 356
},
{
"epoch": 1.3478260869565217,
"grad_norm": 1.8778375387191772,
"learning_rate": 4.769811899456935e-06,
"loss": 0.0033,
"step": 357
},
{
"epoch": 1.3516068052930057,
"grad_norm": 0.2216220498085022,
"learning_rate": 4.767506037004344e-06,
"loss": 0.0028,
"step": 358
},
{
"epoch": 1.3553875236294897,
"grad_norm": 0.12414943426847458,
"learning_rate": 4.765189245845436e-06,
"loss": 0.0016,
"step": 359
},
{
"epoch": 1.3591682419659734,
"grad_norm": 0.12621334195137024,
"learning_rate": 4.762861537146458e-06,
"loss": 0.0017,
"step": 360
},
{
"epoch": 1.3629489603024574,
"grad_norm": 0.27140238881111145,
"learning_rate": 4.760522922126276e-06,
"loss": 0.0061,
"step": 361
},
{
"epoch": 1.3667296786389413,
"grad_norm": 0.39732441306114197,
"learning_rate": 4.75817341205632e-06,
"loss": 0.0061,
"step": 362
},
{
"epoch": 1.3705103969754253,
"grad_norm": 0.30900445580482483,
"learning_rate": 4.755813018260532e-06,
"loss": 0.0074,
"step": 363
},
{
"epoch": 1.3742911153119093,
"grad_norm": 0.2651839852333069,
"learning_rate": 4.753441752115309e-06,
"loss": 0.0058,
"step": 364
},
{
"epoch": 1.3780718336483933,
"grad_norm": 0.49944934248924255,
"learning_rate": 4.75105962504945e-06,
"loss": 0.0085,
"step": 365
},
{
"epoch": 1.3818525519848772,
"grad_norm": 0.38199540972709656,
"learning_rate": 4.748666648544101e-06,
"loss": 0.0035,
"step": 366
},
{
"epoch": 1.3856332703213612,
"grad_norm": 0.3837318420410156,
"learning_rate": 4.7462628341327e-06,
"loss": 0.0096,
"step": 367
},
{
"epoch": 1.389413988657845,
"grad_norm": 0.37649115920066833,
"learning_rate": 4.743848193400918e-06,
"loss": 0.0025,
"step": 368
},
{
"epoch": 1.393194706994329,
"grad_norm": 0.15870973467826843,
"learning_rate": 4.741422737986606e-06,
"loss": 0.0029,
"step": 369
},
{
"epoch": 1.3969754253308129,
"grad_norm": 0.21215665340423584,
"learning_rate": 4.73898647957974e-06,
"loss": 0.0029,
"step": 370
},
{
"epoch": 1.4007561436672968,
"grad_norm": 0.12150968611240387,
"learning_rate": 4.736539429922365e-06,
"loss": 0.0017,
"step": 371
},
{
"epoch": 1.4045368620037806,
"grad_norm": 0.22840923070907593,
"learning_rate": 4.734081600808531e-06,
"loss": 0.0031,
"step": 372
},
{
"epoch": 1.4083175803402646,
"grad_norm": 0.20265617966651917,
"learning_rate": 4.731613004084247e-06,
"loss": 0.0027,
"step": 373
},
{
"epoch": 1.4120982986767485,
"grad_norm": 0.3497235178947449,
"learning_rate": 4.7291336516474165e-06,
"loss": 0.0041,
"step": 374
},
{
"epoch": 1.4158790170132325,
"grad_norm": 0.18451634049415588,
"learning_rate": 4.726643555447782e-06,
"loss": 0.0023,
"step": 375
},
{
"epoch": 1.4196597353497165,
"grad_norm": 0.49125194549560547,
"learning_rate": 4.724142727486869e-06,
"loss": 0.002,
"step": 376
},
{
"epoch": 1.4234404536862004,
"grad_norm": 0.23641876876354218,
"learning_rate": 4.721631179817925e-06,
"loss": 0.0038,
"step": 377
},
{
"epoch": 1.4272211720226844,
"grad_norm": 0.650691032409668,
"learning_rate": 4.719108924545866e-06,
"loss": 0.0031,
"step": 378
},
{
"epoch": 1.4310018903591684,
"grad_norm": 0.21047154068946838,
"learning_rate": 4.716575973827214e-06,
"loss": 0.0051,
"step": 379
},
{
"epoch": 1.434782608695652,
"grad_norm": 0.3125157952308655,
"learning_rate": 4.714032339870041e-06,
"loss": 0.004,
"step": 380
},
{
"epoch": 1.438563327032136,
"grad_norm": 0.29793235659599304,
"learning_rate": 4.711478034933904e-06,
"loss": 0.008,
"step": 381
},
{
"epoch": 1.44234404536862,
"grad_norm": 0.25925037264823914,
"learning_rate": 4.708913071329798e-06,
"loss": 0.0024,
"step": 382
},
{
"epoch": 1.446124763705104,
"grad_norm": 0.1826639175415039,
"learning_rate": 4.706337461420087e-06,
"loss": 0.0024,
"step": 383
},
{
"epoch": 1.449905482041588,
"grad_norm": 0.2966572940349579,
"learning_rate": 4.703751217618444e-06,
"loss": 0.0077,
"step": 384
},
{
"epoch": 1.4536862003780717,
"grad_norm": 0.32406798005104065,
"learning_rate": 4.7011543523898e-06,
"loss": 0.0028,
"step": 385
},
{
"epoch": 1.4574669187145557,
"grad_norm": 0.34746837615966797,
"learning_rate": 4.698546878250272e-06,
"loss": 0.0048,
"step": 386
},
{
"epoch": 1.4612476370510397,
"grad_norm": 1.0227371454238892,
"learning_rate": 4.695928807767113e-06,
"loss": 0.0048,
"step": 387
},
{
"epoch": 1.4650283553875236,
"grad_norm": 0.40959542989730835,
"learning_rate": 4.693300153558647e-06,
"loss": 0.0072,
"step": 388
},
{
"epoch": 1.4688090737240076,
"grad_norm": 0.1713046133518219,
"learning_rate": 4.690660928294204e-06,
"loss": 0.0032,
"step": 389
},
{
"epoch": 1.4725897920604916,
"grad_norm": 0.24153202772140503,
"learning_rate": 4.68801114469407e-06,
"loss": 0.0064,
"step": 390
},
{
"epoch": 1.4763705103969755,
"grad_norm": 0.30539390444755554,
"learning_rate": 4.685350815529415e-06,
"loss": 0.0021,
"step": 391
},
{
"epoch": 1.4801512287334593,
"grad_norm": 0.19090750813484192,
"learning_rate": 4.682679953622234e-06,
"loss": 0.0031,
"step": 392
},
{
"epoch": 1.4839319470699432,
"grad_norm": 0.3017534613609314,
"learning_rate": 4.679998571845293e-06,
"loss": 0.004,
"step": 393
},
{
"epoch": 1.4877126654064272,
"grad_norm": 0.21019820868968964,
"learning_rate": 4.677306683122054e-06,
"loss": 0.0026,
"step": 394
},
{
"epoch": 1.4914933837429112,
"grad_norm": 0.37846195697784424,
"learning_rate": 4.674604300426621e-06,
"loss": 0.0072,
"step": 395
},
{
"epoch": 1.4952741020793952,
"grad_norm": 0.515809178352356,
"learning_rate": 4.671891436783679e-06,
"loss": 0.0125,
"step": 396
},
{
"epoch": 1.499054820415879,
"grad_norm": 0.30968862771987915,
"learning_rate": 4.669168105268425e-06,
"loss": 0.0034,
"step": 397
},
{
"epoch": 1.5028355387523629,
"grad_norm": 0.28533419966697693,
"learning_rate": 4.666434319006508e-06,
"loss": 0.0025,
"step": 398
},
{
"epoch": 1.5066162570888468,
"grad_norm": 0.3805786967277527,
"learning_rate": 4.663690091173968e-06,
"loss": 0.0041,
"step": 399
},
{
"epoch": 1.5103969754253308,
"grad_norm": 0.43205830454826355,
"learning_rate": 4.660935434997168e-06,
"loss": 0.0042,
"step": 400
},
{
"epoch": 1.5141776937618148,
"grad_norm": 0.2526014745235443,
"learning_rate": 4.658170363752734e-06,
"loss": 0.0048,
"step": 401
},
{
"epoch": 1.5179584120982987,
"grad_norm": 0.2991236746311188,
"learning_rate": 4.655394890767489e-06,
"loss": 0.0052,
"step": 402
},
{
"epoch": 1.5217391304347827,
"grad_norm": 0.23038192093372345,
"learning_rate": 4.652609029418389e-06,
"loss": 0.0028,
"step": 403
},
{
"epoch": 1.5255198487712667,
"grad_norm": 0.15317422151565552,
"learning_rate": 4.649812793132459e-06,
"loss": 0.002,
"step": 404
},
{
"epoch": 1.5293005671077504,
"grad_norm": 0.16759291291236877,
"learning_rate": 4.647006195386728e-06,
"loss": 0.0018,
"step": 405
},
{
"epoch": 1.5330812854442344,
"grad_norm": 0.29501622915267944,
"learning_rate": 4.644189249708167e-06,
"loss": 0.0045,
"step": 406
},
{
"epoch": 1.5368620037807184,
"grad_norm": 0.3393457233905792,
"learning_rate": 4.641361969673614e-06,
"loss": 0.0047,
"step": 407
},
{
"epoch": 1.5406427221172023,
"grad_norm": 0.06118448078632355,
"learning_rate": 4.638524368909723e-06,
"loss": 0.0008,
"step": 408
},
{
"epoch": 1.544423440453686,
"grad_norm": 0.18522684276103973,
"learning_rate": 4.635676461092886e-06,
"loss": 0.0025,
"step": 409
},
{
"epoch": 1.54820415879017,
"grad_norm": 0.27855828404426575,
"learning_rate": 4.632818259949175e-06,
"loss": 0.0048,
"step": 410
},
{
"epoch": 1.551984877126654,
"grad_norm": 0.15087196230888367,
"learning_rate": 4.6299497792542714e-06,
"loss": 0.0014,
"step": 411
},
{
"epoch": 1.555765595463138,
"grad_norm": 0.08206247538328171,
"learning_rate": 4.627071032833401e-06,
"loss": 0.0012,
"step": 412
},
{
"epoch": 1.559546313799622,
"grad_norm": 0.2290770560503006,
"learning_rate": 4.624182034561266e-06,
"loss": 0.0038,
"step": 413
},
{
"epoch": 1.563327032136106,
"grad_norm": 0.19741228222846985,
"learning_rate": 4.621282798361983e-06,
"loss": 0.0017,
"step": 414
},
{
"epoch": 1.5671077504725899,
"grad_norm": 0.21542416512966156,
"learning_rate": 4.618373338209008e-06,
"loss": 0.0029,
"step": 415
},
{
"epoch": 1.5708884688090738,
"grad_norm": 0.21594731509685516,
"learning_rate": 4.615453668125077e-06,
"loss": 0.0022,
"step": 416
},
{
"epoch": 1.5746691871455578,
"grad_norm": 0.13286042213439941,
"learning_rate": 4.612523802182134e-06,
"loss": 0.0019,
"step": 417
},
{
"epoch": 1.5784499054820416,
"grad_norm": 0.2678548991680145,
"learning_rate": 4.609583754501263e-06,
"loss": 0.0035,
"step": 418
},
{
"epoch": 1.5822306238185255,
"grad_norm": 0.19654716551303864,
"learning_rate": 4.6066335392526215e-06,
"loss": 0.0028,
"step": 419
},
{
"epoch": 1.5860113421550095,
"grad_norm": 0.13747157156467438,
"learning_rate": 4.603673170655372e-06,
"loss": 0.0014,
"step": 420
},
{
"epoch": 1.5897920604914932,
"grad_norm": 0.15026327967643738,
"learning_rate": 4.600702662977611e-06,
"loss": 0.0016,
"step": 421
},
{
"epoch": 1.5935727788279772,
"grad_norm": 0.18871524930000305,
"learning_rate": 4.597722030536306e-06,
"loss": 0.0032,
"step": 422
},
{
"epoch": 1.5973534971644612,
"grad_norm": 0.21016493439674377,
"learning_rate": 4.5947312876972215e-06,
"loss": 0.002,
"step": 423
},
{
"epoch": 1.6011342155009451,
"grad_norm": 0.2256452441215515,
"learning_rate": 4.591730448874849e-06,
"loss": 0.0016,
"step": 424
},
{
"epoch": 1.6049149338374291,
"grad_norm": 0.2980256676673889,
"learning_rate": 4.588719528532342e-06,
"loss": 0.0022,
"step": 425
},
{
"epoch": 1.608695652173913,
"grad_norm": 0.11986065655946732,
"learning_rate": 4.585698541181443e-06,
"loss": 0.0021,
"step": 426
},
{
"epoch": 1.612476370510397,
"grad_norm": 0.45475128293037415,
"learning_rate": 4.582667501382414e-06,
"loss": 0.0064,
"step": 427
},
{
"epoch": 1.616257088846881,
"grad_norm": 0.23436374962329865,
"learning_rate": 4.579626423743969e-06,
"loss": 0.0018,
"step": 428
},
{
"epoch": 1.620037807183365,
"grad_norm": 0.3097490668296814,
"learning_rate": 4.576575322923199e-06,
"loss": 0.0032,
"step": 429
},
{
"epoch": 1.6238185255198487,
"grad_norm": 0.3248469829559326,
"learning_rate": 4.573514213625505e-06,
"loss": 0.0029,
"step": 430
},
{
"epoch": 1.6275992438563327,
"grad_norm": 0.39466020464897156,
"learning_rate": 4.570443110604525e-06,
"loss": 0.0032,
"step": 431
},
{
"epoch": 1.6313799621928167,
"grad_norm": 0.11133938282728195,
"learning_rate": 4.567362028662066e-06,
"loss": 0.001,
"step": 432
},
{
"epoch": 1.6351606805293004,
"grad_norm": 0.15209349989891052,
"learning_rate": 4.564270982648026e-06,
"loss": 0.0029,
"step": 433
},
{
"epoch": 1.6389413988657844,
"grad_norm": 0.5543721318244934,
"learning_rate": 4.561169987460332e-06,
"loss": 0.0033,
"step": 434
},
{
"epoch": 1.6427221172022684,
"grad_norm": 0.29907462000846863,
"learning_rate": 4.55805905804486e-06,
"loss": 0.0034,
"step": 435
},
{
"epoch": 1.6465028355387523,
"grad_norm": 0.19079849123954773,
"learning_rate": 4.554938209395366e-06,
"loss": 0.0041,
"step": 436
},
{
"epoch": 1.6502835538752363,
"grad_norm": 0.3431627154350281,
"learning_rate": 4.551807456553416e-06,
"loss": 0.0054,
"step": 437
},
{
"epoch": 1.6540642722117203,
"grad_norm": 0.17760495841503143,
"learning_rate": 4.548666814608309e-06,
"loss": 0.0014,
"step": 438
},
{
"epoch": 1.6578449905482042,
"grad_norm": 0.18841500580310822,
"learning_rate": 4.545516298697006e-06,
"loss": 0.0025,
"step": 439
},
{
"epoch": 1.6616257088846882,
"grad_norm": 0.3963446021080017,
"learning_rate": 4.542355924004061e-06,
"loss": 0.0045,
"step": 440
},
{
"epoch": 1.6654064272211722,
"grad_norm": 0.3891541063785553,
"learning_rate": 4.5391857057615426e-06,
"loss": 0.006,
"step": 441
},
{
"epoch": 1.669187145557656,
"grad_norm": 0.28442084789276123,
"learning_rate": 4.53600565924896e-06,
"loss": 0.0035,
"step": 442
},
{
"epoch": 1.6729678638941399,
"grad_norm": 0.1774737536907196,
"learning_rate": 4.532815799793196e-06,
"loss": 0.001,
"step": 443
},
{
"epoch": 1.6767485822306238,
"grad_norm": 0.15262679755687714,
"learning_rate": 4.529616142768424e-06,
"loss": 0.0023,
"step": 444
},
{
"epoch": 1.6805293005671076,
"grad_norm": 0.21102535724639893,
"learning_rate": 4.526406703596043e-06,
"loss": 0.0044,
"step": 445
},
{
"epoch": 1.6843100189035916,
"grad_norm": 0.392592191696167,
"learning_rate": 4.523187497744599e-06,
"loss": 0.0057,
"step": 446
},
{
"epoch": 1.6880907372400755,
"grad_norm": 0.2138652503490448,
"learning_rate": 4.519958540729708e-06,
"loss": 0.0031,
"step": 447
},
{
"epoch": 1.6918714555765595,
"grad_norm": 0.2835381329059601,
"learning_rate": 4.516719848113983e-06,
"loss": 0.0033,
"step": 448
},
{
"epoch": 1.6956521739130435,
"grad_norm": 0.2596120834350586,
"learning_rate": 4.5134714355069635e-06,
"loss": 0.0028,
"step": 449
},
{
"epoch": 1.6994328922495274,
"grad_norm": 0.23696757853031158,
"learning_rate": 4.510213318565034e-06,
"loss": 0.0019,
"step": 450
},
{
"epoch": 1.7032136105860114,
"grad_norm": 0.35938960313796997,
"learning_rate": 4.506945512991352e-06,
"loss": 0.0034,
"step": 451
},
{
"epoch": 1.7069943289224954,
"grad_norm": 0.22387456893920898,
"learning_rate": 4.503668034535768e-06,
"loss": 0.0047,
"step": 452
},
{
"epoch": 1.7107750472589793,
"grad_norm": 0.276939332485199,
"learning_rate": 4.500380898994761e-06,
"loss": 0.004,
"step": 453
},
{
"epoch": 1.7145557655954633,
"grad_norm": 0.09006724506616592,
"learning_rate": 4.4970841222113435e-06,
"loss": 0.001,
"step": 454
},
{
"epoch": 1.718336483931947,
"grad_norm": 0.09450830519199371,
"learning_rate": 4.4937777200750045e-06,
"loss": 0.0012,
"step": 455
},
{
"epoch": 1.722117202268431,
"grad_norm": 0.8761259317398071,
"learning_rate": 4.490461708521621e-06,
"loss": 0.0062,
"step": 456
},
{
"epoch": 1.725897920604915,
"grad_norm": 0.21306881308555603,
"learning_rate": 4.487136103533383e-06,
"loss": 0.0026,
"step": 457
},
{
"epoch": 1.7296786389413987,
"grad_norm": 0.16684308648109436,
"learning_rate": 4.483800921138722e-06,
"loss": 0.0021,
"step": 458
},
{
"epoch": 1.7334593572778827,
"grad_norm": 0.354653537273407,
"learning_rate": 4.4804561774122275e-06,
"loss": 0.0059,
"step": 459
},
{
"epoch": 1.7372400756143667,
"grad_norm": 0.241005077958107,
"learning_rate": 4.4771018884745705e-06,
"loss": 0.0016,
"step": 460
},
{
"epoch": 1.7410207939508506,
"grad_norm": 0.23569397628307343,
"learning_rate": 4.473738070492428e-06,
"loss": 0.0035,
"step": 461
},
{
"epoch": 1.7448015122873346,
"grad_norm": 0.3624860942363739,
"learning_rate": 4.470364739678407e-06,
"loss": 0.0041,
"step": 462
},
{
"epoch": 1.7485822306238186,
"grad_norm": 0.1583845615386963,
"learning_rate": 4.466981912290959e-06,
"loss": 0.0021,
"step": 463
},
{
"epoch": 1.7523629489603025,
"grad_norm": 0.26447030901908875,
"learning_rate": 4.46358960463431e-06,
"loss": 0.0049,
"step": 464
},
{
"epoch": 1.7561436672967865,
"grad_norm": 0.2515549659729004,
"learning_rate": 4.4601878330583744e-06,
"loss": 0.002,
"step": 465
},
{
"epoch": 1.7599243856332705,
"grad_norm": 0.2806968688964844,
"learning_rate": 4.456776613958683e-06,
"loss": 0.0025,
"step": 466
},
{
"epoch": 1.7637051039697542,
"grad_norm": 0.08327832818031311,
"learning_rate": 4.453355963776301e-06,
"loss": 0.001,
"step": 467
},
{
"epoch": 1.7674858223062382,
"grad_norm": 0.2449251115322113,
"learning_rate": 4.449925898997745e-06,
"loss": 0.0017,
"step": 468
},
{
"epoch": 1.7712665406427222,
"grad_norm": 0.1794063001871109,
"learning_rate": 4.4464864361549105e-06,
"loss": 0.0021,
"step": 469
},
{
"epoch": 1.775047258979206,
"grad_norm": 0.24746963381767273,
"learning_rate": 4.443037591824988e-06,
"loss": 0.0052,
"step": 470
},
{
"epoch": 1.7788279773156899,
"grad_norm": 0.15880246460437775,
"learning_rate": 4.439579382630384e-06,
"loss": 0.0015,
"step": 471
},
{
"epoch": 1.7826086956521738,
"grad_norm": 0.3914218544960022,
"learning_rate": 4.43611182523864e-06,
"loss": 0.0027,
"step": 472
},
{
"epoch": 1.7863894139886578,
"grad_norm": 0.26020798087120056,
"learning_rate": 4.432634936362354e-06,
"loss": 0.0029,
"step": 473
},
{
"epoch": 1.7901701323251418,
"grad_norm": 0.23592509329319,
"learning_rate": 4.429148732759101e-06,
"loss": 0.0022,
"step": 474
},
{
"epoch": 1.7939508506616257,
"grad_norm": 0.20257288217544556,
"learning_rate": 4.425653231231344e-06,
"loss": 0.002,
"step": 475
},
{
"epoch": 1.7977315689981097,
"grad_norm": 0.338589072227478,
"learning_rate": 4.4221484486263664e-06,
"loss": 0.0017,
"step": 476
},
{
"epoch": 1.8015122873345937,
"grad_norm": 0.2656640410423279,
"learning_rate": 4.418634401836179e-06,
"loss": 0.005,
"step": 477
},
{
"epoch": 1.8052930056710776,
"grad_norm": 0.2045784294605255,
"learning_rate": 4.415111107797445e-06,
"loss": 0.0029,
"step": 478
},
{
"epoch": 1.8090737240075614,
"grad_norm": 0.3493967354297638,
"learning_rate": 4.411578583491397e-06,
"loss": 0.0037,
"step": 479
},
{
"epoch": 1.8128544423440454,
"grad_norm": 0.20509131252765656,
"learning_rate": 4.408036845943753e-06,
"loss": 0.0032,
"step": 480
},
{
"epoch": 1.8166351606805293,
"grad_norm": 0.23410679399967194,
"learning_rate": 4.404485912224638e-06,
"loss": 0.0029,
"step": 481
},
{
"epoch": 1.820415879017013,
"grad_norm": 0.21431049704551697,
"learning_rate": 4.400925799448498e-06,
"loss": 0.0015,
"step": 482
},
{
"epoch": 1.824196597353497,
"grad_norm": 0.2684635519981384,
"learning_rate": 4.39735652477402e-06,
"loss": 0.0028,
"step": 483
},
{
"epoch": 1.827977315689981,
"grad_norm": 0.1810218095779419,
"learning_rate": 4.393778105404051e-06,
"loss": 0.0018,
"step": 484
},
{
"epoch": 1.831758034026465,
"grad_norm": 0.32294052839279175,
"learning_rate": 4.390190558585508e-06,
"loss": 0.0028,
"step": 485
},
{
"epoch": 1.835538752362949,
"grad_norm": 0.23815307021141052,
"learning_rate": 4.386593901609304e-06,
"loss": 0.0033,
"step": 486
},
{
"epoch": 1.839319470699433,
"grad_norm": 0.19764836132526398,
"learning_rate": 4.382988151810258e-06,
"loss": 0.0027,
"step": 487
},
{
"epoch": 1.8431001890359169,
"grad_norm": 0.10845785588026047,
"learning_rate": 4.379373326567015e-06,
"loss": 0.0011,
"step": 488
},
{
"epoch": 1.8468809073724008,
"grad_norm": 0.17199721932411194,
"learning_rate": 4.375749443301957e-06,
"loss": 0.0018,
"step": 489
},
{
"epoch": 1.8506616257088848,
"grad_norm": 1.2053210735321045,
"learning_rate": 4.372116519481131e-06,
"loss": 0.003,
"step": 490
},
{
"epoch": 1.8544423440453688,
"grad_norm": 0.13807225227355957,
"learning_rate": 4.3684745726141474e-06,
"loss": 0.0011,
"step": 491
},
{
"epoch": 1.8582230623818525,
"grad_norm": 0.43164512515068054,
"learning_rate": 4.364823620254113e-06,
"loss": 0.0029,
"step": 492
},
{
"epoch": 1.8620037807183365,
"grad_norm": 0.4380972683429718,
"learning_rate": 4.361163679997532e-06,
"loss": 0.0085,
"step": 493
},
{
"epoch": 1.8657844990548205,
"grad_norm": 0.3626115322113037,
"learning_rate": 4.357494769484232e-06,
"loss": 0.0042,
"step": 494
},
{
"epoch": 1.8695652173913042,
"grad_norm": 0.3318653106689453,
"learning_rate": 4.3538169063972735e-06,
"loss": 0.0047,
"step": 495
},
{
"epoch": 1.8733459357277882,
"grad_norm": 0.12301299721002579,
"learning_rate": 4.350130108462864e-06,
"loss": 0.001,
"step": 496
},
{
"epoch": 1.8771266540642721,
"grad_norm": 0.10507173091173172,
"learning_rate": 4.346434393450277e-06,
"loss": 0.0012,
"step": 497
},
{
"epoch": 1.8809073724007561,
"grad_norm": 0.08294566720724106,
"learning_rate": 4.342729779171762e-06,
"loss": 0.0009,
"step": 498
},
{
"epoch": 1.88468809073724,
"grad_norm": 0.34475982189178467,
"learning_rate": 4.33901628348246e-06,
"loss": 0.0024,
"step": 499
},
{
"epoch": 1.888468809073724,
"grad_norm": 0.2793658375740051,
"learning_rate": 4.335293924280318e-06,
"loss": 0.0032,
"step": 500
},
{
"epoch": 1.892249527410208,
"grad_norm": 0.2560276687145233,
"learning_rate": 4.331562719506005e-06,
"loss": 0.0029,
"step": 501
},
{
"epoch": 1.896030245746692,
"grad_norm": 0.18784171342849731,
"learning_rate": 4.327822687142818e-06,
"loss": 0.0017,
"step": 502
},
{
"epoch": 1.899810964083176,
"grad_norm": 0.11437463760375977,
"learning_rate": 4.324073845216606e-06,
"loss": 0.0013,
"step": 503
},
{
"epoch": 1.9035916824196597,
"grad_norm": 0.22978849709033966,
"learning_rate": 4.320316211795674e-06,
"loss": 0.0026,
"step": 504
},
{
"epoch": 1.9073724007561437,
"grad_norm": 0.253154993057251,
"learning_rate": 4.3165498049907e-06,
"loss": 0.0078,
"step": 505
},
{
"epoch": 1.9111531190926276,
"grad_norm": 0.37678200006484985,
"learning_rate": 4.3127746429546465e-06,
"loss": 0.0044,
"step": 506
},
{
"epoch": 1.9149338374291114,
"grad_norm": 0.4084177613258362,
"learning_rate": 4.308990743882675e-06,
"loss": 0.0053,
"step": 507
},
{
"epoch": 1.9187145557655954,
"grad_norm": 0.2297290414571762,
"learning_rate": 4.305198126012057e-06,
"loss": 0.003,
"step": 508
},
{
"epoch": 1.9224952741020793,
"grad_norm": 0.22575493156909943,
"learning_rate": 4.301396807622084e-06,
"loss": 0.0057,
"step": 509
},
{
"epoch": 1.9262759924385633,
"grad_norm": 0.3964064419269562,
"learning_rate": 4.297586807033983e-06,
"loss": 0.0044,
"step": 510
},
{
"epoch": 1.9300567107750473,
"grad_norm": 0.2161191701889038,
"learning_rate": 4.293768142610828e-06,
"loss": 0.002,
"step": 511
},
{
"epoch": 1.9338374291115312,
"grad_norm": 0.22545190155506134,
"learning_rate": 4.2899408327574475e-06,
"loss": 0.0018,
"step": 512
},
{
"epoch": 1.9376181474480152,
"grad_norm": 0.32777416706085205,
"learning_rate": 4.2861048959203386e-06,
"loss": 0.0086,
"step": 513
},
{
"epoch": 1.9413988657844992,
"grad_norm": 0.18625925481319427,
"learning_rate": 4.282260350587579e-06,
"loss": 0.0024,
"step": 514
},
{
"epoch": 1.9451795841209831,
"grad_norm": 0.3894895911216736,
"learning_rate": 4.278407215288738e-06,
"loss": 0.0096,
"step": 515
},
{
"epoch": 1.9489603024574669,
"grad_norm": 0.1433977335691452,
"learning_rate": 4.274545508594783e-06,
"loss": 0.0025,
"step": 516
},
{
"epoch": 1.9527410207939508,
"grad_norm": 0.14154686033725739,
"learning_rate": 4.270675249117995e-06,
"loss": 0.0016,
"step": 517
},
{
"epoch": 1.9565217391304348,
"grad_norm": 0.8607096076011658,
"learning_rate": 4.2667964555118755e-06,
"loss": 0.0028,
"step": 518
},
{
"epoch": 1.9603024574669186,
"grad_norm": 0.1729510873556137,
"learning_rate": 4.26290914647106e-06,
"loss": 0.0051,
"step": 519
},
{
"epoch": 1.9640831758034025,
"grad_norm": 0.5080238580703735,
"learning_rate": 4.259013340731224e-06,
"loss": 0.0016,
"step": 520
},
{
"epoch": 1.9678638941398865,
"grad_norm": 0.19814470410346985,
"learning_rate": 4.255109057068995e-06,
"loss": 0.0018,
"step": 521
},
{
"epoch": 1.9716446124763705,
"grad_norm": 0.42453914880752563,
"learning_rate": 4.251196314301864e-06,
"loss": 0.0041,
"step": 522
},
{
"epoch": 1.9754253308128544,
"grad_norm": 0.2547927498817444,
"learning_rate": 4.2472751312880866e-06,
"loss": 0.0045,
"step": 523
},
{
"epoch": 1.9792060491493384,
"grad_norm": 0.20119120180606842,
"learning_rate": 4.243345526926605e-06,
"loss": 0.0019,
"step": 524
},
{
"epoch": 1.9829867674858224,
"grad_norm": 0.17681175470352173,
"learning_rate": 4.239407520156943e-06,
"loss": 0.0044,
"step": 525
},
{
"epoch": 1.9867674858223063,
"grad_norm": 0.1287936270236969,
"learning_rate": 4.235461129959128e-06,
"loss": 0.0017,
"step": 526
},
{
"epoch": 1.9905482041587903,
"grad_norm": 0.18217135965824127,
"learning_rate": 4.231506375353588e-06,
"loss": 0.0021,
"step": 527
},
{
"epoch": 1.994328922495274,
"grad_norm": 0.2492142617702484,
"learning_rate": 4.227543275401067e-06,
"loss": 0.0015,
"step": 528
},
{
"epoch": 1.998109640831758,
"grad_norm": 0.35620054602622986,
"learning_rate": 4.22357184920253e-06,
"loss": 0.0015,
"step": 529
},
{
"epoch": 2.0,
"grad_norm": 0.42932477593421936,
"learning_rate": 4.219592115899075e-06,
"loss": 0.0036,
"step": 530
},
{
"epoch": 2.003780718336484,
"grad_norm": 0.18517857789993286,
"learning_rate": 4.215604094671835e-06,
"loss": 0.0041,
"step": 531
},
{
"epoch": 2.007561436672968,
"grad_norm": 0.3995055556297302,
"learning_rate": 4.211607804741888e-06,
"loss": 0.0039,
"step": 532
},
{
"epoch": 2.011342155009452,
"grad_norm": 0.2439192533493042,
"learning_rate": 4.207603265370166e-06,
"loss": 0.0033,
"step": 533
},
{
"epoch": 2.015122873345936,
"grad_norm": 0.34398382902145386,
"learning_rate": 4.2035904958573615e-06,
"loss": 0.0018,
"step": 534
},
{
"epoch": 2.01890359168242,
"grad_norm": 0.07287435233592987,
"learning_rate": 4.1995695155438325e-06,
"loss": 0.0007,
"step": 535
},
{
"epoch": 2.022684310018904,
"grad_norm": 0.04657262936234474,
"learning_rate": 4.195540343809511e-06,
"loss": 0.0005,
"step": 536
},
{
"epoch": 2.0264650283553873,
"grad_norm": 0.1362239122390747,
"learning_rate": 4.19150300007381e-06,
"loss": 0.0008,
"step": 537
},
{
"epoch": 2.0302457466918713,
"grad_norm": 0.34500372409820557,
"learning_rate": 4.187457503795526e-06,
"loss": 0.0019,
"step": 538
},
{
"epoch": 2.0340264650283553,
"grad_norm": 0.27130597829818726,
"learning_rate": 4.183403874472754e-06,
"loss": 0.0071,
"step": 539
},
{
"epoch": 2.0378071833648392,
"grad_norm": 0.042355943471193314,
"learning_rate": 4.179342131642783e-06,
"loss": 0.0004,
"step": 540
},
{
"epoch": 2.041587901701323,
"grad_norm": 0.27351462841033936,
"learning_rate": 4.175272294882008e-06,
"loss": 0.0064,
"step": 541
},
{
"epoch": 2.045368620037807,
"grad_norm": 0.14808684587478638,
"learning_rate": 4.1711943838058346e-06,
"loss": 0.001,
"step": 542
},
{
"epoch": 2.049149338374291,
"grad_norm": 0.3332342207431793,
"learning_rate": 4.167108418068585e-06,
"loss": 0.0017,
"step": 543
},
{
"epoch": 2.052930056710775,
"grad_norm": 0.10606346279382706,
"learning_rate": 4.163014417363402e-06,
"loss": 0.0013,
"step": 544
},
{
"epoch": 2.056710775047259,
"grad_norm": 0.22433032095432281,
"learning_rate": 4.158912401422152e-06,
"loss": 0.0022,
"step": 545
},
{
"epoch": 2.060491493383743,
"grad_norm": 0.08767926692962646,
"learning_rate": 4.154802390015338e-06,
"loss": 0.0007,
"step": 546
},
{
"epoch": 2.064272211720227,
"grad_norm": 0.25233975052833557,
"learning_rate": 4.150684402951994e-06,
"loss": 0.0025,
"step": 547
},
{
"epoch": 2.068052930056711,
"grad_norm": 0.2354487180709839,
"learning_rate": 4.146558460079595e-06,
"loss": 0.0014,
"step": 548
},
{
"epoch": 2.0718336483931945,
"grad_norm": 0.1968374401330948,
"learning_rate": 4.142424581283962e-06,
"loss": 0.0016,
"step": 549
},
{
"epoch": 2.0756143667296785,
"grad_norm": 0.1780686229467392,
"learning_rate": 4.138282786489165e-06,
"loss": 0.0015,
"step": 550
},
{
"epoch": 2.0793950850661624,
"grad_norm": 0.14376260340213776,
"learning_rate": 4.134133095657424e-06,
"loss": 0.001,
"step": 551
},
{
"epoch": 2.0831758034026464,
"grad_norm": 0.42116665840148926,
"learning_rate": 4.129975528789021e-06,
"loss": 0.0018,
"step": 552
},
{
"epoch": 2.0869565217391304,
"grad_norm": 0.36534008383750916,
"learning_rate": 4.125810105922192e-06,
"loss": 0.0032,
"step": 553
},
{
"epoch": 2.0907372400756143,
"grad_norm": 0.261728972196579,
"learning_rate": 4.121636847133039e-06,
"loss": 0.0056,
"step": 554
},
{
"epoch": 2.0945179584120983,
"grad_norm": 0.10244060307741165,
"learning_rate": 4.117455772535433e-06,
"loss": 0.001,
"step": 555
},
{
"epoch": 2.0982986767485823,
"grad_norm": 0.34060749411582947,
"learning_rate": 4.113266902280914e-06,
"loss": 0.0024,
"step": 556
},
{
"epoch": 2.1020793950850662,
"grad_norm": 0.6971008777618408,
"learning_rate": 4.109070256558589e-06,
"loss": 0.0057,
"step": 557
},
{
"epoch": 2.10586011342155,
"grad_norm": 0.48935261368751526,
"learning_rate": 4.10486585559505e-06,
"loss": 0.0035,
"step": 558
},
{
"epoch": 2.109640831758034,
"grad_norm": 0.27988186478614807,
"learning_rate": 4.100653719654259e-06,
"loss": 0.0015,
"step": 559
},
{
"epoch": 2.113421550094518,
"grad_norm": 0.13683204352855682,
"learning_rate": 4.096433869037464e-06,
"loss": 0.0019,
"step": 560
},
{
"epoch": 2.1172022684310017,
"grad_norm": 0.17368103563785553,
"learning_rate": 4.0922063240830914e-06,
"loss": 0.0016,
"step": 561
},
{
"epoch": 2.1209829867674856,
"grad_norm": 0.3833687901496887,
"learning_rate": 4.087971105166654e-06,
"loss": 0.0027,
"step": 562
},
{
"epoch": 2.1247637051039696,
"grad_norm": 0.2219664752483368,
"learning_rate": 4.08372823270065e-06,
"loss": 0.0026,
"step": 563
},
{
"epoch": 2.1285444234404536,
"grad_norm": 0.291406512260437,
"learning_rate": 4.079477727134466e-06,
"loss": 0.0029,
"step": 564
},
{
"epoch": 2.1323251417769375,
"grad_norm": 0.3142309784889221,
"learning_rate": 4.075219608954279e-06,
"loss": 0.0064,
"step": 565
},
{
"epoch": 2.1361058601134215,
"grad_norm": 0.2063475400209427,
"learning_rate": 4.070953898682952e-06,
"loss": 0.0042,
"step": 566
},
{
"epoch": 2.1398865784499055,
"grad_norm": 0.05330202728509903,
"learning_rate": 4.066680616879946e-06,
"loss": 0.0006,
"step": 567
},
{
"epoch": 2.1436672967863895,
"grad_norm": 0.22444003820419312,
"learning_rate": 4.062399784141209e-06,
"loss": 0.002,
"step": 568
},
{
"epoch": 2.1474480151228734,
"grad_norm": 0.22138641774654388,
"learning_rate": 4.058111421099087e-06,
"loss": 0.0017,
"step": 569
},
{
"epoch": 2.1512287334593574,
"grad_norm": 0.12596635520458221,
"learning_rate": 4.0538155484222156e-06,
"loss": 0.0007,
"step": 570
},
{
"epoch": 2.1550094517958414,
"grad_norm": 0.10118561238050461,
"learning_rate": 4.0495121868154275e-06,
"loss": 0.0013,
"step": 571
},
{
"epoch": 2.1587901701323253,
"grad_norm": 0.0691247284412384,
"learning_rate": 4.045201357019648e-06,
"loss": 0.0009,
"step": 572
},
{
"epoch": 2.1625708884688093,
"grad_norm": 0.14483240246772766,
"learning_rate": 4.040883079811799e-06,
"loss": 0.0012,
"step": 573
},
{
"epoch": 2.166351606805293,
"grad_norm": 0.08602052927017212,
"learning_rate": 4.036557376004694e-06,
"loss": 0.0008,
"step": 574
},
{
"epoch": 2.170132325141777,
"grad_norm": 0.13205234706401825,
"learning_rate": 4.032224266446943e-06,
"loss": 0.0009,
"step": 575
},
{
"epoch": 2.1739130434782608,
"grad_norm": 0.039761174470186234,
"learning_rate": 4.027883772022847e-06,
"loss": 0.0005,
"step": 576
},
{
"epoch": 2.1776937618147447,
"grad_norm": 0.35107317566871643,
"learning_rate": 4.0235359136523024e-06,
"loss": 0.0027,
"step": 577
},
{
"epoch": 2.1814744801512287,
"grad_norm": 0.17431484162807465,
"learning_rate": 4.0191807122906965e-06,
"loss": 0.0016,
"step": 578
},
{
"epoch": 2.1852551984877127,
"grad_norm": 0.3038300573825836,
"learning_rate": 4.014818188928808e-06,
"loss": 0.002,
"step": 579
},
{
"epoch": 2.1890359168241966,
"grad_norm": 0.0982559323310852,
"learning_rate": 4.010448364592703e-06,
"loss": 0.0008,
"step": 580
},
{
"epoch": 2.1928166351606806,
"grad_norm": 0.19003631174564362,
"learning_rate": 4.00607126034364e-06,
"loss": 0.0024,
"step": 581
},
{
"epoch": 2.1965973534971646,
"grad_norm": 0.23104579746723175,
"learning_rate": 4.0016868972779636e-06,
"loss": 0.0027,
"step": 582
},
{
"epoch": 2.2003780718336485,
"grad_norm": 0.1018993929028511,
"learning_rate": 3.997295296527001e-06,
"loss": 0.0008,
"step": 583
},
{
"epoch": 2.2041587901701325,
"grad_norm": 0.2123573124408722,
"learning_rate": 3.992896479256966e-06,
"loss": 0.0017,
"step": 584
},
{
"epoch": 2.207939508506616,
"grad_norm": 0.5353474617004395,
"learning_rate": 3.988490466668852e-06,
"loss": 0.0044,
"step": 585
},
{
"epoch": 2.2117202268431,
"grad_norm": 0.15907587110996246,
"learning_rate": 3.984077279998333e-06,
"loss": 0.0026,
"step": 586
},
{
"epoch": 2.215500945179584,
"grad_norm": 0.09038275480270386,
"learning_rate": 3.979656940515659e-06,
"loss": 0.0008,
"step": 587
},
{
"epoch": 2.219281663516068,
"grad_norm": 0.15547309815883636,
"learning_rate": 3.975229469525555e-06,
"loss": 0.0018,
"step": 588
},
{
"epoch": 2.223062381852552,
"grad_norm": 0.17092019319534302,
"learning_rate": 3.970794888367116e-06,
"loss": 0.0024,
"step": 589
},
{
"epoch": 2.226843100189036,
"grad_norm": 0.17940843105316162,
"learning_rate": 3.96635321841371e-06,
"loss": 0.0011,
"step": 590
},
{
"epoch": 2.23062381852552,
"grad_norm": 0.08405204862356186,
"learning_rate": 3.961904481072867e-06,
"loss": 0.0009,
"step": 591
},
{
"epoch": 2.234404536862004,
"grad_norm": 0.16749833524227142,
"learning_rate": 3.957448697786181e-06,
"loss": 0.0016,
"step": 592
},
{
"epoch": 2.2381852551984878,
"grad_norm": 0.1705680787563324,
"learning_rate": 3.952985890029205e-06,
"loss": 0.0026,
"step": 593
},
{
"epoch": 2.2419659735349717,
"grad_norm": 0.2567225396633148,
"learning_rate": 3.948516079311349e-06,
"loss": 0.003,
"step": 594
},
{
"epoch": 2.2457466918714557,
"grad_norm": 0.06781842559576035,
"learning_rate": 3.9440392871757735e-06,
"loss": 0.0008,
"step": 595
},
{
"epoch": 2.2495274102079397,
"grad_norm": 0.08968048542737961,
"learning_rate": 3.93955553519929e-06,
"loss": 0.0011,
"step": 596
},
{
"epoch": 2.2533081285444236,
"grad_norm": 0.28453701734542847,
"learning_rate": 3.935064844992251e-06,
"loss": 0.0038,
"step": 597
},
{
"epoch": 2.257088846880907,
"grad_norm": 0.04295709729194641,
"learning_rate": 3.930567238198451e-06,
"loss": 0.0005,
"step": 598
},
{
"epoch": 2.260869565217391,
"grad_norm": 0.14463184773921967,
"learning_rate": 3.926062736495021e-06,
"loss": 0.0008,
"step": 599
},
{
"epoch": 2.264650283553875,
"grad_norm": 0.07073153555393219,
"learning_rate": 3.921551361592323e-06,
"loss": 0.0007,
"step": 600
},
{
"epoch": 2.268431001890359,
"grad_norm": 0.15077605843544006,
"learning_rate": 3.917033135233845e-06,
"loss": 0.0004,
"step": 601
},
{
"epoch": 2.272211720226843,
"grad_norm": 0.24590805172920227,
"learning_rate": 3.912508079196097e-06,
"loss": 0.0024,
"step": 602
},
{
"epoch": 2.275992438563327,
"grad_norm": 1.3826171159744263,
"learning_rate": 3.907976215288507e-06,
"loss": 0.0028,
"step": 603
},
{
"epoch": 2.279773156899811,
"grad_norm": 0.17305316030979156,
"learning_rate": 3.903437565353314e-06,
"loss": 0.0008,
"step": 604
},
{
"epoch": 2.283553875236295,
"grad_norm": 0.13404357433319092,
"learning_rate": 3.8988921512654644e-06,
"loss": 0.001,
"step": 605
},
{
"epoch": 2.287334593572779,
"grad_norm": 0.07687821239233017,
"learning_rate": 3.8943399949325055e-06,
"loss": 0.0007,
"step": 606
},
{
"epoch": 2.291115311909263,
"grad_norm": 0.13530321419239044,
"learning_rate": 3.88978111829448e-06,
"loss": 0.0009,
"step": 607
},
{
"epoch": 2.294896030245747,
"grad_norm": 0.06520530581474304,
"learning_rate": 3.885215543323821e-06,
"loss": 0.0005,
"step": 608
},
{
"epoch": 2.298676748582231,
"grad_norm": 0.24392560124397278,
"learning_rate": 3.880643292025245e-06,
"loss": 0.0024,
"step": 609
},
{
"epoch": 2.302457466918715,
"grad_norm": 0.376128613948822,
"learning_rate": 3.876064386435646e-06,
"loss": 0.0024,
"step": 610
},
{
"epoch": 2.3062381852551983,
"grad_norm": 0.15951082110404968,
"learning_rate": 3.871478848623992e-06,
"loss": 0.0009,
"step": 611
},
{
"epoch": 2.3100189035916823,
"grad_norm": 0.1541188359260559,
"learning_rate": 3.8668867006912126e-06,
"loss": 0.0026,
"step": 612
},
{
"epoch": 2.3137996219281662,
"grad_norm": 0.16861151158809662,
"learning_rate": 3.862287964770099e-06,
"loss": 0.0022,
"step": 613
},
{
"epoch": 2.31758034026465,
"grad_norm": 0.05262811481952667,
"learning_rate": 3.857682663025195e-06,
"loss": 0.0005,
"step": 614
},
{
"epoch": 2.321361058601134,
"grad_norm": 0.15682393312454224,
"learning_rate": 3.853070817652688e-06,
"loss": 0.0019,
"step": 615
},
{
"epoch": 2.325141776937618,
"grad_norm": 0.2507023513317108,
"learning_rate": 3.848452450880303e-06,
"loss": 0.0019,
"step": 616
},
{
"epoch": 2.328922495274102,
"grad_norm": 0.2650613784790039,
"learning_rate": 3.8438275849671994e-06,
"loss": 0.0041,
"step": 617
},
{
"epoch": 2.332703213610586,
"grad_norm": 0.1116451546549797,
"learning_rate": 3.839196242203859e-06,
"loss": 0.001,
"step": 618
},
{
"epoch": 2.33648393194707,
"grad_norm": 0.022393720224499702,
"learning_rate": 3.834558444911978e-06,
"loss": 0.0003,
"step": 619
},
{
"epoch": 2.340264650283554,
"grad_norm": 0.02391130104660988,
"learning_rate": 3.829914215444363e-06,
"loss": 0.0003,
"step": 620
},
{
"epoch": 2.344045368620038,
"grad_norm": 0.2857953906059265,
"learning_rate": 3.825263576184823e-06,
"loss": 0.0021,
"step": 621
},
{
"epoch": 2.3478260869565215,
"grad_norm": 0.08123797178268433,
"learning_rate": 3.82060654954806e-06,
"loss": 0.0005,
"step": 622
},
{
"epoch": 2.351606805293006,
"grad_norm": 0.30385783314704895,
"learning_rate": 3.8159431579795605e-06,
"loss": 0.0026,
"step": 623
},
{
"epoch": 2.3553875236294894,
"grad_norm": 0.29088184237480164,
"learning_rate": 3.8112734239554884e-06,
"loss": 0.0035,
"step": 624
},
{
"epoch": 2.3591682419659734,
"grad_norm": 0.19305771589279175,
"learning_rate": 3.806597369982574e-06,
"loss": 0.0012,
"step": 625
},
{
"epoch": 2.3629489603024574,
"grad_norm": 0.12433846294879913,
"learning_rate": 3.8019150185980125e-06,
"loss": 0.001,
"step": 626
},
{
"epoch": 2.3667296786389413,
"grad_norm": 0.07855097204446793,
"learning_rate": 3.7972263923693477e-06,
"loss": 0.0006,
"step": 627
},
{
"epoch": 2.3705103969754253,
"grad_norm": 0.12908008694648743,
"learning_rate": 3.792531513894365e-06,
"loss": 0.0006,
"step": 628
},
{
"epoch": 2.3742911153119093,
"grad_norm": 0.14044421911239624,
"learning_rate": 3.7878304058009883e-06,
"loss": 0.001,
"step": 629
},
{
"epoch": 2.3780718336483933,
"grad_norm": 0.12183783203363419,
"learning_rate": 3.7831230907471626e-06,
"loss": 0.001,
"step": 630
},
{
"epoch": 2.381852551984877,
"grad_norm": 0.03276016563177109,
"learning_rate": 3.778409591420749e-06,
"loss": 0.0003,
"step": 631
},
{
"epoch": 2.385633270321361,
"grad_norm": 0.19104357063770294,
"learning_rate": 3.7736899305394178e-06,
"loss": 0.0026,
"step": 632
},
{
"epoch": 2.389413988657845,
"grad_norm": 0.16137288510799408,
"learning_rate": 3.768964130850532e-06,
"loss": 0.0017,
"step": 633
},
{
"epoch": 2.393194706994329,
"grad_norm": 0.11547023802995682,
"learning_rate": 3.7642322151310453e-06,
"loss": 0.0009,
"step": 634
},
{
"epoch": 2.3969754253308126,
"grad_norm": 0.06410176306962967,
"learning_rate": 3.7594942061873863e-06,
"loss": 0.0004,
"step": 635
},
{
"epoch": 2.4007561436672966,
"grad_norm": 0.0972183421254158,
"learning_rate": 3.754750126855353e-06,
"loss": 0.0016,
"step": 636
},
{
"epoch": 2.4045368620037806,
"grad_norm": 0.25424033403396606,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.0017,
"step": 637
},
{
"epoch": 2.4083175803402646,
"grad_norm": 0.0990934744477272,
"learning_rate": 3.7452438485155305e-06,
"loss": 0.0006,
"step": 638
},
{
"epoch": 2.4120982986767485,
"grad_norm": 0.2217491716146469,
"learning_rate": 3.740481695325181e-06,
"loss": 0.002,
"step": 639
},
{
"epoch": 2.4158790170132325,
"grad_norm": 0.10147841274738312,
"learning_rate": 3.7357135633811198e-06,
"loss": 0.0007,
"step": 640
},
{
"epoch": 2.4196597353497165,
"grad_norm": 0.32936543226242065,
"learning_rate": 3.7309394756643265e-06,
"loss": 0.0017,
"step": 641
},
{
"epoch": 2.4234404536862004,
"grad_norm": 0.15416279435157776,
"learning_rate": 3.7261594551844863e-06,
"loss": 0.0013,
"step": 642
},
{
"epoch": 2.4272211720226844,
"grad_norm": 0.11692031472921371,
"learning_rate": 3.7213735249798826e-06,
"loss": 0.0011,
"step": 643
},
{
"epoch": 2.4310018903591684,
"grad_norm": 0.24383437633514404,
"learning_rate": 3.716581708117276e-06,
"loss": 0.0035,
"step": 644
},
{
"epoch": 2.4347826086956523,
"grad_norm": 0.3260003924369812,
"learning_rate": 3.711784027691805e-06,
"loss": 0.0026,
"step": 645
},
{
"epoch": 2.4385633270321363,
"grad_norm": 0.4543023407459259,
"learning_rate": 3.7069805068268626e-06,
"loss": 0.0027,
"step": 646
},
{
"epoch": 2.4423440453686203,
"grad_norm": 0.42455440759658813,
"learning_rate": 3.7021711686739963e-06,
"loss": 0.0048,
"step": 647
},
{
"epoch": 2.446124763705104,
"grad_norm": 0.2571258544921875,
"learning_rate": 3.697356036412788e-06,
"loss": 0.0023,
"step": 648
},
{
"epoch": 2.4499054820415878,
"grad_norm": 0.409236341714859,
"learning_rate": 3.6925351332507477e-06,
"loss": 0.0036,
"step": 649
},
{
"epoch": 2.4536862003780717,
"grad_norm": 0.0866311714053154,
"learning_rate": 3.6877084824231962e-06,
"loss": 0.0005,
"step": 650
},
{
"epoch": 2.4574669187145557,
"grad_norm": 0.22943198680877686,
"learning_rate": 3.6828761071931605e-06,
"loss": 0.0014,
"step": 651
},
{
"epoch": 2.4612476370510397,
"grad_norm": 0.2298421412706375,
"learning_rate": 3.6780380308512544e-06,
"loss": 0.0018,
"step": 652
},
{
"epoch": 2.4650283553875236,
"grad_norm": 0.24654293060302734,
"learning_rate": 3.6731942767155703e-06,
"loss": 0.0025,
"step": 653
},
{
"epoch": 2.4688090737240076,
"grad_norm": 0.1594986468553543,
"learning_rate": 3.668344868131567e-06,
"loss": 0.0009,
"step": 654
},
{
"epoch": 2.4725897920604916,
"grad_norm": 0.38067856431007385,
"learning_rate": 3.6634898284719533e-06,
"loss": 0.0023,
"step": 655
},
{
"epoch": 2.4763705103969755,
"grad_norm": 1.169722557067871,
"learning_rate": 3.6586291811365814e-06,
"loss": 0.0038,
"step": 656
},
{
"epoch": 2.4801512287334595,
"grad_norm": 0.2724793553352356,
"learning_rate": 3.653762949552328e-06,
"loss": 0.0056,
"step": 657
},
{
"epoch": 2.4839319470699435,
"grad_norm": 0.4527992606163025,
"learning_rate": 3.648891157172986e-06,
"loss": 0.004,
"step": 658
},
{
"epoch": 2.487712665406427,
"grad_norm": 0.12839777767658234,
"learning_rate": 3.6440138274791505e-06,
"loss": 0.0007,
"step": 659
},
{
"epoch": 2.491493383742911,
"grad_norm": 0.17698079347610474,
"learning_rate": 3.6391309839781014e-06,
"loss": 0.002,
"step": 660
},
{
"epoch": 2.495274102079395,
"grad_norm": 0.3649884760379791,
"learning_rate": 3.6342426502036972e-06,
"loss": 0.003,
"step": 661
},
{
"epoch": 2.499054820415879,
"grad_norm": 0.10224947333335876,
"learning_rate": 3.6293488497162555e-06,
"loss": 0.0007,
"step": 662
},
{
"epoch": 2.502835538752363,
"grad_norm": 0.1708725541830063,
"learning_rate": 3.6244496061024416e-06,
"loss": 0.0012,
"step": 663
},
{
"epoch": 2.506616257088847,
"grad_norm": 0.06363553553819656,
"learning_rate": 3.6195449429751585e-06,
"loss": 0.0005,
"step": 664
},
{
"epoch": 2.510396975425331,
"grad_norm": 0.2798851728439331,
"learning_rate": 3.614634883973425e-06,
"loss": 0.0018,
"step": 665
},
{
"epoch": 2.5141776937618148,
"grad_norm": 0.027707351371645927,
"learning_rate": 3.6097194527622697e-06,
"loss": 0.0003,
"step": 666
},
{
"epoch": 2.5179584120982987,
"grad_norm": 0.25401636958122253,
"learning_rate": 3.6047986730326132e-06,
"loss": 0.0015,
"step": 667
},
{
"epoch": 2.5217391304347827,
"grad_norm": 0.19876894354820251,
"learning_rate": 3.5998725685011526e-06,
"loss": 0.0021,
"step": 668
},
{
"epoch": 2.5255198487712667,
"grad_norm": 0.28308069705963135,
"learning_rate": 3.594941162910251e-06,
"loss": 0.0015,
"step": 669
},
{
"epoch": 2.52930056710775,
"grad_norm": 0.23315759003162384,
"learning_rate": 3.59000448002782e-06,
"loss": 0.0015,
"step": 670
},
{
"epoch": 2.5330812854442346,
"grad_norm": 0.20551222562789917,
"learning_rate": 3.585062543647205e-06,
"loss": 0.0018,
"step": 671
},
{
"epoch": 2.536862003780718,
"grad_norm": 0.4355032742023468,
"learning_rate": 3.580115377587075e-06,
"loss": 0.0081,
"step": 672
},
{
"epoch": 2.5406427221172025,
"grad_norm": 0.055470116436481476,
"learning_rate": 3.5751630056913017e-06,
"loss": 0.0007,
"step": 673
},
{
"epoch": 2.544423440453686,
"grad_norm": 0.21149052679538727,
"learning_rate": 3.5702054518288477e-06,
"loss": 0.0025,
"step": 674
},
{
"epoch": 2.54820415879017,
"grad_norm": 0.09681661427021027,
"learning_rate": 3.5652427398936527e-06,
"loss": 0.0007,
"step": 675
},
{
"epoch": 2.551984877126654,
"grad_norm": 0.12468747794628143,
"learning_rate": 3.560274893804515e-06,
"loss": 0.0007,
"step": 676
},
{
"epoch": 2.555765595463138,
"grad_norm": 0.21462489664554596,
"learning_rate": 3.55530193750498e-06,
"loss": 0.0031,
"step": 677
},
{
"epoch": 2.559546313799622,
"grad_norm": 0.4433956742286682,
"learning_rate": 3.5503238949632202e-06,
"loss": 0.0045,
"step": 678
},
{
"epoch": 2.563327032136106,
"grad_norm": 0.06683538109064102,
"learning_rate": 3.5453407901719232e-06,
"loss": 0.0006,
"step": 679
},
{
"epoch": 2.56710775047259,
"grad_norm": 0.09750392287969589,
"learning_rate": 3.5403526471481777e-06,
"loss": 0.001,
"step": 680
},
{
"epoch": 2.570888468809074,
"grad_norm": 0.3301282823085785,
"learning_rate": 3.535359489933351e-06,
"loss": 0.0025,
"step": 681
},
{
"epoch": 2.574669187145558,
"grad_norm": 0.18829470872879028,
"learning_rate": 3.5303613425929805e-06,
"loss": 0.0017,
"step": 682
},
{
"epoch": 2.5784499054820413,
"grad_norm": 0.24966050684452057,
"learning_rate": 3.5253582292166525e-06,
"loss": 0.0039,
"step": 683
},
{
"epoch": 2.5822306238185257,
"grad_norm": 0.32009822130203247,
"learning_rate": 3.5203501739178904e-06,
"loss": 0.0029,
"step": 684
},
{
"epoch": 2.5860113421550093,
"grad_norm": 0.11968351900577545,
"learning_rate": 3.515337200834034e-06,
"loss": 0.0006,
"step": 685
},
{
"epoch": 2.5897920604914932,
"grad_norm": 0.48058372735977173,
"learning_rate": 3.510319334126126e-06,
"loss": 0.0025,
"step": 686
},
{
"epoch": 2.593572778827977,
"grad_norm": 0.6822208166122437,
"learning_rate": 3.5052965979787952e-06,
"loss": 0.0004,
"step": 687
},
{
"epoch": 2.597353497164461,
"grad_norm": 0.1800459921360016,
"learning_rate": 3.50026901660014e-06,
"loss": 0.0011,
"step": 688
},
{
"epoch": 2.601134215500945,
"grad_norm": 0.032964590936899185,
"learning_rate": 3.4952366142216097e-06,
"loss": 0.0004,
"step": 689
},
{
"epoch": 2.604914933837429,
"grad_norm": 0.8373531103134155,
"learning_rate": 3.4901994150978926e-06,
"loss": 0.0008,
"step": 690
},
{
"epoch": 2.608695652173913,
"grad_norm": 0.061929311603307724,
"learning_rate": 3.4851574435067925e-06,
"loss": 0.0006,
"step": 691
},
{
"epoch": 2.612476370510397,
"grad_norm": 0.19831609725952148,
"learning_rate": 3.480110723749116e-06,
"loss": 0.0015,
"step": 692
},
{
"epoch": 2.616257088846881,
"grad_norm": 0.2698705494403839,
"learning_rate": 3.4750592801485565e-06,
"loss": 0.0026,
"step": 693
},
{
"epoch": 2.620037807183365,
"grad_norm": 0.21397550404071808,
"learning_rate": 3.470003137051573e-06,
"loss": 0.002,
"step": 694
},
{
"epoch": 2.623818525519849,
"grad_norm": 0.26705247163772583,
"learning_rate": 3.4649423188272743e-06,
"loss": 0.0042,
"step": 695
},
{
"epoch": 2.6275992438563325,
"grad_norm": 0.4273588955402374,
"learning_rate": 3.459876849867303e-06,
"loss": 0.0066,
"step": 696
},
{
"epoch": 2.631379962192817,
"grad_norm": 0.47062620520591736,
"learning_rate": 3.454806754585716e-06,
"loss": 0.0025,
"step": 697
},
{
"epoch": 2.6351606805293004,
"grad_norm": 0.057229600846767426,
"learning_rate": 3.4497320574188696e-06,
"loss": 0.0007,
"step": 698
},
{
"epoch": 2.6389413988657844,
"grad_norm": 0.1500159204006195,
"learning_rate": 3.4446527828252975e-06,
"loss": 0.0006,
"step": 699
},
{
"epoch": 2.6427221172022684,
"grad_norm": 0.2159212976694107,
"learning_rate": 3.4395689552855956e-06,
"loss": 0.0011,
"step": 700
},
{
"epoch": 2.6465028355387523,
"grad_norm": 0.07223492860794067,
"learning_rate": 3.4344805993023056e-06,
"loss": 0.0004,
"step": 701
},
{
"epoch": 2.6502835538752363,
"grad_norm": 0.22909198701381683,
"learning_rate": 3.429387739399793e-06,
"loss": 0.0016,
"step": 702
},
{
"epoch": 2.6540642722117203,
"grad_norm": 0.08029157668352127,
"learning_rate": 3.4242904001241313e-06,
"loss": 0.0006,
"step": 703
},
{
"epoch": 2.6578449905482042,
"grad_norm": 0.21342195570468903,
"learning_rate": 3.4191886060429837e-06,
"loss": 0.002,
"step": 704
},
{
"epoch": 2.661625708884688,
"grad_norm": 0.16814370453357697,
"learning_rate": 3.4140823817454826e-06,
"loss": 0.0022,
"step": 705
},
{
"epoch": 2.665406427221172,
"grad_norm": 0.718342661857605,
"learning_rate": 3.4089717518421168e-06,
"loss": 0.004,
"step": 706
},
{
"epoch": 2.6691871455576557,
"grad_norm": 0.32336297631263733,
"learning_rate": 3.403856740964604e-06,
"loss": 0.0037,
"step": 707
},
{
"epoch": 2.67296786389414,
"grad_norm": 0.24353626370429993,
"learning_rate": 3.3987373737657793e-06,
"loss": 0.0029,
"step": 708
},
{
"epoch": 2.6767485822306236,
"grad_norm": 0.30585017800331116,
"learning_rate": 3.393613674919473e-06,
"loss": 0.0019,
"step": 709
},
{
"epoch": 2.6805293005671076,
"grad_norm": 0.06842097640037537,
"learning_rate": 3.388485669120395e-06,
"loss": 0.001,
"step": 710
},
{
"epoch": 2.6843100189035916,
"grad_norm": 0.2304970771074295,
"learning_rate": 3.3833533810840107e-06,
"loss": 0.0019,
"step": 711
},
{
"epoch": 2.6880907372400755,
"grad_norm": 0.20766133069992065,
"learning_rate": 3.3782168355464263e-06,
"loss": 0.0029,
"step": 712
},
{
"epoch": 2.6918714555765595,
"grad_norm": 0.13076448440551758,
"learning_rate": 3.3730760572642656e-06,
"loss": 0.0015,
"step": 713
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.3618428111076355,
"learning_rate": 3.367931071014557e-06,
"loss": 0.0024,
"step": 714
},
{
"epoch": 2.6994328922495274,
"grad_norm": 0.05557963624596596,
"learning_rate": 3.3627819015946062e-06,
"loss": 0.0004,
"step": 715
},
{
"epoch": 2.7032136105860114,
"grad_norm": 0.22211109101772308,
"learning_rate": 3.3576285738218816e-06,
"loss": 0.0042,
"step": 716
},
{
"epoch": 2.7069943289224954,
"grad_norm": 0.35389330983161926,
"learning_rate": 3.3524711125338954e-06,
"loss": 0.0017,
"step": 717
},
{
"epoch": 2.7107750472589793,
"grad_norm": 0.36975187063217163,
"learning_rate": 3.3473095425880795e-06,
"loss": 0.0032,
"step": 718
},
{
"epoch": 2.7145557655954633,
"grad_norm": 0.1723395735025406,
"learning_rate": 3.3421438888616702e-06,
"loss": 0.0025,
"step": 719
},
{
"epoch": 2.718336483931947,
"grad_norm": 0.09432091563940048,
"learning_rate": 3.336974176251586e-06,
"loss": 0.0007,
"step": 720
},
{
"epoch": 2.7221172022684312,
"grad_norm": 0.06567203253507614,
"learning_rate": 3.331800429674308e-06,
"loss": 0.0007,
"step": 721
},
{
"epoch": 2.7258979206049148,
"grad_norm": 0.1482936590909958,
"learning_rate": 3.3266226740657585e-06,
"loss": 0.001,
"step": 722
},
{
"epoch": 2.7296786389413987,
"grad_norm": 0.271511971950531,
"learning_rate": 3.3214409343811844e-06,
"loss": 0.0018,
"step": 723
},
{
"epoch": 2.7334593572778827,
"grad_norm": 0.15263792872428894,
"learning_rate": 3.3162552355950324e-06,
"loss": 0.0012,
"step": 724
},
{
"epoch": 2.7372400756143667,
"grad_norm": 0.21455731987953186,
"learning_rate": 3.3110656027008333e-06,
"loss": 0.0019,
"step": 725
},
{
"epoch": 2.7410207939508506,
"grad_norm": 0.1563577950000763,
"learning_rate": 3.3058720607110772e-06,
"loss": 0.0012,
"step": 726
},
{
"epoch": 2.7448015122873346,
"grad_norm": 0.2714839577674866,
"learning_rate": 3.300674634657094e-06,
"loss": 0.0028,
"step": 727
},
{
"epoch": 2.7485822306238186,
"grad_norm": 0.17852231860160828,
"learning_rate": 3.295473349588938e-06,
"loss": 0.0017,
"step": 728
},
{
"epoch": 2.7523629489603025,
"grad_norm": 0.11730555444955826,
"learning_rate": 3.2902682305752593e-06,
"loss": 0.0008,
"step": 729
},
{
"epoch": 2.7561436672967865,
"grad_norm": 0.3449864685535431,
"learning_rate": 3.2850593027031852e-06,
"loss": 0.006,
"step": 730
},
{
"epoch": 2.7599243856332705,
"grad_norm": 0.10376542061567307,
"learning_rate": 3.2798465910782053e-06,
"loss": 0.0006,
"step": 731
},
{
"epoch": 2.7637051039697544,
"grad_norm": 0.14629311859607697,
"learning_rate": 3.2746301208240398e-06,
"loss": 0.0008,
"step": 732
},
{
"epoch": 2.767485822306238,
"grad_norm": 0.47744810581207275,
"learning_rate": 3.269409917082531e-06,
"loss": 0.0027,
"step": 733
},
{
"epoch": 2.7712665406427224,
"grad_norm": 0.01659102737903595,
"learning_rate": 3.2641860050135097e-06,
"loss": 0.0002,
"step": 734
},
{
"epoch": 2.775047258979206,
"grad_norm": 0.19318439066410065,
"learning_rate": 3.2589584097946823e-06,
"loss": 0.001,
"step": 735
},
{
"epoch": 2.77882797731569,
"grad_norm": 0.14054666459560394,
"learning_rate": 3.253727156621508e-06,
"loss": 0.0014,
"step": 736
},
{
"epoch": 2.782608695652174,
"grad_norm": 0.15233230590820312,
"learning_rate": 3.248492270707074e-06,
"loss": 0.0011,
"step": 737
},
{
"epoch": 2.786389413988658,
"grad_norm": 0.042402613908052444,
"learning_rate": 3.2432537772819767e-06,
"loss": 0.0004,
"step": 738
},
{
"epoch": 2.7901701323251418,
"grad_norm": 0.21248780190944672,
"learning_rate": 3.2380117015942014e-06,
"loss": 0.0027,
"step": 739
},
{
"epoch": 2.7939508506616257,
"grad_norm": 0.14514854550361633,
"learning_rate": 3.2327660689089967e-06,
"loss": 0.001,
"step": 740
},
{
"epoch": 2.7977315689981097,
"grad_norm": 0.17545442283153534,
"learning_rate": 3.2275169045087563e-06,
"loss": 0.001,
"step": 741
},
{
"epoch": 2.8015122873345937,
"grad_norm": 0.1575508713722229,
"learning_rate": 3.2222642336928943e-06,
"loss": 0.0014,
"step": 742
},
{
"epoch": 2.8052930056710776,
"grad_norm": 0.282153457403183,
"learning_rate": 3.217008081777726e-06,
"loss": 0.0037,
"step": 743
},
{
"epoch": 2.809073724007561,
"grad_norm": 0.9071174263954163,
"learning_rate": 3.2117484740963445e-06,
"loss": 0.0012,
"step": 744
},
{
"epoch": 2.8128544423440456,
"grad_norm": 0.4039877951145172,
"learning_rate": 3.2064854359984976e-06,
"loss": 0.0017,
"step": 745
},
{
"epoch": 2.816635160680529,
"grad_norm": 0.3697907030582428,
"learning_rate": 3.201218992850468e-06,
"loss": 0.0049,
"step": 746
},
{
"epoch": 2.820415879017013,
"grad_norm": 0.12910278141498566,
"learning_rate": 3.195949170034948e-06,
"loss": 0.0009,
"step": 747
},
{
"epoch": 2.824196597353497,
"grad_norm": 0.17991717159748077,
"learning_rate": 3.190675992950921e-06,
"loss": 0.001,
"step": 748
},
{
"epoch": 2.827977315689981,
"grad_norm": 0.1861676275730133,
"learning_rate": 3.1853994870135364e-06,
"loss": 0.0018,
"step": 749
},
{
"epoch": 2.831758034026465,
"grad_norm": 0.22559480369091034,
"learning_rate": 3.180119677653986e-06,
"loss": 0.0057,
"step": 750
},
{
"epoch": 2.835538752362949,
"grad_norm": 0.1633092761039734,
"learning_rate": 3.1748365903193844e-06,
"loss": 0.0009,
"step": 751
},
{
"epoch": 2.839319470699433,
"grad_norm": 0.5486204028129578,
"learning_rate": 3.1695502504726468e-06,
"loss": 0.0039,
"step": 752
},
{
"epoch": 2.843100189035917,
"grad_norm": 0.1651940494775772,
"learning_rate": 3.1642606835923608e-06,
"loss": 0.0021,
"step": 753
},
{
"epoch": 2.846880907372401,
"grad_norm": 0.13890020549297333,
"learning_rate": 3.1589679151726693e-06,
"loss": 0.0009,
"step": 754
},
{
"epoch": 2.850661625708885,
"grad_norm": 0.1909310668706894,
"learning_rate": 3.1536719707231473e-06,
"loss": 0.0012,
"step": 755
},
{
"epoch": 2.854442344045369,
"grad_norm": 0.20282648503780365,
"learning_rate": 3.1483728757686745e-06,
"loss": 0.0011,
"step": 756
},
{
"epoch": 2.8582230623818523,
"grad_norm": 0.057108793407678604,
"learning_rate": 3.143070655849317e-06,
"loss": 0.0004,
"step": 757
},
{
"epoch": 2.8620037807183367,
"grad_norm": 0.24721874296665192,
"learning_rate": 3.137765336520201e-06,
"loss": 0.0028,
"step": 758
},
{
"epoch": 2.8657844990548202,
"grad_norm": 0.14667409658432007,
"learning_rate": 3.1324569433513915e-06,
"loss": 0.0013,
"step": 759
},
{
"epoch": 2.869565217391304,
"grad_norm": 0.32723960280418396,
"learning_rate": 3.1271455019277686e-06,
"loss": 0.0026,
"step": 760
},
{
"epoch": 2.873345935727788,
"grad_norm": 0.07179637998342514,
"learning_rate": 3.1218310378489046e-06,
"loss": 0.0005,
"step": 761
},
{
"epoch": 2.877126654064272,
"grad_norm": 0.049112673848867416,
"learning_rate": 3.116513576728939e-06,
"loss": 0.0004,
"step": 762
},
{
"epoch": 2.880907372400756,
"grad_norm": 0.34955427050590515,
"learning_rate": 3.111193144196457e-06,
"loss": 0.0033,
"step": 763
},
{
"epoch": 2.88468809073724,
"grad_norm": 0.4462724030017853,
"learning_rate": 3.1058697658943644e-06,
"loss": 0.0054,
"step": 764
},
{
"epoch": 2.888468809073724,
"grad_norm": 0.174446702003479,
"learning_rate": 3.100543467479766e-06,
"loss": 0.0008,
"step": 765
},
{
"epoch": 2.892249527410208,
"grad_norm": 0.3103788197040558,
"learning_rate": 3.095214274623839e-06,
"loss": 0.0037,
"step": 766
},
{
"epoch": 2.896030245746692,
"grad_norm": 0.03961129114031792,
"learning_rate": 3.089882213011711e-06,
"loss": 0.0004,
"step": 767
},
{
"epoch": 2.899810964083176,
"grad_norm": 0.8424095511436462,
"learning_rate": 3.0845473083423393e-06,
"loss": 0.0007,
"step": 768
},
{
"epoch": 2.90359168241966,
"grad_norm": 0.19955624639987946,
"learning_rate": 3.0792095863283794e-06,
"loss": 0.0017,
"step": 769
},
{
"epoch": 2.9073724007561434,
"grad_norm": 0.06657801568508148,
"learning_rate": 3.0738690726960683e-06,
"loss": 0.0008,
"step": 770
},
{
"epoch": 2.911153119092628,
"grad_norm": 0.09979762136936188,
"learning_rate": 3.068525793185097e-06,
"loss": 0.0007,
"step": 771
},
{
"epoch": 2.9149338374291114,
"grad_norm": 0.07001560181379318,
"learning_rate": 3.0631797735484877e-06,
"loss": 0.0005,
"step": 772
},
{
"epoch": 2.9187145557655954,
"grad_norm": 0.07472414523363113,
"learning_rate": 3.057831039552469e-06,
"loss": 0.0006,
"step": 773
},
{
"epoch": 2.9224952741020793,
"grad_norm": 0.19127531349658966,
"learning_rate": 3.0524796169763513e-06,
"loss": 0.0011,
"step": 774
},
{
"epoch": 2.9262759924385633,
"grad_norm": 0.19040343165397644,
"learning_rate": 3.047125531612404e-06,
"loss": 0.0014,
"step": 775
},
{
"epoch": 2.9300567107750473,
"grad_norm": 0.10263635218143463,
"learning_rate": 3.0417688092657296e-06,
"loss": 0.0008,
"step": 776
},
{
"epoch": 2.9338374291115312,
"grad_norm": 0.16794386506080627,
"learning_rate": 3.03640947575414e-06,
"loss": 0.0025,
"step": 777
},
{
"epoch": 2.937618147448015,
"grad_norm": 0.222567617893219,
"learning_rate": 3.0310475569080346e-06,
"loss": 0.0032,
"step": 778
},
{
"epoch": 2.941398865784499,
"grad_norm": 0.08296474069356918,
"learning_rate": 3.0256830785702682e-06,
"loss": 0.0006,
"step": 779
},
{
"epoch": 2.945179584120983,
"grad_norm": 0.0924961268901825,
"learning_rate": 3.020316066596037e-06,
"loss": 0.0005,
"step": 780
},
{
"epoch": 2.9489603024574667,
"grad_norm": 0.14127139747142792,
"learning_rate": 3.014946546852746e-06,
"loss": 0.001,
"step": 781
},
{
"epoch": 2.952741020793951,
"grad_norm": 0.2660104036331177,
"learning_rate": 3.009574545219887e-06,
"loss": 0.0029,
"step": 782
},
{
"epoch": 2.9565217391304346,
"grad_norm": 0.2899058163166046,
"learning_rate": 3.0042000875889136e-06,
"loss": 0.0022,
"step": 783
},
{
"epoch": 2.9603024574669186,
"grad_norm": 0.12538635730743408,
"learning_rate": 2.998823199863118e-06,
"loss": 0.0008,
"step": 784
},
{
"epoch": 2.9640831758034025,
"grad_norm": 0.3452463150024414,
"learning_rate": 2.993443907957502e-06,
"loss": 0.0028,
"step": 785
},
{
"epoch": 2.9678638941398865,
"grad_norm": 0.17313146591186523,
"learning_rate": 2.988062237798657e-06,
"loss": 0.0012,
"step": 786
},
{
"epoch": 2.9716446124763705,
"grad_norm": 0.26183098554611206,
"learning_rate": 2.982678215324638e-06,
"loss": 0.0029,
"step": 787
},
{
"epoch": 2.9754253308128544,
"grad_norm": 0.06372971832752228,
"learning_rate": 2.9772918664848334e-06,
"loss": 0.0006,
"step": 788
},
{
"epoch": 2.9792060491493384,
"grad_norm": 0.22375071048736572,
"learning_rate": 2.9719032172398484e-06,
"loss": 0.0015,
"step": 789
},
{
"epoch": 2.9829867674858224,
"grad_norm": 0.22053028643131256,
"learning_rate": 2.9665122935613727e-06,
"loss": 0.002,
"step": 790
},
{
"epoch": 2.9867674858223063,
"grad_norm": 0.20250220596790314,
"learning_rate": 2.961119121432059e-06,
"loss": 0.001,
"step": 791
},
{
"epoch": 2.9905482041587903,
"grad_norm": 0.13320282101631165,
"learning_rate": 2.955723726845397e-06,
"loss": 0.001,
"step": 792
},
{
"epoch": 2.9943289224952743,
"grad_norm": 0.16205881536006927,
"learning_rate": 2.950326135805587e-06,
"loss": 0.0026,
"step": 793
},
{
"epoch": 2.998109640831758,
"grad_norm": 0.09543368965387344,
"learning_rate": 2.944926374327418e-06,
"loss": 0.0008,
"step": 794
},
{
"epoch": 3.0,
"grad_norm": 0.10140597820281982,
"learning_rate": 2.9395244684361366e-06,
"loss": 0.0007,
"step": 795
},
{
"epoch": 3.003780718336484,
"grad_norm": 0.13640156388282776,
"learning_rate": 2.9341204441673267e-06,
"loss": 0.0011,
"step": 796
},
{
"epoch": 3.007561436672968,
"grad_norm": 0.03600619360804558,
"learning_rate": 2.928714327566781e-06,
"loss": 0.0004,
"step": 797
},
{
"epoch": 3.011342155009452,
"grad_norm": 0.2623865306377411,
"learning_rate": 2.92330614469038e-06,
"loss": 0.0016,
"step": 798
},
{
"epoch": 3.015122873345936,
"grad_norm": 0.15679103136062622,
"learning_rate": 2.917895921603958e-06,
"loss": 0.002,
"step": 799
},
{
"epoch": 3.01890359168242,
"grad_norm": 0.07765256613492966,
"learning_rate": 2.9124836843831873e-06,
"loss": 0.0007,
"step": 800
},
{
"epoch": 3.022684310018904,
"grad_norm": 0.44910237193107605,
"learning_rate": 2.9070694591134435e-06,
"loss": 0.0024,
"step": 801
},
{
"epoch": 3.0264650283553873,
"grad_norm": 0.18752925097942352,
"learning_rate": 2.901653271889687e-06,
"loss": 0.0024,
"step": 802
},
{
"epoch": 3.0302457466918713,
"grad_norm": 0.2537623345851898,
"learning_rate": 2.896235148816333e-06,
"loss": 0.0019,
"step": 803
},
{
"epoch": 3.0340264650283553,
"grad_norm": 0.07914618402719498,
"learning_rate": 2.8908151160071266e-06,
"loss": 0.0007,
"step": 804
},
{
"epoch": 3.0378071833648392,
"grad_norm": 0.05549820512533188,
"learning_rate": 2.885393199585019e-06,
"loss": 0.0004,
"step": 805
},
{
"epoch": 3.041587901701323,
"grad_norm": 0.2862958610057831,
"learning_rate": 2.879969425682037e-06,
"loss": 0.002,
"step": 806
},
{
"epoch": 3.045368620037807,
"grad_norm": 0.19830483198165894,
"learning_rate": 2.8745438204391623e-06,
"loss": 0.0008,
"step": 807
},
{
"epoch": 3.049149338374291,
"grad_norm": 0.018718168139457703,
"learning_rate": 2.8691164100062035e-06,
"loss": 0.0002,
"step": 808
},
{
"epoch": 3.052930056710775,
"grad_norm": 0.14809346199035645,
"learning_rate": 2.8636872205416666e-06,
"loss": 0.0011,
"step": 809
},
{
"epoch": 3.056710775047259,
"grad_norm": 0.09784464538097382,
"learning_rate": 2.8582562782126356e-06,
"loss": 0.0008,
"step": 810
},
{
"epoch": 3.060491493383743,
"grad_norm": 0.216910719871521,
"learning_rate": 2.8528236091946406e-06,
"loss": 0.0017,
"step": 811
},
{
"epoch": 3.064272211720227,
"grad_norm": 0.05447099730372429,
"learning_rate": 2.847389239671533e-06,
"loss": 0.0007,
"step": 812
},
{
"epoch": 3.068052930056711,
"grad_norm": 0.17010824382305145,
"learning_rate": 2.8419531958353636e-06,
"loss": 0.0017,
"step": 813
},
{
"epoch": 3.0718336483931945,
"grad_norm": 0.175946906208992,
"learning_rate": 2.836515503886249e-06,
"loss": 0.0028,
"step": 814
},
{
"epoch": 3.0756143667296785,
"grad_norm": 0.07030730694532394,
"learning_rate": 2.831076190032252e-06,
"loss": 0.0003,
"step": 815
},
{
"epoch": 3.0793950850661624,
"grad_norm": 0.23115147650241852,
"learning_rate": 2.825635280489252e-06,
"loss": 0.003,
"step": 816
},
{
"epoch": 3.0831758034026464,
"grad_norm": 0.19651320576667786,
"learning_rate": 2.820192801480817e-06,
"loss": 0.0013,
"step": 817
},
{
"epoch": 3.0869565217391304,
"grad_norm": 0.030265018343925476,
"learning_rate": 2.8147487792380835e-06,
"loss": 0.0003,
"step": 818
},
{
"epoch": 3.0907372400756143,
"grad_norm": 0.26177728176116943,
"learning_rate": 2.809303239999621e-06,
"loss": 0.0012,
"step": 819
},
{
"epoch": 3.0945179584120983,
"grad_norm": 0.20297202467918396,
"learning_rate": 2.8038562100113135e-06,
"loss": 0.002,
"step": 820
},
{
"epoch": 3.0982986767485823,
"grad_norm": 0.26535096764564514,
"learning_rate": 2.7984077155262305e-06,
"loss": 0.0017,
"step": 821
},
{
"epoch": 3.1020793950850662,
"grad_norm": 0.04992718622088432,
"learning_rate": 2.7929577828044972e-06,
"loss": 0.0003,
"step": 822
},
{
"epoch": 3.10586011342155,
"grad_norm": 0.2694036066532135,
"learning_rate": 2.787506438113173e-06,
"loss": 0.0016,
"step": 823
},
{
"epoch": 3.109640831758034,
"grad_norm": 0.17114634811878204,
"learning_rate": 2.7820537077261236e-06,
"loss": 0.0011,
"step": 824
},
{
"epoch": 3.113421550094518,
"grad_norm": 0.06140950322151184,
"learning_rate": 2.7765996179238885e-06,
"loss": 0.0004,
"step": 825
},
{
"epoch": 3.1172022684310017,
"grad_norm": 0.19029368460178375,
"learning_rate": 2.771144194993564e-06,
"loss": 0.0007,
"step": 826
},
{
"epoch": 3.1209829867674856,
"grad_norm": 0.030096733942627907,
"learning_rate": 2.7656874652286703e-06,
"loss": 0.0002,
"step": 827
},
{
"epoch": 3.1247637051039696,
"grad_norm": 0.012109048664569855,
"learning_rate": 2.7602294549290242e-06,
"loss": 0.0002,
"step": 828
},
{
"epoch": 3.1285444234404536,
"grad_norm": 0.12381824851036072,
"learning_rate": 2.754770190400618e-06,
"loss": 0.001,
"step": 829
},
{
"epoch": 3.1323251417769375,
"grad_norm": 0.13012413680553436,
"learning_rate": 2.7493096979554858e-06,
"loss": 0.0009,
"step": 830
},
{
"epoch": 3.1361058601134215,
"grad_norm": 0.4504835903644562,
"learning_rate": 2.7438480039115807e-06,
"loss": 0.0061,
"step": 831
},
{
"epoch": 3.1398865784499055,
"grad_norm": 0.15395177900791168,
"learning_rate": 2.7383851345926478e-06,
"loss": 0.0004,
"step": 832
},
{
"epoch": 3.1436672967863895,
"grad_norm": 0.14107590913772583,
"learning_rate": 2.732921116328097e-06,
"loss": 0.0007,
"step": 833
},
{
"epoch": 3.1474480151228734,
"grad_norm": 0.21490655839443207,
"learning_rate": 2.727455975452874e-06,
"loss": 0.0011,
"step": 834
},
{
"epoch": 3.1512287334593574,
"grad_norm": 0.14863553643226624,
"learning_rate": 2.721989738307337e-06,
"loss": 0.0007,
"step": 835
},
{
"epoch": 3.1550094517958414,
"grad_norm": 0.05071425437927246,
"learning_rate": 2.7165224312371276e-06,
"loss": 0.0003,
"step": 836
},
{
"epoch": 3.1587901701323253,
"grad_norm": 0.3041730225086212,
"learning_rate": 2.711054080593043e-06,
"loss": 0.0023,
"step": 837
},
{
"epoch": 3.1625708884688093,
"grad_norm": 0.13519695401191711,
"learning_rate": 2.705584712730911e-06,
"loss": 0.0009,
"step": 838
},
{
"epoch": 3.166351606805293,
"grad_norm": 0.027860593050718307,
"learning_rate": 2.700114354011461e-06,
"loss": 0.0002,
"step": 839
},
{
"epoch": 3.170132325141777,
"grad_norm": 0.1974245011806488,
"learning_rate": 2.694643030800201e-06,
"loss": 0.0027,
"step": 840
},
{
"epoch": 3.1739130434782608,
"grad_norm": 0.6723803281784058,
"learning_rate": 2.6891707694672833e-06,
"loss": 0.0004,
"step": 841
},
{
"epoch": 3.1776937618147447,
"grad_norm": 0.2227015495300293,
"learning_rate": 2.6836975963873846e-06,
"loss": 0.003,
"step": 842
},
{
"epoch": 3.1814744801512287,
"grad_norm": 0.17940707504749298,
"learning_rate": 2.6782235379395766e-06,
"loss": 0.0015,
"step": 843
},
{
"epoch": 3.1852551984877127,
"grad_norm": 2.6559791564941406,
"learning_rate": 2.672748620507195e-06,
"loss": 0.0029,
"step": 844
},
{
"epoch": 3.1890359168241966,
"grad_norm": 0.18533027172088623,
"learning_rate": 2.6672728704777184e-06,
"loss": 0.0031,
"step": 845
},
{
"epoch": 3.1928166351606806,
"grad_norm": 0.028909239917993546,
"learning_rate": 2.6617963142426373e-06,
"loss": 0.0002,
"step": 846
},
{
"epoch": 3.1965973534971646,
"grad_norm": 0.524781346321106,
"learning_rate": 2.656318978197327e-06,
"loss": 0.0019,
"step": 847
},
{
"epoch": 3.2003780718336485,
"grad_norm": 0.14690302312374115,
"learning_rate": 2.650840888740923e-06,
"loss": 0.001,
"step": 848
},
{
"epoch": 3.2041587901701325,
"grad_norm": 0.2171226441860199,
"learning_rate": 2.6453620722761897e-06,
"loss": 0.0007,
"step": 849
},
{
"epoch": 3.207939508506616,
"grad_norm": 0.1151246502995491,
"learning_rate": 2.6398825552093982e-06,
"loss": 0.0015,
"step": 850
},
{
"epoch": 3.2117202268431,
"grad_norm": 0.14347006380558014,
"learning_rate": 2.634402363950195e-06,
"loss": 0.0009,
"step": 851
},
{
"epoch": 3.215500945179584,
"grad_norm": 0.1755191683769226,
"learning_rate": 2.6289215249114753e-06,
"loss": 0.0029,
"step": 852
},
{
"epoch": 3.219281663516068,
"grad_norm": 0.07147383689880371,
"learning_rate": 2.6234400645092576e-06,
"loss": 0.0004,
"step": 853
},
{
"epoch": 3.223062381852552,
"grad_norm": 0.22565598785877228,
"learning_rate": 2.6179580091625543e-06,
"loss": 0.0023,
"step": 854
},
{
"epoch": 3.226843100189036,
"grad_norm": 0.35696810483932495,
"learning_rate": 2.6124753852932457e-06,
"loss": 0.0026,
"step": 855
},
{
"epoch": 3.23062381852552,
"grad_norm": 0.056309353560209274,
"learning_rate": 2.6069922193259523e-06,
"loss": 0.0004,
"step": 856
},
{
"epoch": 3.234404536862004,
"grad_norm": 0.14044925570487976,
"learning_rate": 2.6015085376879063e-06,
"loss": 0.0009,
"step": 857
},
{
"epoch": 3.2381852551984878,
"grad_norm": 0.12725995481014252,
"learning_rate": 2.596024366808827e-06,
"loss": 0.0015,
"step": 858
},
{
"epoch": 3.2419659735349717,
"grad_norm": 0.2350405901670456,
"learning_rate": 2.590539733120791e-06,
"loss": 0.0016,
"step": 859
},
{
"epoch": 3.2457466918714557,
"grad_norm": 0.1963028609752655,
"learning_rate": 2.5850546630581048e-06,
"loss": 0.0013,
"step": 860
},
{
"epoch": 3.2495274102079397,
"grad_norm": 0.11657372117042542,
"learning_rate": 2.579569183057179e-06,
"loss": 0.0005,
"step": 861
},
{
"epoch": 3.2533081285444236,
"grad_norm": 0.030969172716140747,
"learning_rate": 2.5740833195563996e-06,
"loss": 0.0003,
"step": 862
},
{
"epoch": 3.257088846880907,
"grad_norm": 0.2936573624610901,
"learning_rate": 2.5685970989960008e-06,
"loss": 0.0025,
"step": 863
},
{
"epoch": 3.260869565217391,
"grad_norm": 0.16479003429412842,
"learning_rate": 2.5631105478179385e-06,
"loss": 0.0007,
"step": 864
},
{
"epoch": 3.264650283553875,
"grad_norm": 0.10733142495155334,
"learning_rate": 2.55762369246576e-06,
"loss": 0.0006,
"step": 865
},
{
"epoch": 3.268431001890359,
"grad_norm": 0.05159861594438553,
"learning_rate": 2.552136559384482e-06,
"loss": 0.0003,
"step": 866
},
{
"epoch": 3.272211720226843,
"grad_norm": 0.18454059958457947,
"learning_rate": 2.5466491750204564e-06,
"loss": 0.0011,
"step": 867
},
{
"epoch": 3.275992438563327,
"grad_norm": 0.1398025006055832,
"learning_rate": 2.541161565821248e-06,
"loss": 0.0011,
"step": 868
},
{
"epoch": 3.279773156899811,
"grad_norm": 0.0885324478149414,
"learning_rate": 2.535673758235506e-06,
"loss": 0.0004,
"step": 869
},
{
"epoch": 3.283553875236295,
"grad_norm": 0.29177770018577576,
"learning_rate": 2.5301857787128344e-06,
"loss": 0.0019,
"step": 870
},
{
"epoch": 3.287334593572779,
"grad_norm": 0.3835117220878601,
"learning_rate": 2.5246976537036646e-06,
"loss": 0.0017,
"step": 871
},
{
"epoch": 3.291115311909263,
"grad_norm": 0.3790607750415802,
"learning_rate": 2.5192094096591323e-06,
"loss": 0.0023,
"step": 872
},
{
"epoch": 3.294896030245747,
"grad_norm": 0.08609069138765335,
"learning_rate": 2.5137210730309445e-06,
"loss": 0.0004,
"step": 873
},
{
"epoch": 3.298676748582231,
"grad_norm": 0.018595052883028984,
"learning_rate": 2.508232670271257e-06,
"loss": 0.0002,
"step": 874
},
{
"epoch": 3.302457466918715,
"grad_norm": 0.1749623566865921,
"learning_rate": 2.502744227832541e-06,
"loss": 0.0008,
"step": 875
},
{
"epoch": 3.3062381852551983,
"grad_norm": 0.2559996247291565,
"learning_rate": 2.49725577216746e-06,
"loss": 0.0005,
"step": 876
},
{
"epoch": 3.3100189035916823,
"grad_norm": 0.09897667914628983,
"learning_rate": 2.4917673297287443e-06,
"loss": 0.0009,
"step": 877
},
{
"epoch": 3.3137996219281662,
"grad_norm": 0.1699284166097641,
"learning_rate": 2.486278926969056e-06,
"loss": 0.0009,
"step": 878
},
{
"epoch": 3.31758034026465,
"grad_norm": 0.33824798464775085,
"learning_rate": 2.480790590340869e-06,
"loss": 0.0042,
"step": 879
},
{
"epoch": 3.321361058601134,
"grad_norm": 0.19574449956417084,
"learning_rate": 2.4753023462963363e-06,
"loss": 0.0014,
"step": 880
},
{
"epoch": 3.325141776937618,
"grad_norm": 0.058739885687828064,
"learning_rate": 2.4698142212871665e-06,
"loss": 0.0003,
"step": 881
},
{
"epoch": 3.328922495274102,
"grad_norm": 0.03225823491811752,
"learning_rate": 2.4643262417644943e-06,
"loss": 0.0002,
"step": 882
},
{
"epoch": 3.332703213610586,
"grad_norm": 0.039619289338588715,
"learning_rate": 2.458838434178752e-06,
"loss": 0.0002,
"step": 883
},
{
"epoch": 3.33648393194707,
"grad_norm": 0.13394464552402496,
"learning_rate": 2.4533508249795445e-06,
"loss": 0.0008,
"step": 884
},
{
"epoch": 3.340264650283554,
"grad_norm": 0.041759148240089417,
"learning_rate": 2.4478634406155188e-06,
"loss": 0.0003,
"step": 885
},
{
"epoch": 3.344045368620038,
"grad_norm": 0.08425647765398026,
"learning_rate": 2.4423763075342404e-06,
"loss": 0.0008,
"step": 886
},
{
"epoch": 3.3478260869565215,
"grad_norm": 0.05781193822622299,
"learning_rate": 2.4368894521820623e-06,
"loss": 0.0005,
"step": 887
},
{
"epoch": 3.351606805293006,
"grad_norm": 0.010408837348222733,
"learning_rate": 2.4314029010040005e-06,
"loss": 0.0001,
"step": 888
},
{
"epoch": 3.3553875236294894,
"grad_norm": 0.034971896559000015,
"learning_rate": 2.4259166804436008e-06,
"loss": 0.0002,
"step": 889
},
{
"epoch": 3.3591682419659734,
"grad_norm": 0.27988651394844055,
"learning_rate": 2.420430816942821e-06,
"loss": 0.0027,
"step": 890
},
{
"epoch": 3.3629489603024574,
"grad_norm": 0.11298784613609314,
"learning_rate": 2.414945336941896e-06,
"loss": 0.0007,
"step": 891
},
{
"epoch": 3.3667296786389413,
"grad_norm": 0.17393267154693604,
"learning_rate": 2.4094602668792092e-06,
"loss": 0.0012,
"step": 892
},
{
"epoch": 3.3705103969754253,
"grad_norm": 0.020736075937747955,
"learning_rate": 2.403975633191174e-06,
"loss": 0.0002,
"step": 893
},
{
"epoch": 3.3742911153119093,
"grad_norm": 0.20499488711357117,
"learning_rate": 2.3984914623120945e-06,
"loss": 0.0015,
"step": 894
},
{
"epoch": 3.3780718336483933,
"grad_norm": 0.0892459973692894,
"learning_rate": 2.393007780674049e-06,
"loss": 0.0008,
"step": 895
},
{
"epoch": 3.381852551984877,
"grad_norm": 0.14485645294189453,
"learning_rate": 2.387524614706755e-06,
"loss": 0.0009,
"step": 896
},
{
"epoch": 3.385633270321361,
"grad_norm": 0.24436050653457642,
"learning_rate": 2.3820419908374465e-06,
"loss": 0.0008,
"step": 897
},
{
"epoch": 3.389413988657845,
"grad_norm": 0.03550902381539345,
"learning_rate": 2.376559935490743e-06,
"loss": 0.0003,
"step": 898
},
{
"epoch": 3.393194706994329,
"grad_norm": 0.06936739385128021,
"learning_rate": 2.3710784750885247e-06,
"loss": 0.0007,
"step": 899
},
{
"epoch": 3.3969754253308126,
"grad_norm": 0.08910365402698517,
"learning_rate": 2.3655976360498053e-06,
"loss": 0.0009,
"step": 900
},
{
"epoch": 3.4007561436672966,
"grad_norm": 0.1739652007818222,
"learning_rate": 2.3601174447906017e-06,
"loss": 0.0012,
"step": 901
},
{
"epoch": 3.4045368620037806,
"grad_norm": 0.05551964044570923,
"learning_rate": 2.3546379277238107e-06,
"loss": 0.0005,
"step": 902
},
{
"epoch": 3.4083175803402646,
"grad_norm": 0.05429123714566231,
"learning_rate": 2.3491591112590776e-06,
"loss": 0.0003,
"step": 903
},
{
"epoch": 3.4120982986767485,
"grad_norm": 0.050854556262493134,
"learning_rate": 2.3436810218026744e-06,
"loss": 0.0004,
"step": 904
},
{
"epoch": 3.4158790170132325,
"grad_norm": 0.031142516061663628,
"learning_rate": 2.3382036857573635e-06,
"loss": 0.0004,
"step": 905
},
{
"epoch": 3.4196597353497165,
"grad_norm": 0.18706762790679932,
"learning_rate": 2.3327271295222824e-06,
"loss": 0.0035,
"step": 906
},
{
"epoch": 3.4234404536862004,
"grad_norm": 0.18152789771556854,
"learning_rate": 2.3272513794928055e-06,
"loss": 0.0011,
"step": 907
},
{
"epoch": 3.4272211720226844,
"grad_norm": 0.031270433217287064,
"learning_rate": 2.3217764620604234e-06,
"loss": 0.0002,
"step": 908
},
{
"epoch": 3.4310018903591684,
"grad_norm": 0.08286745101213455,
"learning_rate": 2.316302403612616e-06,
"loss": 0.0007,
"step": 909
},
{
"epoch": 3.4347826086956523,
"grad_norm": 0.35267341136932373,
"learning_rate": 2.3108292305327176e-06,
"loss": 0.003,
"step": 910
},
{
"epoch": 3.4385633270321363,
"grad_norm": 0.4903669059276581,
"learning_rate": 2.3053569691998003e-06,
"loss": 0.0036,
"step": 911
},
{
"epoch": 3.4423440453686203,
"grad_norm": 0.4287071228027344,
"learning_rate": 2.2998856459885397e-06,
"loss": 0.0017,
"step": 912
},
{
"epoch": 3.446124763705104,
"grad_norm": 0.172041654586792,
"learning_rate": 2.29441528726909e-06,
"loss": 0.0012,
"step": 913
},
{
"epoch": 3.4499054820415878,
"grad_norm": 0.08805087953805923,
"learning_rate": 2.288945919406958e-06,
"loss": 0.0008,
"step": 914
},
{
"epoch": 3.4536862003780717,
"grad_norm": 0.04952337220311165,
"learning_rate": 2.2834775687628732e-06,
"loss": 0.0005,
"step": 915
},
{
"epoch": 3.4574669187145557,
"grad_norm": 0.022441262379288673,
"learning_rate": 2.278010261692663e-06,
"loss": 0.0002,
"step": 916
},
{
"epoch": 3.4612476370510397,
"grad_norm": 0.05646921694278717,
"learning_rate": 2.272544024547126e-06,
"loss": 0.0005,
"step": 917
},
{
"epoch": 3.4650283553875236,
"grad_norm": 0.20496559143066406,
"learning_rate": 2.267078883671904e-06,
"loss": 0.0012,
"step": 918
},
{
"epoch": 3.4688090737240076,
"grad_norm": 0.07436178624629974,
"learning_rate": 2.2616148654073527e-06,
"loss": 0.0004,
"step": 919
},
{
"epoch": 3.4725897920604916,
"grad_norm": 0.08458953350782394,
"learning_rate": 2.2561519960884205e-06,
"loss": 0.001,
"step": 920
},
{
"epoch": 3.4763705103969755,
"grad_norm": 0.15990053117275238,
"learning_rate": 2.2506903020445146e-06,
"loss": 0.0008,
"step": 921
},
{
"epoch": 3.4801512287334595,
"grad_norm": 0.352651983499527,
"learning_rate": 2.245229809599383e-06,
"loss": 0.0028,
"step": 922
},
{
"epoch": 3.4839319470699435,
"grad_norm": 0.23983141779899597,
"learning_rate": 2.2397705450709766e-06,
"loss": 0.001,
"step": 923
},
{
"epoch": 3.487712665406427,
"grad_norm": 0.04083551466464996,
"learning_rate": 2.2343125347713305e-06,
"loss": 0.0003,
"step": 924
},
{
"epoch": 3.491493383742911,
"grad_norm": 0.13663507997989655,
"learning_rate": 2.2288558050064366e-06,
"loss": 0.0015,
"step": 925
},
{
"epoch": 3.495274102079395,
"grad_norm": 0.0250815711915493,
"learning_rate": 2.2234003820761123e-06,
"loss": 0.0002,
"step": 926
},
{
"epoch": 3.499054820415879,
"grad_norm": 0.0626959577202797,
"learning_rate": 2.2179462922738777e-06,
"loss": 0.0003,
"step": 927
},
{
"epoch": 3.502835538752363,
"grad_norm": 0.2886582612991333,
"learning_rate": 2.212493561886827e-06,
"loss": 0.001,
"step": 928
},
{
"epoch": 3.506616257088847,
"grad_norm": 0.1235971450805664,
"learning_rate": 2.207042217195503e-06,
"loss": 0.0007,
"step": 929
},
{
"epoch": 3.510396975425331,
"grad_norm": 0.012822418473660946,
"learning_rate": 2.2015922844737703e-06,
"loss": 0.0002,
"step": 930
},
{
"epoch": 3.5141776937618148,
"grad_norm": 3.3414323329925537,
"learning_rate": 2.196143789988687e-06,
"loss": 0.0036,
"step": 931
},
{
"epoch": 3.5179584120982987,
"grad_norm": 0.11547354608774185,
"learning_rate": 2.1906967600003796e-06,
"loss": 0.0011,
"step": 932
},
{
"epoch": 3.5217391304347827,
"grad_norm": 0.22768059372901917,
"learning_rate": 2.185251220761917e-06,
"loss": 0.0008,
"step": 933
},
{
"epoch": 3.5255198487712667,
"grad_norm": 0.0345073863863945,
"learning_rate": 2.1798071985191833e-06,
"loss": 0.0004,
"step": 934
},
{
"epoch": 3.52930056710775,
"grad_norm": 0.024516673758625984,
"learning_rate": 2.174364719510748e-06,
"loss": 0.0002,
"step": 935
},
{
"epoch": 3.5330812854442346,
"grad_norm": 0.028897911310195923,
"learning_rate": 2.1689238099677482e-06,
"loss": 0.0002,
"step": 936
},
{
"epoch": 3.536862003780718,
"grad_norm": 0.10890685021877289,
"learning_rate": 2.1634844961137517e-06,
"loss": 0.0006,
"step": 937
},
{
"epoch": 3.5406427221172025,
"grad_norm": 0.11050219088792801,
"learning_rate": 2.1580468041646377e-06,
"loss": 0.0006,
"step": 938
},
{
"epoch": 3.544423440453686,
"grad_norm": 0.1272352635860443,
"learning_rate": 2.1526107603284675e-06,
"loss": 0.0007,
"step": 939
},
{
"epoch": 3.54820415879017,
"grad_norm": 0.2891867458820343,
"learning_rate": 2.147176390805361e-06,
"loss": 0.0008,
"step": 940
},
{
"epoch": 3.551984877126654,
"grad_norm": 0.06173471733927727,
"learning_rate": 2.1417437217873653e-06,
"loss": 0.0003,
"step": 941
},
{
"epoch": 3.555765595463138,
"grad_norm": 0.17375104129314423,
"learning_rate": 2.1363127794583334e-06,
"loss": 0.0011,
"step": 942
},
{
"epoch": 3.559546313799622,
"grad_norm": 0.18656082451343536,
"learning_rate": 2.1308835899937974e-06,
"loss": 0.0012,
"step": 943
},
{
"epoch": 3.563327032136106,
"grad_norm": 0.10677118599414825,
"learning_rate": 2.1254561795608377e-06,
"loss": 0.0006,
"step": 944
},
{
"epoch": 3.56710775047259,
"grad_norm": 0.205882728099823,
"learning_rate": 2.1200305743179637e-06,
"loss": 0.0016,
"step": 945
},
{
"epoch": 3.570888468809074,
"grad_norm": 0.17761781811714172,
"learning_rate": 2.1146068004149815e-06,
"loss": 0.0011,
"step": 946
},
{
"epoch": 3.574669187145558,
"grad_norm": 0.18611370027065277,
"learning_rate": 2.109184883992874e-06,
"loss": 0.0009,
"step": 947
},
{
"epoch": 3.5784499054820413,
"grad_norm": 0.14028282463550568,
"learning_rate": 2.1037648511836674e-06,
"loss": 0.0006,
"step": 948
},
{
"epoch": 3.5822306238185257,
"grad_norm": 0.08887307345867157,
"learning_rate": 2.098346728110314e-06,
"loss": 0.0004,
"step": 949
},
{
"epoch": 3.5860113421550093,
"grad_norm": 0.1713380068540573,
"learning_rate": 2.0929305408865573e-06,
"loss": 0.0007,
"step": 950
},
{
"epoch": 3.5897920604914932,
"grad_norm": 0.0907624289393425,
"learning_rate": 2.0875163156168127e-06,
"loss": 0.0004,
"step": 951
},
{
"epoch": 3.593572778827977,
"grad_norm": 0.32220035791397095,
"learning_rate": 2.0821040783960422e-06,
"loss": 0.0012,
"step": 952
},
{
"epoch": 3.597353497164461,
"grad_norm": 0.02104608714580536,
"learning_rate": 2.0766938553096204e-06,
"loss": 0.0002,
"step": 953
},
{
"epoch": 3.601134215500945,
"grad_norm": 0.17797155678272247,
"learning_rate": 2.0712856724332193e-06,
"loss": 0.0013,
"step": 954
},
{
"epoch": 3.604914933837429,
"grad_norm": 0.1625778079032898,
"learning_rate": 2.0658795558326745e-06,
"loss": 0.0008,
"step": 955
},
{
"epoch": 3.608695652173913,
"grad_norm": 0.050108522176742554,
"learning_rate": 2.0604755315638646e-06,
"loss": 0.0003,
"step": 956
},
{
"epoch": 3.612476370510397,
"grad_norm": 0.06516983360052109,
"learning_rate": 2.0550736256725827e-06,
"loss": 0.0004,
"step": 957
},
{
"epoch": 3.616257088846881,
"grad_norm": 0.013083672150969505,
"learning_rate": 2.0496738641944132e-06,
"loss": 0.0001,
"step": 958
},
{
"epoch": 3.620037807183365,
"grad_norm": 0.18956373631954193,
"learning_rate": 2.0442762731546033e-06,
"loss": 0.0009,
"step": 959
},
{
"epoch": 3.623818525519849,
"grad_norm": 0.1957995742559433,
"learning_rate": 2.038880878567941e-06,
"loss": 0.0013,
"step": 960
},
{
"epoch": 3.6275992438563325,
"grad_norm": 0.24313993752002716,
"learning_rate": 2.0334877064386277e-06,
"loss": 0.0007,
"step": 961
},
{
"epoch": 3.631379962192817,
"grad_norm": 0.1375015825033188,
"learning_rate": 2.028096782760152e-06,
"loss": 0.0009,
"step": 962
},
{
"epoch": 3.6351606805293004,
"grad_norm": 0.0393967442214489,
"learning_rate": 2.022708133515168e-06,
"loss": 0.0002,
"step": 963
},
{
"epoch": 3.6389413988657844,
"grad_norm": 0.17369344830513,
"learning_rate": 2.017321784675363e-06,
"loss": 0.001,
"step": 964
},
{
"epoch": 3.6427221172022684,
"grad_norm": 0.06712903827428818,
"learning_rate": 2.011937762201344e-06,
"loss": 0.0005,
"step": 965
},
{
"epoch": 3.6465028355387523,
"grad_norm": 0.20035861432552338,
"learning_rate": 2.0065560920424994e-06,
"loss": 0.0014,
"step": 966
},
{
"epoch": 3.6502835538752363,
"grad_norm": 0.06654126197099686,
"learning_rate": 2.0011768001368836e-06,
"loss": 0.0003,
"step": 967
},
{
"epoch": 3.6540642722117203,
"grad_norm": 0.2571471333503723,
"learning_rate": 1.9957999124110872e-06,
"loss": 0.0013,
"step": 968
},
{
"epoch": 3.6578449905482042,
"grad_norm": 0.14951904118061066,
"learning_rate": 1.9904254547801134e-06,
"loss": 0.0009,
"step": 969
},
{
"epoch": 3.661625708884688,
"grad_norm": 0.06377850472927094,
"learning_rate": 1.9850534531472544e-06,
"loss": 0.0004,
"step": 970
},
{
"epoch": 3.665406427221172,
"grad_norm": 0.04174032434821129,
"learning_rate": 1.979683933403963e-06,
"loss": 0.0002,
"step": 971
},
{
"epoch": 3.6691871455576557,
"grad_norm": 0.2826174199581146,
"learning_rate": 1.9743169214297326e-06,
"loss": 0.0013,
"step": 972
},
{
"epoch": 3.67296786389414,
"grad_norm": 0.9276911616325378,
"learning_rate": 1.9689524430919667e-06,
"loss": 0.0008,
"step": 973
},
{
"epoch": 3.6767485822306236,
"grad_norm": 0.08827374875545502,
"learning_rate": 1.9635905242458607e-06,
"loss": 0.0005,
"step": 974
},
{
"epoch": 3.6805293005671076,
"grad_norm": 0.2523500621318817,
"learning_rate": 1.958231190734271e-06,
"loss": 0.0015,
"step": 975
},
{
"epoch": 3.6843100189035916,
"grad_norm": 0.01701711304485798,
"learning_rate": 1.9528744683875973e-06,
"loss": 0.0002,
"step": 976
},
{
"epoch": 3.6880907372400755,
"grad_norm": 0.2163919359445572,
"learning_rate": 1.947520383023649e-06,
"loss": 0.0011,
"step": 977
},
{
"epoch": 3.6918714555765595,
"grad_norm": 0.10253291577100754,
"learning_rate": 1.942168960447531e-06,
"loss": 0.0008,
"step": 978
},
{
"epoch": 3.6956521739130435,
"grad_norm": 0.11419998109340668,
"learning_rate": 1.9368202264515127e-06,
"loss": 0.0002,
"step": 979
},
{
"epoch": 3.6994328922495274,
"grad_norm": 0.02250787802040577,
"learning_rate": 1.931474206814903e-06,
"loss": 0.0002,
"step": 980
},
{
"epoch": 3.7032136105860114,
"grad_norm": 0.08538919687271118,
"learning_rate": 1.926130927303933e-06,
"loss": 0.0006,
"step": 981
},
{
"epoch": 3.7069943289224954,
"grad_norm": 0.06906883418560028,
"learning_rate": 1.9207904136716215e-06,
"loss": 0.0004,
"step": 982
},
{
"epoch": 3.7107750472589793,
"grad_norm": 0.2188524454832077,
"learning_rate": 1.915452691657662e-06,
"loss": 0.0014,
"step": 983
},
{
"epoch": 3.7145557655954633,
"grad_norm": 0.2538103759288788,
"learning_rate": 1.9101177869882893e-06,
"loss": 0.0027,
"step": 984
},
{
"epoch": 3.718336483931947,
"grad_norm": 0.40686437487602234,
"learning_rate": 1.9047857253761622e-06,
"loss": 0.0053,
"step": 985
},
{
"epoch": 3.7221172022684312,
"grad_norm": 0.3122040927410126,
"learning_rate": 1.8994565325202346e-06,
"loss": 0.0007,
"step": 986
},
{
"epoch": 3.7258979206049148,
"grad_norm": 0.08798515051603317,
"learning_rate": 1.8941302341056356e-06,
"loss": 0.0004,
"step": 987
},
{
"epoch": 3.7296786389413987,
"grad_norm": 0.28242072463035583,
"learning_rate": 1.8888068558035435e-06,
"loss": 0.002,
"step": 988
},
{
"epoch": 3.7334593572778827,
"grad_norm": 0.07137515395879745,
"learning_rate": 1.8834864232710614e-06,
"loss": 0.0006,
"step": 989
},
{
"epoch": 3.7372400756143667,
"grad_norm": 0.09302951395511627,
"learning_rate": 1.8781689621510962e-06,
"loss": 0.0004,
"step": 990
},
{
"epoch": 3.7410207939508506,
"grad_norm": 0.06287765502929688,
"learning_rate": 1.8728544980722318e-06,
"loss": 0.0002,
"step": 991
},
{
"epoch": 3.7448015122873346,
"grad_norm": 0.23519393801689148,
"learning_rate": 1.8675430566486102e-06,
"loss": 0.0004,
"step": 992
},
{
"epoch": 3.7485822306238186,
"grad_norm": 0.17336173355579376,
"learning_rate": 1.8622346634798e-06,
"loss": 0.0006,
"step": 993
},
{
"epoch": 3.7523629489603025,
"grad_norm": 0.17733007669448853,
"learning_rate": 1.8569293441506847e-06,
"loss": 0.0005,
"step": 994
},
{
"epoch": 3.7561436672967865,
"grad_norm": 0.05134982615709305,
"learning_rate": 1.8516271242313267e-06,
"loss": 0.0003,
"step": 995
},
{
"epoch": 3.7599243856332705,
"grad_norm": 0.014910437166690826,
"learning_rate": 1.8463280292768529e-06,
"loss": 0.0002,
"step": 996
},
{
"epoch": 3.7637051039697544,
"grad_norm": 0.04615437984466553,
"learning_rate": 1.8410320848273315e-06,
"loss": 0.0004,
"step": 997
},
{
"epoch": 3.767485822306238,
"grad_norm": 0.013714149594306946,
"learning_rate": 1.8357393164076403e-06,
"loss": 0.0002,
"step": 998
},
{
"epoch": 3.7712665406427224,
"grad_norm": 0.14500373601913452,
"learning_rate": 1.8304497495273543e-06,
"loss": 0.0004,
"step": 999
},
{
"epoch": 3.775047258979206,
"grad_norm": 0.09468162804841995,
"learning_rate": 1.8251634096806158e-06,
"loss": 0.0007,
"step": 1000
},
{
"epoch": 3.77882797731569,
"grad_norm": 0.32062816619873047,
"learning_rate": 1.8198803223460148e-06,
"loss": 0.0022,
"step": 1001
},
{
"epoch": 3.782608695652174,
"grad_norm": 0.17651709914207458,
"learning_rate": 1.8146005129864644e-06,
"loss": 0.0005,
"step": 1002
},
{
"epoch": 3.786389413988658,
"grad_norm": 0.3067803382873535,
"learning_rate": 1.8093240070490791e-06,
"loss": 0.0015,
"step": 1003
},
{
"epoch": 3.7901701323251418,
"grad_norm": 0.20555636286735535,
"learning_rate": 1.8040508299650524e-06,
"loss": 0.0014,
"step": 1004
},
{
"epoch": 3.7939508506616257,
"grad_norm": 0.41955652832984924,
"learning_rate": 1.7987810071495327e-06,
"loss": 0.0023,
"step": 1005
},
{
"epoch": 3.7977315689981097,
"grad_norm": 0.11061103641986847,
"learning_rate": 1.793514564001503e-06,
"loss": 0.0005,
"step": 1006
},
{
"epoch": 3.8015122873345937,
"grad_norm": 0.028922656551003456,
"learning_rate": 1.788251525903656e-06,
"loss": 0.0003,
"step": 1007
},
{
"epoch": 3.8052930056710776,
"grad_norm": 0.27011823654174805,
"learning_rate": 1.7829919182222752e-06,
"loss": 0.0026,
"step": 1008
},
{
"epoch": 3.809073724007561,
"grad_norm": 0.18338823318481445,
"learning_rate": 1.7777357663071063e-06,
"loss": 0.002,
"step": 1009
},
{
"epoch": 3.8128544423440456,
"grad_norm": 0.2944319546222687,
"learning_rate": 1.7724830954912454e-06,
"loss": 0.0007,
"step": 1010
},
{
"epoch": 3.816635160680529,
"grad_norm": 0.06173175200819969,
"learning_rate": 1.767233931091004e-06,
"loss": 0.0003,
"step": 1011
},
{
"epoch": 3.820415879017013,
"grad_norm": 0.03877865895628929,
"learning_rate": 1.7619882984057988e-06,
"loss": 0.0004,
"step": 1012
},
{
"epoch": 3.824196597353497,
"grad_norm": 0.05438617616891861,
"learning_rate": 1.7567462227180241e-06,
"loss": 0.0003,
"step": 1013
},
{
"epoch": 3.827977315689981,
"grad_norm": 0.37924039363861084,
"learning_rate": 1.7515077292929267e-06,
"loss": 0.0016,
"step": 1014
},
{
"epoch": 3.831758034026465,
"grad_norm": 0.07925333082675934,
"learning_rate": 1.746272843378493e-06,
"loss": 0.0003,
"step": 1015
},
{
"epoch": 3.835538752362949,
"grad_norm": 0.43481874465942383,
"learning_rate": 1.7410415902053179e-06,
"loss": 0.0066,
"step": 1016
},
{
"epoch": 3.839319470699433,
"grad_norm": 0.05568358302116394,
"learning_rate": 1.7358139949864913e-06,
"loss": 0.0004,
"step": 1017
},
{
"epoch": 3.843100189035917,
"grad_norm": 0.13646312057971954,
"learning_rate": 1.73059008291747e-06,
"loss": 0.001,
"step": 1018
},
{
"epoch": 3.846880907372401,
"grad_norm": 0.03888719156384468,
"learning_rate": 1.7253698791759604e-06,
"loss": 0.0003,
"step": 1019
},
{
"epoch": 3.850661625708885,
"grad_norm": 0.12511153519153595,
"learning_rate": 1.720153408921796e-06,
"loss": 0.0004,
"step": 1020
},
{
"epoch": 3.854442344045369,
"grad_norm": 0.055930424481630325,
"learning_rate": 1.7149406972968148e-06,
"loss": 0.0003,
"step": 1021
},
{
"epoch": 3.8582230623818523,
"grad_norm": 0.257201611995697,
"learning_rate": 1.7097317694247418e-06,
"loss": 0.006,
"step": 1022
},
{
"epoch": 3.8620037807183367,
"grad_norm": 0.05855580419301987,
"learning_rate": 1.704526650411062e-06,
"loss": 0.0004,
"step": 1023
},
{
"epoch": 3.8657844990548202,
"grad_norm": 0.07037494331598282,
"learning_rate": 1.6993253653429064e-06,
"loss": 0.0004,
"step": 1024
},
{
"epoch": 3.869565217391304,
"grad_norm": 0.5156080722808838,
"learning_rate": 1.6941279392889238e-06,
"loss": 0.0037,
"step": 1025
},
{
"epoch": 3.873345935727788,
"grad_norm": 1.0936269760131836,
"learning_rate": 1.6889343972991682e-06,
"loss": 0.0017,
"step": 1026
},
{
"epoch": 3.877126654064272,
"grad_norm": 0.06437109410762787,
"learning_rate": 1.6837447644049682e-06,
"loss": 0.0003,
"step": 1027
},
{
"epoch": 3.880907372400756,
"grad_norm": 0.006052945740520954,
"learning_rate": 1.678559065618817e-06,
"loss": 0.0001,
"step": 1028
},
{
"epoch": 3.88468809073724,
"grad_norm": 0.022189904004335403,
"learning_rate": 1.6733773259342424e-06,
"loss": 0.0002,
"step": 1029
},
{
"epoch": 3.888468809073724,
"grad_norm": 0.11311095207929611,
"learning_rate": 1.6681995703256925e-06,
"loss": 0.0007,
"step": 1030
},
{
"epoch": 3.892249527410208,
"grad_norm": 0.10053358972072601,
"learning_rate": 1.6630258237484144e-06,
"loss": 0.0003,
"step": 1031
},
{
"epoch": 3.896030245746692,
"grad_norm": 0.10737577825784683,
"learning_rate": 1.6578561111383298e-06,
"loss": 0.0004,
"step": 1032
},
{
"epoch": 3.899810964083176,
"grad_norm": 0.38978490233421326,
"learning_rate": 1.6526904574119213e-06,
"loss": 0.0017,
"step": 1033
},
{
"epoch": 3.90359168241966,
"grad_norm": 0.21502213180065155,
"learning_rate": 1.6475288874661053e-06,
"loss": 0.0018,
"step": 1034
},
{
"epoch": 3.9073724007561434,
"grad_norm": 0.013299698010087013,
"learning_rate": 1.6423714261781196e-06,
"loss": 0.0001,
"step": 1035
},
{
"epoch": 3.911153119092628,
"grad_norm": 0.0985122099518776,
"learning_rate": 1.6372180984053946e-06,
"loss": 0.0008,
"step": 1036
},
{
"epoch": 3.9149338374291114,
"grad_norm": 0.1399480402469635,
"learning_rate": 1.632068928985444e-06,
"loss": 0.0008,
"step": 1037
},
{
"epoch": 3.9187145557655954,
"grad_norm": 0.05103133246302605,
"learning_rate": 1.6269239427357348e-06,
"loss": 0.0004,
"step": 1038
},
{
"epoch": 3.9224952741020793,
"grad_norm": 0.17584289610385895,
"learning_rate": 1.6217831644535743e-06,
"loss": 0.0014,
"step": 1039
},
{
"epoch": 3.9262759924385633,
"grad_norm": 0.11641789227724075,
"learning_rate": 1.6166466189159901e-06,
"loss": 0.0009,
"step": 1040
},
{
"epoch": 3.9300567107750473,
"grad_norm": 0.5359205007553101,
"learning_rate": 1.6115143308796054e-06,
"loss": 0.0017,
"step": 1041
},
{
"epoch": 3.9338374291115312,
"grad_norm": 0.06901949644088745,
"learning_rate": 1.6063863250805279e-06,
"loss": 0.0006,
"step": 1042
},
{
"epoch": 3.937618147448015,
"grad_norm": 0.148621067404747,
"learning_rate": 1.6012626262342217e-06,
"loss": 0.0007,
"step": 1043
},
{
"epoch": 3.941398865784499,
"grad_norm": 0.11555682867765427,
"learning_rate": 1.5961432590353973e-06,
"loss": 0.0013,
"step": 1044
},
{
"epoch": 3.945179584120983,
"grad_norm": 0.0854395255446434,
"learning_rate": 1.5910282481578843e-06,
"loss": 0.0005,
"step": 1045
},
{
"epoch": 3.9489603024574667,
"grad_norm": 0.021322818472981453,
"learning_rate": 1.5859176182545178e-06,
"loss": 0.0002,
"step": 1046
},
{
"epoch": 3.952741020793951,
"grad_norm": 0.1617099493741989,
"learning_rate": 1.5808113939570175e-06,
"loss": 0.0011,
"step": 1047
},
{
"epoch": 3.9565217391304346,
"grad_norm": 0.1138630211353302,
"learning_rate": 1.5757095998758693e-06,
"loss": 0.0011,
"step": 1048
},
{
"epoch": 3.9603024574669186,
"grad_norm": 0.18166384100914001,
"learning_rate": 1.5706122606002078e-06,
"loss": 0.0027,
"step": 1049
},
{
"epoch": 3.9640831758034025,
"grad_norm": 0.04953674599528313,
"learning_rate": 1.5655194006976948e-06,
"loss": 0.0004,
"step": 1050
},
{
"epoch": 3.9678638941398865,
"grad_norm": 0.26992782950401306,
"learning_rate": 1.5604310447144052e-06,
"loss": 0.0007,
"step": 1051
},
{
"epoch": 3.9716446124763705,
"grad_norm": 0.028692858293652534,
"learning_rate": 1.5553472171747031e-06,
"loss": 0.0003,
"step": 1052
},
{
"epoch": 3.9754253308128544,
"grad_norm": 0.11208011955022812,
"learning_rate": 1.5502679425811317e-06,
"loss": 0.0005,
"step": 1053
},
{
"epoch": 3.9792060491493384,
"grad_norm": 0.04451301321387291,
"learning_rate": 1.5451932454142847e-06,
"loss": 0.0003,
"step": 1054
},
{
"epoch": 3.9829867674858224,
"grad_norm": 0.15611913800239563,
"learning_rate": 1.5401231501326985e-06,
"loss": 0.0008,
"step": 1055
},
{
"epoch": 3.9867674858223063,
"grad_norm": 0.022839756682515144,
"learning_rate": 1.535057681172727e-06,
"loss": 0.0002,
"step": 1056
},
{
"epoch": 3.9905482041587903,
"grad_norm": 0.16348697245121002,
"learning_rate": 1.5299968629484277e-06,
"loss": 0.0008,
"step": 1057
},
{
"epoch": 3.9943289224952743,
"grad_norm": 0.018139483407139778,
"learning_rate": 1.524940719851444e-06,
"loss": 0.0001,
"step": 1058
},
{
"epoch": 3.998109640831758,
"grad_norm": 0.045861661434173584,
"learning_rate": 1.5198892762508843e-06,
"loss": 0.0003,
"step": 1059
},
{
"epoch": 4.0,
"grad_norm": 0.03191845491528511,
"learning_rate": 1.5148425564932085e-06,
"loss": 0.0002,
"step": 1060
},
{
"epoch": 4.0037807183364835,
"grad_norm": 0.1410960704088211,
"learning_rate": 1.509800584902108e-06,
"loss": 0.0008,
"step": 1061
},
{
"epoch": 4.007561436672968,
"grad_norm": 0.05028901994228363,
"learning_rate": 1.504763385778391e-06,
"loss": 0.0003,
"step": 1062
},
{
"epoch": 4.0113421550094515,
"grad_norm": 0.8279242515563965,
"learning_rate": 1.4997309833998608e-06,
"loss": 0.0003,
"step": 1063
},
{
"epoch": 4.015122873345936,
"grad_norm": 0.15353278815746307,
"learning_rate": 1.4947034020212062e-06,
"loss": 0.0011,
"step": 1064
},
{
"epoch": 4.018903591682419,
"grad_norm": 0.23645780980587006,
"learning_rate": 1.4896806658738747e-06,
"loss": 0.0018,
"step": 1065
},
{
"epoch": 4.022684310018904,
"grad_norm": 0.07189466804265976,
"learning_rate": 1.4846627991659662e-06,
"loss": 0.0004,
"step": 1066
},
{
"epoch": 4.026465028355387,
"grad_norm": 0.09000851958990097,
"learning_rate": 1.4796498260821104e-06,
"loss": 0.0004,
"step": 1067
},
{
"epoch": 4.030245746691872,
"grad_norm": 0.05539466440677643,
"learning_rate": 1.4746417707833472e-06,
"loss": 0.0004,
"step": 1068
},
{
"epoch": 4.034026465028355,
"grad_norm": 0.07071535289287567,
"learning_rate": 1.4696386574070203e-06,
"loss": 0.0004,
"step": 1069
},
{
"epoch": 4.03780718336484,
"grad_norm": 0.03256288170814514,
"learning_rate": 1.4646405100666498e-06,
"loss": 0.0003,
"step": 1070
},
{
"epoch": 4.041587901701323,
"grad_norm": 0.1477307826280594,
"learning_rate": 1.4596473528518236e-06,
"loss": 0.001,
"step": 1071
},
{
"epoch": 4.045368620037808,
"grad_norm": 0.11566184461116791,
"learning_rate": 1.4546592098280772e-06,
"loss": 0.0005,
"step": 1072
},
{
"epoch": 4.049149338374291,
"grad_norm": 0.031164227053523064,
"learning_rate": 1.449676105036781e-06,
"loss": 0.0002,
"step": 1073
},
{
"epoch": 4.052930056710775,
"grad_norm": 0.41042619943618774,
"learning_rate": 1.4446980624950201e-06,
"loss": 0.0007,
"step": 1074
},
{
"epoch": 4.056710775047259,
"grad_norm": 0.23981332778930664,
"learning_rate": 1.4397251061954848e-06,
"loss": 0.0014,
"step": 1075
},
{
"epoch": 4.060491493383743,
"grad_norm": 0.06305637955665588,
"learning_rate": 1.4347572601063477e-06,
"loss": 0.0006,
"step": 1076
},
{
"epoch": 4.064272211720227,
"grad_norm": 0.5029850006103516,
"learning_rate": 1.4297945481711527e-06,
"loss": 0.0008,
"step": 1077
},
{
"epoch": 4.0680529300567105,
"grad_norm": 0.21877774596214294,
"learning_rate": 1.4248369943086997e-06,
"loss": 0.0007,
"step": 1078
},
{
"epoch": 4.071833648393195,
"grad_norm": 0.08199340105056763,
"learning_rate": 1.4198846224129255e-06,
"loss": 0.0004,
"step": 1079
},
{
"epoch": 4.0756143667296785,
"grad_norm": 0.025111818686127663,
"learning_rate": 1.4149374563527957e-06,
"loss": 0.0002,
"step": 1080
},
{
"epoch": 4.079395085066163,
"grad_norm": 0.013005358166992664,
"learning_rate": 1.4099955199721816e-06,
"loss": 0.0001,
"step": 1081
},
{
"epoch": 4.083175803402646,
"grad_norm": 0.041962917894124985,
"learning_rate": 1.4050588370897493e-06,
"loss": 0.0003,
"step": 1082
},
{
"epoch": 4.086956521739131,
"grad_norm": 0.20963317155838013,
"learning_rate": 1.4001274314988476e-06,
"loss": 0.0009,
"step": 1083
},
{
"epoch": 4.090737240075614,
"grad_norm": 0.06594974547624588,
"learning_rate": 1.3952013269673872e-06,
"loss": 0.0006,
"step": 1084
},
{
"epoch": 4.094517958412098,
"grad_norm": 0.20333515107631683,
"learning_rate": 1.3902805472377307e-06,
"loss": 0.0014,
"step": 1085
},
{
"epoch": 4.098298676748582,
"grad_norm": 0.01559424214065075,
"learning_rate": 1.3853651160265752e-06,
"loss": 0.0002,
"step": 1086
},
{
"epoch": 4.102079395085066,
"grad_norm": 0.00993192009627819,
"learning_rate": 1.3804550570248431e-06,
"loss": 0.0001,
"step": 1087
},
{
"epoch": 4.10586011342155,
"grad_norm": 0.033074699342250824,
"learning_rate": 1.375550393897559e-06,
"loss": 0.0002,
"step": 1088
},
{
"epoch": 4.109640831758034,
"grad_norm": 0.46920835971832275,
"learning_rate": 1.370651150283746e-06,
"loss": 0.0032,
"step": 1089
},
{
"epoch": 4.113421550094518,
"grad_norm": 0.034286096692085266,
"learning_rate": 1.3657573497963042e-06,
"loss": 0.0002,
"step": 1090
},
{
"epoch": 4.117202268431002,
"grad_norm": 0.042510658502578735,
"learning_rate": 1.360869016021899e-06,
"loss": 0.0004,
"step": 1091
},
{
"epoch": 4.120982986767486,
"grad_norm": 0.015354749746620655,
"learning_rate": 1.3559861725208501e-06,
"loss": 0.0001,
"step": 1092
},
{
"epoch": 4.12476370510397,
"grad_norm": 0.11133018881082535,
"learning_rate": 1.3511088428270141e-06,
"loss": 0.0008,
"step": 1093
},
{
"epoch": 4.128544423440454,
"grad_norm": 0.0947665274143219,
"learning_rate": 1.3462370504476729e-06,
"loss": 0.0007,
"step": 1094
},
{
"epoch": 4.1323251417769375,
"grad_norm": 0.19103285670280457,
"learning_rate": 1.3413708188634192e-06,
"loss": 0.0011,
"step": 1095
},
{
"epoch": 4.136105860113422,
"grad_norm": 0.015423089265823364,
"learning_rate": 1.3365101715280473e-06,
"loss": 0.0001,
"step": 1096
},
{
"epoch": 4.1398865784499055,
"grad_norm": 0.04986581578850746,
"learning_rate": 1.3316551318684342e-06,
"loss": 0.0003,
"step": 1097
},
{
"epoch": 4.143667296786389,
"grad_norm": 0.22506923973560333,
"learning_rate": 1.3268057232844305e-06,
"loss": 0.001,
"step": 1098
},
{
"epoch": 4.147448015122873,
"grad_norm": 0.14979712665081024,
"learning_rate": 1.321961969148746e-06,
"loss": 0.0007,
"step": 1099
},
{
"epoch": 4.151228733459357,
"grad_norm": 0.08199136704206467,
"learning_rate": 1.3171238928068403e-06,
"loss": 0.0004,
"step": 1100
},
{
"epoch": 4.155009451795841,
"grad_norm": 0.06098520755767822,
"learning_rate": 1.3122915175768044e-06,
"loss": 0.0003,
"step": 1101
},
{
"epoch": 4.158790170132325,
"grad_norm": 0.02282070182263851,
"learning_rate": 1.307464866749253e-06,
"loss": 0.0002,
"step": 1102
},
{
"epoch": 4.162570888468809,
"grad_norm": 0.10259565711021423,
"learning_rate": 1.3026439635872132e-06,
"loss": 0.0004,
"step": 1103
},
{
"epoch": 4.166351606805293,
"grad_norm": 0.23500141501426697,
"learning_rate": 1.2978288313260045e-06,
"loss": 0.0002,
"step": 1104
},
{
"epoch": 4.170132325141777,
"grad_norm": 0.18757376074790955,
"learning_rate": 1.2930194931731382e-06,
"loss": 0.0012,
"step": 1105
},
{
"epoch": 4.173913043478261,
"grad_norm": 0.061834666877985,
"learning_rate": 1.2882159723081967e-06,
"loss": 0.0002,
"step": 1106
},
{
"epoch": 4.177693761814745,
"grad_norm": 0.157730370759964,
"learning_rate": 1.2834182918827248e-06,
"loss": 0.0005,
"step": 1107
},
{
"epoch": 4.181474480151229,
"grad_norm": 0.14092953503131866,
"learning_rate": 1.2786264750201182e-06,
"loss": 0.0007,
"step": 1108
},
{
"epoch": 4.185255198487713,
"grad_norm": 0.2177150845527649,
"learning_rate": 1.2738405448155139e-06,
"loss": 0.001,
"step": 1109
},
{
"epoch": 4.189035916824197,
"grad_norm": 0.09654669463634491,
"learning_rate": 1.269060524335675e-06,
"loss": 0.0007,
"step": 1110
},
{
"epoch": 4.19281663516068,
"grad_norm": 0.12041306495666504,
"learning_rate": 1.2642864366188806e-06,
"loss": 0.0005,
"step": 1111
},
{
"epoch": 4.196597353497165,
"grad_norm": 0.05897996574640274,
"learning_rate": 1.2595183046748187e-06,
"loss": 0.0003,
"step": 1112
},
{
"epoch": 4.200378071833648,
"grad_norm": 0.02310796268284321,
"learning_rate": 1.2547561514844706e-06,
"loss": 0.0002,
"step": 1113
},
{
"epoch": 4.2041587901701325,
"grad_norm": 0.1802576780319214,
"learning_rate": 1.2500000000000007e-06,
"loss": 0.001,
"step": 1114
},
{
"epoch": 4.207939508506616,
"grad_norm": 0.09202495962381363,
"learning_rate": 1.245249873144647e-06,
"loss": 0.0009,
"step": 1115
},
{
"epoch": 4.2117202268431,
"grad_norm": 0.1162513941526413,
"learning_rate": 1.2405057938126152e-06,
"loss": 0.0006,
"step": 1116
},
{
"epoch": 4.215500945179584,
"grad_norm": 0.20377464592456818,
"learning_rate": 1.235767784868956e-06,
"loss": 0.0007,
"step": 1117
},
{
"epoch": 4.219281663516068,
"grad_norm": 0.16681860387325287,
"learning_rate": 1.231035869149468e-06,
"loss": 0.0007,
"step": 1118
},
{
"epoch": 4.223062381852552,
"grad_norm": 0.04913189262151718,
"learning_rate": 1.2263100694605837e-06,
"loss": 0.0003,
"step": 1119
},
{
"epoch": 4.226843100189036,
"grad_norm": 0.043777771294116974,
"learning_rate": 1.221590408579251e-06,
"loss": 0.0003,
"step": 1120
},
{
"epoch": 4.23062381852552,
"grad_norm": 0.19229836761951447,
"learning_rate": 1.2168769092528382e-06,
"loss": 0.0006,
"step": 1121
},
{
"epoch": 4.234404536862003,
"grad_norm": 0.18815527856349945,
"learning_rate": 1.2121695941990126e-06,
"loss": 0.0015,
"step": 1122
},
{
"epoch": 4.238185255198488,
"grad_norm": 0.02042614109814167,
"learning_rate": 1.207468486105636e-06,
"loss": 0.0001,
"step": 1123
},
{
"epoch": 4.241965973534971,
"grad_norm": 0.4299137592315674,
"learning_rate": 1.2027736076306533e-06,
"loss": 0.0023,
"step": 1124
},
{
"epoch": 4.245746691871456,
"grad_norm": 0.18861791491508484,
"learning_rate": 1.1980849814019881e-06,
"loss": 0.0019,
"step": 1125
},
{
"epoch": 4.249527410207939,
"grad_norm": 0.03366239741444588,
"learning_rate": 1.1934026300174265e-06,
"loss": 0.0002,
"step": 1126
},
{
"epoch": 4.253308128544424,
"grad_norm": 0.05086700618267059,
"learning_rate": 1.1887265760445122e-06,
"loss": 0.0002,
"step": 1127
},
{
"epoch": 4.257088846880907,
"grad_norm": 0.015901854261755943,
"learning_rate": 1.1840568420204393e-06,
"loss": 0.0001,
"step": 1128
},
{
"epoch": 4.260869565217392,
"grad_norm": 0.3603084683418274,
"learning_rate": 1.17939345045194e-06,
"loss": 0.0059,
"step": 1129
},
{
"epoch": 4.264650283553875,
"grad_norm": 0.16659899055957794,
"learning_rate": 1.1747364238151774e-06,
"loss": 0.0009,
"step": 1130
},
{
"epoch": 4.2684310018903595,
"grad_norm": 0.025750087574124336,
"learning_rate": 1.1700857845556374e-06,
"loss": 0.0002,
"step": 1131
},
{
"epoch": 4.272211720226843,
"grad_norm": 0.016796128824353218,
"learning_rate": 1.1654415550880245e-06,
"loss": 0.0002,
"step": 1132
},
{
"epoch": 4.2759924385633274,
"grad_norm": 0.014483626000583172,
"learning_rate": 1.1608037577961423e-06,
"loss": 0.0001,
"step": 1133
},
{
"epoch": 4.279773156899811,
"grad_norm": 0.13883107900619507,
"learning_rate": 1.156172415032801e-06,
"loss": 0.0008,
"step": 1134
},
{
"epoch": 4.2835538752362945,
"grad_norm": 0.12188896536827087,
"learning_rate": 1.1515475491196975e-06,
"loss": 0.0006,
"step": 1135
},
{
"epoch": 4.287334593572779,
"grad_norm": 0.4809183180332184,
"learning_rate": 1.1469291823473127e-06,
"loss": 0.0027,
"step": 1136
},
{
"epoch": 4.291115311909262,
"grad_norm": 0.4040108621120453,
"learning_rate": 1.1423173369748056e-06,
"loss": 0.0009,
"step": 1137
},
{
"epoch": 4.294896030245747,
"grad_norm": 0.010789219290018082,
"learning_rate": 1.1377120352299015e-06,
"loss": 0.0001,
"step": 1138
},
{
"epoch": 4.29867674858223,
"grad_norm": 0.02084983140230179,
"learning_rate": 1.1331132993087887e-06,
"loss": 0.0002,
"step": 1139
},
{
"epoch": 4.302457466918715,
"grad_norm": 0.061494551599025726,
"learning_rate": 1.1285211513760088e-06,
"loss": 0.0005,
"step": 1140
},
{
"epoch": 4.306238185255198,
"grad_norm": 0.027528882026672363,
"learning_rate": 1.1239356135643544e-06,
"loss": 0.0002,
"step": 1141
},
{
"epoch": 4.310018903591683,
"grad_norm": 0.16053670644760132,
"learning_rate": 1.119356707974756e-06,
"loss": 0.0022,
"step": 1142
},
{
"epoch": 4.313799621928166,
"grad_norm": 0.0065573155879974365,
"learning_rate": 1.11478445667618e-06,
"loss": 0.0001,
"step": 1143
},
{
"epoch": 4.317580340264651,
"grad_norm": 0.0785282552242279,
"learning_rate": 1.1102188817055202e-06,
"loss": 0.0004,
"step": 1144
},
{
"epoch": 4.321361058601134,
"grad_norm": 0.0666346326470375,
"learning_rate": 1.105660005067495e-06,
"loss": 0.0004,
"step": 1145
},
{
"epoch": 4.325141776937619,
"grad_norm": 0.2753641903400421,
"learning_rate": 1.1011078487345362e-06,
"loss": 0.0015,
"step": 1146
},
{
"epoch": 4.328922495274102,
"grad_norm": 0.11533135175704956,
"learning_rate": 1.096562434646686e-06,
"loss": 0.0007,
"step": 1147
},
{
"epoch": 4.332703213610586,
"grad_norm": 0.16574276983737946,
"learning_rate": 1.0920237847114945e-06,
"loss": 0.0009,
"step": 1148
},
{
"epoch": 4.33648393194707,
"grad_norm": 0.45153912901878357,
"learning_rate": 1.0874919208039036e-06,
"loss": 0.0033,
"step": 1149
},
{
"epoch": 4.340264650283554,
"grad_norm": 0.038779936730861664,
"learning_rate": 1.0829668647661558e-06,
"loss": 0.0002,
"step": 1150
},
{
"epoch": 4.344045368620038,
"grad_norm": 1.6438484191894531,
"learning_rate": 1.0784486384076778e-06,
"loss": 0.0034,
"step": 1151
},
{
"epoch": 4.3478260869565215,
"grad_norm": 0.1677672415971756,
"learning_rate": 1.0739372635049796e-06,
"loss": 0.002,
"step": 1152
},
{
"epoch": 4.351606805293006,
"grad_norm": 0.21162758767604828,
"learning_rate": 1.0694327618015492e-06,
"loss": 0.0014,
"step": 1153
},
{
"epoch": 4.355387523629489,
"grad_norm": 0.10020237416028976,
"learning_rate": 1.06493515500775e-06,
"loss": 0.0006,
"step": 1154
},
{
"epoch": 4.359168241965974,
"grad_norm": 0.2266089767217636,
"learning_rate": 1.0604444648007114e-06,
"loss": 0.003,
"step": 1155
},
{
"epoch": 4.362948960302457,
"grad_norm": 0.07692354917526245,
"learning_rate": 1.0559607128242267e-06,
"loss": 0.0004,
"step": 1156
},
{
"epoch": 4.366729678638942,
"grad_norm": 0.11673581600189209,
"learning_rate": 1.0514839206886519e-06,
"loss": 0.0005,
"step": 1157
},
{
"epoch": 4.370510396975425,
"grad_norm": 0.03876563906669617,
"learning_rate": 1.0470141099707958e-06,
"loss": 0.0003,
"step": 1158
},
{
"epoch": 4.374291115311909,
"grad_norm": 0.16376714408397675,
"learning_rate": 1.0425513022138203e-06,
"loss": 0.0009,
"step": 1159
},
{
"epoch": 4.378071833648393,
"grad_norm": 0.07068619877099991,
"learning_rate": 1.0380955189271336e-06,
"loss": 0.0004,
"step": 1160
},
{
"epoch": 4.381852551984877,
"grad_norm": 0.1028299331665039,
"learning_rate": 1.0336467815862905e-06,
"loss": 0.0012,
"step": 1161
},
{
"epoch": 4.385633270321361,
"grad_norm": 0.08210930973291397,
"learning_rate": 1.0292051116328844e-06,
"loss": 0.0005,
"step": 1162
},
{
"epoch": 4.389413988657845,
"grad_norm": 0.09313171356916428,
"learning_rate": 1.0247705304744458e-06,
"loss": 0.0005,
"step": 1163
},
{
"epoch": 4.393194706994329,
"grad_norm": 0.06403695791959763,
"learning_rate": 1.0203430594843426e-06,
"loss": 0.0005,
"step": 1164
},
{
"epoch": 4.396975425330813,
"grad_norm": 0.11548646539449692,
"learning_rate": 1.0159227200016677e-06,
"loss": 0.0004,
"step": 1165
},
{
"epoch": 4.400756143667297,
"grad_norm": 0.025654347613453865,
"learning_rate": 1.0115095333311487e-06,
"loss": 0.0001,
"step": 1166
},
{
"epoch": 4.404536862003781,
"grad_norm": 0.2518896162509918,
"learning_rate": 1.0071035207430352e-06,
"loss": 0.0014,
"step": 1167
},
{
"epoch": 4.408317580340265,
"grad_norm": 0.09397520869970322,
"learning_rate": 1.002704703473e-06,
"loss": 0.0005,
"step": 1168
},
{
"epoch": 4.4120982986767485,
"grad_norm": 0.11244060099124908,
"learning_rate": 9.983131027220373e-07,
"loss": 0.0005,
"step": 1169
},
{
"epoch": 4.415879017013232,
"grad_norm": 0.05858241766691208,
"learning_rate": 9.939287396563604e-07,
"loss": 0.0004,
"step": 1170
},
{
"epoch": 4.4196597353497165,
"grad_norm": 0.17480628192424774,
"learning_rate": 9.895516354072981e-07,
"loss": 0.0011,
"step": 1171
},
{
"epoch": 4.4234404536862,
"grad_norm": 0.12224049121141434,
"learning_rate": 9.85181811071193e-07,
"loss": 0.0006,
"step": 1172
},
{
"epoch": 4.427221172022684,
"grad_norm": 0.039147842675447464,
"learning_rate": 9.808192877093039e-07,
"loss": 0.0002,
"step": 1173
},
{
"epoch": 4.431001890359168,
"grad_norm": 0.04021478816866875,
"learning_rate": 9.764640863476982e-07,
"loss": 0.0002,
"step": 1174
},
{
"epoch": 4.434782608695652,
"grad_norm": 0.06257487833499908,
"learning_rate": 9.72116227977154e-07,
"loss": 0.0002,
"step": 1175
},
{
"epoch": 4.438563327032136,
"grad_norm": 0.05208725109696388,
"learning_rate": 9.677757335530577e-07,
"loss": 0.0002,
"step": 1176
},
{
"epoch": 4.44234404536862,
"grad_norm": 0.05006309598684311,
"learning_rate": 9.634426239953074e-07,
"loss": 0.0003,
"step": 1177
},
{
"epoch": 4.446124763705104,
"grad_norm": 0.054312851279973984,
"learning_rate": 9.59116920188202e-07,
"loss": 0.0003,
"step": 1178
},
{
"epoch": 4.449905482041588,
"grad_norm": 0.07308554649353027,
"learning_rate": 9.547986429803518e-07,
"loss": 0.0003,
"step": 1179
},
{
"epoch": 4.453686200378072,
"grad_norm": 0.21417483687400818,
"learning_rate": 9.504878131845738e-07,
"loss": 0.0009,
"step": 1180
},
{
"epoch": 4.457466918714556,
"grad_norm": 0.01975918561220169,
"learning_rate": 9.461844515777849e-07,
"loss": 0.0001,
"step": 1181
},
{
"epoch": 4.46124763705104,
"grad_norm": 0.27177906036376953,
"learning_rate": 9.418885789009139e-07,
"loss": 0.0019,
"step": 1182
},
{
"epoch": 4.465028355387524,
"grad_norm": 0.08854660391807556,
"learning_rate": 9.376002158587915e-07,
"loss": 0.0005,
"step": 1183
},
{
"epoch": 4.468809073724008,
"grad_norm": 0.19646334648132324,
"learning_rate": 9.333193831200557e-07,
"loss": 0.0007,
"step": 1184
},
{
"epoch": 4.472589792060491,
"grad_norm": 0.25866439938545227,
"learning_rate": 9.290461013170485e-07,
"loss": 0.0012,
"step": 1185
},
{
"epoch": 4.4763705103969755,
"grad_norm": 0.05048421025276184,
"learning_rate": 9.247803910457226e-07,
"loss": 0.0003,
"step": 1186
},
{
"epoch": 4.480151228733459,
"grad_norm": 0.16501876711845398,
"learning_rate": 9.205222728655344e-07,
"loss": 0.0008,
"step": 1187
},
{
"epoch": 4.4839319470699435,
"grad_norm": 0.045157257467508316,
"learning_rate": 9.162717672993499e-07,
"loss": 0.0004,
"step": 1188
},
{
"epoch": 4.487712665406427,
"grad_norm": 0.1297757625579834,
"learning_rate": 9.120288948333464e-07,
"loss": 0.0008,
"step": 1189
},
{
"epoch": 4.491493383742911,
"grad_norm": 0.011051077395677567,
"learning_rate": 9.07793675916909e-07,
"loss": 0.0001,
"step": 1190
},
{
"epoch": 4.495274102079395,
"grad_norm": 0.20022499561309814,
"learning_rate": 9.035661309625366e-07,
"loss": 0.0011,
"step": 1191
},
{
"epoch": 4.499054820415879,
"grad_norm": 0.006567796692252159,
"learning_rate": 8.993462803457403e-07,
"loss": 0.0001,
"step": 1192
},
{
"epoch": 4.502835538752363,
"grad_norm": 0.061368178576231,
"learning_rate": 8.951341444049513e-07,
"loss": 0.0003,
"step": 1193
},
{
"epoch": 4.506616257088847,
"grad_norm": 0.09380974620580673,
"learning_rate": 8.90929743441411e-07,
"loss": 0.0004,
"step": 1194
},
{
"epoch": 4.510396975425331,
"grad_norm": 0.16075399518013,
"learning_rate": 8.867330977190877e-07,
"loss": 0.0011,
"step": 1195
},
{
"epoch": 4.514177693761814,
"grad_norm": 0.08030504733324051,
"learning_rate": 8.825442274645671e-07,
"loss": 0.0003,
"step": 1196
},
{
"epoch": 4.517958412098299,
"grad_norm": 0.1006259024143219,
"learning_rate": 8.783631528669607e-07,
"loss": 0.0004,
"step": 1197
},
{
"epoch": 4.521739130434782,
"grad_norm": 0.2878094017505646,
"learning_rate": 8.741898940778088e-07,
"loss": 0.0014,
"step": 1198
},
{
"epoch": 4.525519848771267,
"grad_norm": 0.08849186450242996,
"learning_rate": 8.700244712109798e-07,
"loss": 0.0004,
"step": 1199
},
{
"epoch": 4.52930056710775,
"grad_norm": 0.1939917802810669,
"learning_rate": 8.658669043425763e-07,
"loss": 0.0018,
"step": 1200
},
{
"epoch": 4.533081285444235,
"grad_norm": 0.019765350967645645,
"learning_rate": 8.617172135108356e-07,
"loss": 0.0002,
"step": 1201
},
{
"epoch": 4.536862003780718,
"grad_norm": 0.0034924966748803854,
"learning_rate": 8.575754187160385e-07,
"loss": 0.0001,
"step": 1202
},
{
"epoch": 4.5406427221172025,
"grad_norm": 0.07683593779802322,
"learning_rate": 8.53441539920406e-07,
"loss": 0.0003,
"step": 1203
},
{
"epoch": 4.544423440453686,
"grad_norm": 0.7124339938163757,
"learning_rate": 8.493155970480074e-07,
"loss": 0.0017,
"step": 1204
},
{
"epoch": 4.5482041587901705,
"grad_norm": 0.12068726867437363,
"learning_rate": 8.45197609984662e-07,
"loss": 0.0004,
"step": 1205
},
{
"epoch": 4.551984877126654,
"grad_norm": 0.02341364324092865,
"learning_rate": 8.410875985778477e-07,
"loss": 0.0002,
"step": 1206
},
{
"epoch": 4.5557655954631375,
"grad_norm": 0.2545037567615509,
"learning_rate": 8.369855826365988e-07,
"loss": 0.0011,
"step": 1207
},
{
"epoch": 4.559546313799622,
"grad_norm": 0.03259057179093361,
"learning_rate": 8.328915819314149e-07,
"loss": 0.0002,
"step": 1208
},
{
"epoch": 4.563327032136106,
"grad_norm": 0.1389022320508957,
"learning_rate": 8.288056161941663e-07,
"loss": 0.0009,
"step": 1209
},
{
"epoch": 4.56710775047259,
"grad_norm": 0.1874510645866394,
"learning_rate": 8.247277051179925e-07,
"loss": 0.0001,
"step": 1210
},
{
"epoch": 4.570888468809073,
"grad_norm": 0.08008835464715958,
"learning_rate": 8.206578683572178e-07,
"loss": 0.0004,
"step": 1211
},
{
"epoch": 4.574669187145558,
"grad_norm": 0.01835073158144951,
"learning_rate": 8.165961255272467e-07,
"loss": 0.0001,
"step": 1212
},
{
"epoch": 4.578449905482041,
"grad_norm": 0.0242909025400877,
"learning_rate": 8.125424962044742e-07,
"loss": 0.0002,
"step": 1213
},
{
"epoch": 4.582230623818526,
"grad_norm": 0.05294472351670265,
"learning_rate": 8.084969999261907e-07,
"loss": 0.0003,
"step": 1214
},
{
"epoch": 4.586011342155009,
"grad_norm": 0.022286243736743927,
"learning_rate": 8.044596561904894e-07,
"loss": 0.0001,
"step": 1215
},
{
"epoch": 4.589792060491494,
"grad_norm": 0.014280243776738644,
"learning_rate": 8.004304844561684e-07,
"loss": 0.0001,
"step": 1216
},
{
"epoch": 4.593572778827977,
"grad_norm": 0.006016590166836977,
"learning_rate": 7.964095041426387e-07,
"loss": 0.0001,
"step": 1217
},
{
"epoch": 4.597353497164462,
"grad_norm": 0.0817185491323471,
"learning_rate": 7.923967346298345e-07,
"loss": 0.0007,
"step": 1218
},
{
"epoch": 4.601134215500945,
"grad_norm": 0.0028400784358382225,
"learning_rate": 7.883921952581131e-07,
"loss": 0.0001,
"step": 1219
},
{
"epoch": 4.60491493383743,
"grad_norm": 0.09645567089319229,
"learning_rate": 7.843959053281663e-07,
"loss": 0.0005,
"step": 1220
},
{
"epoch": 4.608695652173913,
"grad_norm": 0.0480729341506958,
"learning_rate": 7.804078841009252e-07,
"loss": 0.0003,
"step": 1221
},
{
"epoch": 4.612476370510397,
"grad_norm": 0.2541545331478119,
"learning_rate": 7.764281507974711e-07,
"loss": 0.001,
"step": 1222
},
{
"epoch": 4.616257088846881,
"grad_norm": 0.3360908329486847,
"learning_rate": 7.724567245989342e-07,
"loss": 0.0016,
"step": 1223
},
{
"epoch": 4.6200378071833645,
"grad_norm": 0.397393137216568,
"learning_rate": 7.684936246464125e-07,
"loss": 0.0032,
"step": 1224
},
{
"epoch": 4.623818525519849,
"grad_norm": 0.2237563133239746,
"learning_rate": 7.645388700408732e-07,
"loss": 0.0006,
"step": 1225
},
{
"epoch": 4.6275992438563325,
"grad_norm": 0.2880302369594574,
"learning_rate": 7.60592479843057e-07,
"loss": 0.0006,
"step": 1226
},
{
"epoch": 4.631379962192817,
"grad_norm": 0.0548987090587616,
"learning_rate": 7.566544730733961e-07,
"loss": 0.0002,
"step": 1227
},
{
"epoch": 4.6351606805293,
"grad_norm": 0.15247085690498352,
"learning_rate": 7.527248687119137e-07,
"loss": 0.0008,
"step": 1228
},
{
"epoch": 4.638941398865785,
"grad_norm": 0.02119935303926468,
"learning_rate": 7.488036856981373e-07,
"loss": 0.0002,
"step": 1229
},
{
"epoch": 4.642722117202268,
"grad_norm": 0.0966901108622551,
"learning_rate": 7.448909429310047e-07,
"loss": 0.0004,
"step": 1230
},
{
"epoch": 4.646502835538753,
"grad_norm": 0.4204804301261902,
"learning_rate": 7.409866592687768e-07,
"loss": 0.0031,
"step": 1231
},
{
"epoch": 4.650283553875236,
"grad_norm": 0.24543237686157227,
"learning_rate": 7.370908535289409e-07,
"loss": 0.0011,
"step": 1232
},
{
"epoch": 4.65406427221172,
"grad_norm": 0.003815317992120981,
"learning_rate": 7.332035444881247e-07,
"loss": 0.0001,
"step": 1233
},
{
"epoch": 4.657844990548204,
"grad_norm": 0.11462830752134323,
"learning_rate": 7.293247508820059e-07,
"loss": 0.0005,
"step": 1234
},
{
"epoch": 4.661625708884688,
"grad_norm": 0.2189127504825592,
"learning_rate": 7.254544914052178e-07,
"loss": 0.0018,
"step": 1235
},
{
"epoch": 4.665406427221172,
"grad_norm": 0.021884597837924957,
"learning_rate": 7.215927847112628e-07,
"loss": 0.0002,
"step": 1236
},
{
"epoch": 4.669187145557656,
"grad_norm": 0.04036756604909897,
"learning_rate": 7.177396494124206e-07,
"loss": 0.0003,
"step": 1237
},
{
"epoch": 4.67296786389414,
"grad_norm": 0.022519908845424652,
"learning_rate": 7.138951040796627e-07,
"loss": 0.0002,
"step": 1238
},
{
"epoch": 4.676748582230624,
"grad_norm": 0.11803640425205231,
"learning_rate": 7.100591672425533e-07,
"loss": 0.0006,
"step": 1239
},
{
"epoch": 4.680529300567108,
"grad_norm": 0.0525188073515892,
"learning_rate": 7.062318573891716e-07,
"loss": 0.0004,
"step": 1240
},
{
"epoch": 4.6843100189035916,
"grad_norm": 0.14717620611190796,
"learning_rate": 7.024131929660172e-07,
"loss": 0.0008,
"step": 1241
},
{
"epoch": 4.688090737240076,
"grad_norm": 0.11175039410591125,
"learning_rate": 6.986031923779166e-07,
"loss": 0.0005,
"step": 1242
},
{
"epoch": 4.6918714555765595,
"grad_norm": 0.7206166386604309,
"learning_rate": 6.948018739879439e-07,
"loss": 0.0029,
"step": 1243
},
{
"epoch": 4.695652173913043,
"grad_norm": 0.0034987463150173426,
"learning_rate": 6.910092561173256e-07,
"loss": 0.0001,
"step": 1244
},
{
"epoch": 4.699432892249527,
"grad_norm": 0.2348700314760208,
"learning_rate": 6.872253570453546e-07,
"loss": 0.001,
"step": 1245
},
{
"epoch": 4.703213610586012,
"grad_norm": 0.025731360539793968,
"learning_rate": 6.834501950093006e-07,
"loss": 0.0002,
"step": 1246
},
{
"epoch": 4.706994328922495,
"grad_norm": 0.2358698844909668,
"learning_rate": 6.796837882043261e-07,
"loss": 0.0011,
"step": 1247
},
{
"epoch": 4.710775047258979,
"grad_norm": 0.27968379855155945,
"learning_rate": 6.75926154783394e-07,
"loss": 0.0026,
"step": 1248
},
{
"epoch": 4.714555765595463,
"grad_norm": 0.04314443841576576,
"learning_rate": 6.721773128571812e-07,
"loss": 0.0003,
"step": 1249
},
{
"epoch": 4.718336483931947,
"grad_norm": 0.15619735419750214,
"learning_rate": 6.684372804939954e-07,
"loss": 0.0009,
"step": 1250
},
{
"epoch": 4.722117202268431,
"grad_norm": 0.07120101153850555,
"learning_rate": 6.647060757196818e-07,
"loss": 0.0005,
"step": 1251
},
{
"epoch": 4.725897920604915,
"grad_norm": 0.03989715874195099,
"learning_rate": 6.609837165175408e-07,
"loss": 0.0003,
"step": 1252
},
{
"epoch": 4.729678638941399,
"grad_norm": 0.14208559691905975,
"learning_rate": 6.57270220828238e-07,
"loss": 0.0007,
"step": 1253
},
{
"epoch": 4.733459357277883,
"grad_norm": 0.024005800485610962,
"learning_rate": 6.535656065497237e-07,
"loss": 0.0002,
"step": 1254
},
{
"epoch": 4.737240075614367,
"grad_norm": 0.24465472996234894,
"learning_rate": 6.49869891537136e-07,
"loss": 0.0007,
"step": 1255
},
{
"epoch": 4.741020793950851,
"grad_norm": 0.07883734256029129,
"learning_rate": 6.461830936027272e-07,
"loss": 0.0002,
"step": 1256
},
{
"epoch": 4.744801512287335,
"grad_norm": 0.0853784829378128,
"learning_rate": 6.425052305157683e-07,
"loss": 0.0004,
"step": 1257
},
{
"epoch": 4.748582230623819,
"grad_norm": 0.037580493837594986,
"learning_rate": 6.38836320002468e-07,
"loss": 0.0004,
"step": 1258
},
{
"epoch": 4.752362948960302,
"grad_norm": 0.04207819327712059,
"learning_rate": 6.351763797458879e-07,
"loss": 0.0003,
"step": 1259
},
{
"epoch": 4.7561436672967865,
"grad_norm": 0.1094450056552887,
"learning_rate": 6.315254273858525e-07,
"loss": 0.0005,
"step": 1260
},
{
"epoch": 4.75992438563327,
"grad_norm": 0.1623501181602478,
"learning_rate": 6.278834805188699e-07,
"loss": 0.0012,
"step": 1261
},
{
"epoch": 4.763705103969754,
"grad_norm": 0.044797904789447784,
"learning_rate": 6.242505566980423e-07,
"loss": 0.0003,
"step": 1262
},
{
"epoch": 4.767485822306238,
"grad_norm": 0.008014979772269726,
"learning_rate": 6.20626673432986e-07,
"loss": 0.0001,
"step": 1263
},
{
"epoch": 4.771266540642722,
"grad_norm": 0.17691481113433838,
"learning_rate": 6.170118481897422e-07,
"loss": 0.0009,
"step": 1264
},
{
"epoch": 4.775047258979206,
"grad_norm": 0.10001453012228012,
"learning_rate": 6.134060983906964e-07,
"loss": 0.0006,
"step": 1265
},
{
"epoch": 4.77882797731569,
"grad_norm": 0.029246237128973007,
"learning_rate": 6.09809441414492e-07,
"loss": 0.0002,
"step": 1266
},
{
"epoch": 4.782608695652174,
"grad_norm": 0.3046590983867645,
"learning_rate": 6.062218945959497e-07,
"loss": 0.003,
"step": 1267
},
{
"epoch": 4.786389413988658,
"grad_norm": 0.06525318324565887,
"learning_rate": 6.026434752259802e-07,
"loss": 0.0003,
"step": 1268
},
{
"epoch": 4.790170132325142,
"grad_norm": 0.03284177929162979,
"learning_rate": 5.990742005515024e-07,
"loss": 0.0003,
"step": 1269
},
{
"epoch": 4.793950850661625,
"grad_norm": 0.057276710867881775,
"learning_rate": 5.955140877753635e-07,
"loss": 0.0003,
"step": 1270
},
{
"epoch": 4.79773156899811,
"grad_norm": 0.033256255090236664,
"learning_rate": 5.919631540562473e-07,
"loss": 0.0003,
"step": 1271
},
{
"epoch": 4.801512287334593,
"grad_norm": 0.11409135162830353,
"learning_rate": 5.884214165086036e-07,
"loss": 0.0006,
"step": 1272
},
{
"epoch": 4.805293005671078,
"grad_norm": 0.19041526317596436,
"learning_rate": 5.848888922025553e-07,
"loss": 0.0006,
"step": 1273
},
{
"epoch": 4.809073724007561,
"grad_norm": 0.0136854387819767,
"learning_rate": 5.813655981638217e-07,
"loss": 0.0001,
"step": 1274
},
{
"epoch": 4.812854442344046,
"grad_norm": 0.12821362912654877,
"learning_rate": 5.778515513736338e-07,
"loss": 0.0005,
"step": 1275
},
{
"epoch": 4.816635160680529,
"grad_norm": 0.018250375986099243,
"learning_rate": 5.743467687686563e-07,
"loss": 0.0001,
"step": 1276
},
{
"epoch": 4.8204158790170135,
"grad_norm": 0.3716386556625366,
"learning_rate": 5.708512672409002e-07,
"loss": 0.001,
"step": 1277
},
{
"epoch": 4.824196597353497,
"grad_norm": 0.07315511256456375,
"learning_rate": 5.673650636376457e-07,
"loss": 0.0005,
"step": 1278
},
{
"epoch": 4.8279773156899815,
"grad_norm": 0.031180553138256073,
"learning_rate": 5.638881747613603e-07,
"loss": 0.0002,
"step": 1279
},
{
"epoch": 4.831758034026465,
"grad_norm": 0.03074919618666172,
"learning_rate": 5.604206173696169e-07,
"loss": 0.0002,
"step": 1280
},
{
"epoch": 4.8355387523629485,
"grad_norm": 0.2015416920185089,
"learning_rate": 5.56962408175013e-07,
"loss": 0.0004,
"step": 1281
},
{
"epoch": 4.839319470699433,
"grad_norm": 0.01151141058653593,
"learning_rate": 5.5351356384509e-07,
"loss": 0.0001,
"step": 1282
},
{
"epoch": 4.843100189035916,
"grad_norm": 0.19769836962223053,
"learning_rate": 5.500741010022564e-07,
"loss": 0.0003,
"step": 1283
},
{
"epoch": 4.846880907372401,
"grad_norm": 0.022758327424526215,
"learning_rate": 5.466440362237e-07,
"loss": 0.0002,
"step": 1284
},
{
"epoch": 4.850661625708884,
"grad_norm": 0.10503081232309341,
"learning_rate": 5.432233860413172e-07,
"loss": 0.0006,
"step": 1285
},
{
"epoch": 4.854442344045369,
"grad_norm": 0.13454023003578186,
"learning_rate": 5.398121669416262e-07,
"loss": 0.0005,
"step": 1286
},
{
"epoch": 4.858223062381852,
"grad_norm": 0.16286829113960266,
"learning_rate": 5.364103953656905e-07,
"loss": 0.0011,
"step": 1287
},
{
"epoch": 4.862003780718337,
"grad_norm": 0.6373857855796814,
"learning_rate": 5.33018087709041e-07,
"loss": 0.0015,
"step": 1288
},
{
"epoch": 4.86578449905482,
"grad_norm": 0.005870916415005922,
"learning_rate": 5.296352603215935e-07,
"loss": 0.0001,
"step": 1289
},
{
"epoch": 4.869565217391305,
"grad_norm": 0.24929335713386536,
"learning_rate": 5.262619295075721e-07,
"loss": 0.0011,
"step": 1290
},
{
"epoch": 4.873345935727788,
"grad_norm": 0.07431139051914215,
"learning_rate": 5.2289811152543e-07,
"loss": 0.0006,
"step": 1291
},
{
"epoch": 4.877126654064273,
"grad_norm": 0.07029802352190018,
"learning_rate": 5.195438225877738e-07,
"loss": 0.0004,
"step": 1292
},
{
"epoch": 4.880907372400756,
"grad_norm": 0.009167086333036423,
"learning_rate": 5.161990788612781e-07,
"loss": 0.0001,
"step": 1293
},
{
"epoch": 4.8846880907372405,
"grad_norm": 0.02880323864519596,
"learning_rate": 5.128638964666166e-07,
"loss": 0.0003,
"step": 1294
},
{
"epoch": 4.888468809073724,
"grad_norm": 0.2872433364391327,
"learning_rate": 5.095382914783803e-07,
"loss": 0.0029,
"step": 1295
},
{
"epoch": 4.892249527410208,
"grad_norm": 0.020014256238937378,
"learning_rate": 5.062222799249958e-07,
"loss": 0.0001,
"step": 1296
},
{
"epoch": 4.896030245746692,
"grad_norm": 0.13867512345314026,
"learning_rate": 5.02915877788657e-07,
"loss": 0.0007,
"step": 1297
},
{
"epoch": 4.8998109640831755,
"grad_norm": 0.23288698494434357,
"learning_rate": 4.996191010052404e-07,
"loss": 0.005,
"step": 1298
},
{
"epoch": 4.90359168241966,
"grad_norm": 0.18965700268745422,
"learning_rate": 4.963319654642321e-07,
"loss": 0.0008,
"step": 1299
},
{
"epoch": 4.9073724007561434,
"grad_norm": 0.015174778178334236,
"learning_rate": 4.93054487008649e-07,
"loss": 0.0001,
"step": 1300
},
{
"epoch": 4.911153119092628,
"grad_norm": 0.006329611875116825,
"learning_rate": 4.897866814349667e-07,
"loss": 0.0001,
"step": 1301
},
{
"epoch": 4.914933837429111,
"grad_norm": 0.054992418736219406,
"learning_rate": 4.865285644930373e-07,
"loss": 0.0003,
"step": 1302
},
{
"epoch": 4.918714555765596,
"grad_norm": 0.008632317185401917,
"learning_rate": 4.832801518860175e-07,
"loss": 0.0001,
"step": 1303
},
{
"epoch": 4.922495274102079,
"grad_norm": 0.08638248592615128,
"learning_rate": 4.800414592702934e-07,
"loss": 0.0004,
"step": 1304
},
{
"epoch": 4.926275992438564,
"grad_norm": 0.0244189091026783,
"learning_rate": 4.7681250225540184e-07,
"loss": 0.0002,
"step": 1305
},
{
"epoch": 4.930056710775047,
"grad_norm": 0.0715269073843956,
"learning_rate": 4.7359329640395746e-07,
"loss": 0.0003,
"step": 1306
},
{
"epoch": 4.933837429111531,
"grad_norm": 0.049625638872385025,
"learning_rate": 4.7038385723157666e-07,
"loss": 0.0002,
"step": 1307
},
{
"epoch": 4.937618147448015,
"grad_norm": 1.2531838417053223,
"learning_rate": 4.671842002068061e-07,
"loss": 0.0013,
"step": 1308
},
{
"epoch": 4.941398865784499,
"grad_norm": 0.28482598066329956,
"learning_rate": 4.639943407510408e-07,
"loss": 0.0022,
"step": 1309
},
{
"epoch": 4.945179584120983,
"grad_norm": 0.24505776166915894,
"learning_rate": 4.608142942384586e-07,
"loss": 0.0012,
"step": 1310
},
{
"epoch": 4.948960302457467,
"grad_norm": 0.007503789849579334,
"learning_rate": 4.576440759959394e-07,
"loss": 0.0001,
"step": 1311
},
{
"epoch": 4.952741020793951,
"grad_norm": 0.27511751651763916,
"learning_rate": 4.54483701302994e-07,
"loss": 0.0017,
"step": 1312
},
{
"epoch": 4.956521739130435,
"grad_norm": 0.05660960078239441,
"learning_rate": 4.513331853916922e-07,
"loss": 0.0003,
"step": 1313
},
{
"epoch": 4.960302457466919,
"grad_norm": 0.022548669949173927,
"learning_rate": 4.4819254344658505e-07,
"loss": 0.0002,
"step": 1314
},
{
"epoch": 4.9640831758034025,
"grad_norm": 0.11190716922283173,
"learning_rate": 4.450617906046348e-07,
"loss": 0.0005,
"step": 1315
},
{
"epoch": 4.967863894139887,
"grad_norm": 0.11963574588298798,
"learning_rate": 4.419409419551407e-07,
"loss": 0.0005,
"step": 1316
},
{
"epoch": 4.9716446124763705,
"grad_norm": 0.054456617683172226,
"learning_rate": 4.388300125396686e-07,
"loss": 0.0002,
"step": 1317
},
{
"epoch": 4.975425330812854,
"grad_norm": 0.17984004318714142,
"learning_rate": 4.3572901735197457e-07,
"loss": 0.0034,
"step": 1318
},
{
"epoch": 4.979206049149338,
"grad_norm": 0.03220417723059654,
"learning_rate": 4.32637971337935e-07,
"loss": 0.0002,
"step": 1319
},
{
"epoch": 4.982986767485822,
"grad_norm": 0.0275163222104311,
"learning_rate": 4.2955688939547516e-07,
"loss": 0.0002,
"step": 1320
},
{
"epoch": 4.986767485822306,
"grad_norm": 0.35353925824165344,
"learning_rate": 4.264857863744956e-07,
"loss": 0.0012,
"step": 1321
},
{
"epoch": 4.99054820415879,
"grad_norm": 0.0351225808262825,
"learning_rate": 4.2342467707680163e-07,
"loss": 0.0002,
"step": 1322
},
{
"epoch": 4.994328922495274,
"grad_norm": 0.7056676745414734,
"learning_rate": 4.2037357625603127e-07,
"loss": 0.0012,
"step": 1323
},
{
"epoch": 4.998109640831758,
"grad_norm": 0.1663798689842224,
"learning_rate": 4.1733249861758697e-07,
"loss": 0.0007,
"step": 1324
},
{
"epoch": 5.0,
"grad_norm": 0.002775799483060837,
"learning_rate": 4.143014588185582e-07,
"loss": 0.0001,
"step": 1325
},
{
"epoch": 5.0037807183364835,
"grad_norm": 0.015545940026640892,
"learning_rate": 4.1128047146765936e-07,
"loss": 0.0001,
"step": 1326
},
{
"epoch": 5.007561436672968,
"grad_norm": 0.020587166771292686,
"learning_rate": 4.082695511251522e-07,
"loss": 0.0002,
"step": 1327
},
{
"epoch": 5.0113421550094515,
"grad_norm": 0.0743769034743309,
"learning_rate": 4.0526871230277904e-07,
"loss": 0.0003,
"step": 1328
},
{
"epoch": 5.015122873345936,
"grad_norm": 0.047476768493652344,
"learning_rate": 4.022779694636941e-07,
"loss": 0.0003,
"step": 1329
},
{
"epoch": 5.018903591682419,
"grad_norm": 0.16292473673820496,
"learning_rate": 3.992973370223896e-07,
"loss": 0.0003,
"step": 1330
},
{
"epoch": 5.022684310018904,
"grad_norm": 0.0061319018714129925,
"learning_rate": 3.963268293446296e-07,
"loss": 0.0001,
"step": 1331
},
{
"epoch": 5.026465028355387,
"grad_norm": 0.1270836442708969,
"learning_rate": 3.933664607473789e-07,
"loss": 0.0007,
"step": 1332
},
{
"epoch": 5.030245746691872,
"grad_norm": 0.21252499520778656,
"learning_rate": 3.904162454987373e-07,
"loss": 0.0015,
"step": 1333
},
{
"epoch": 5.034026465028355,
"grad_norm": 0.03182823956012726,
"learning_rate": 3.874761978178665e-07,
"loss": 0.0002,
"step": 1334
},
{
"epoch": 5.03780718336484,
"grad_norm": 0.08775905519723892,
"learning_rate": 3.845463318749235e-07,
"loss": 0.0004,
"step": 1335
},
{
"epoch": 5.041587901701323,
"grad_norm": 0.05888428911566734,
"learning_rate": 3.8162666179099253e-07,
"loss": 0.0003,
"step": 1336
},
{
"epoch": 5.045368620037808,
"grad_norm": 0.07403896749019623,
"learning_rate": 3.7871720163801836e-07,
"loss": 0.0003,
"step": 1337
},
{
"epoch": 5.049149338374291,
"grad_norm": 0.09960105270147324,
"learning_rate": 3.7581796543873476e-07,
"loss": 0.0004,
"step": 1338
},
{
"epoch": 5.052930056710775,
"grad_norm": 0.16964691877365112,
"learning_rate": 3.729289671665998e-07,
"loss": 0.0003,
"step": 1339
},
{
"epoch": 5.056710775047259,
"grad_norm": 0.037955392152071,
"learning_rate": 3.700502207457293e-07,
"loss": 0.0002,
"step": 1340
},
{
"epoch": 5.060491493383743,
"grad_norm": 0.04902094602584839,
"learning_rate": 3.6718174005082515e-07,
"loss": 0.0002,
"step": 1341
},
{
"epoch": 5.064272211720227,
"grad_norm": 0.12219687551259995,
"learning_rate": 3.6432353890711426e-07,
"loss": 0.0007,
"step": 1342
},
{
"epoch": 5.0680529300567105,
"grad_norm": 0.2989348769187927,
"learning_rate": 3.614756310902781e-07,
"loss": 0.0015,
"step": 1343
},
{
"epoch": 5.071833648393195,
"grad_norm": 0.07826798409223557,
"learning_rate": 3.58638030326387e-07,
"loss": 0.0004,
"step": 1344
},
{
"epoch": 5.0756143667296785,
"grad_norm": 0.3352700173854828,
"learning_rate": 3.5581075029183427e-07,
"loss": 0.0004,
"step": 1345
},
{
"epoch": 5.079395085066163,
"grad_norm": 0.10831774771213531,
"learning_rate": 3.529938046132717e-07,
"loss": 0.0005,
"step": 1346
},
{
"epoch": 5.083175803402646,
"grad_norm": 0.2404409795999527,
"learning_rate": 3.501872068675413e-07,
"loss": 0.0003,
"step": 1347
},
{
"epoch": 5.086956521739131,
"grad_norm": 0.007612737361341715,
"learning_rate": 3.4739097058161116e-07,
"loss": 0.0001,
"step": 1348
},
{
"epoch": 5.090737240075614,
"grad_norm": 0.027310775592923164,
"learning_rate": 3.446051092325112e-07,
"loss": 0.0002,
"step": 1349
},
{
"epoch": 5.094517958412098,
"grad_norm": 0.08922844380140305,
"learning_rate": 3.4182963624726616e-07,
"loss": 0.0006,
"step": 1350
},
{
"epoch": 5.098298676748582,
"grad_norm": 0.07636455446481705,
"learning_rate": 3.390645650028324e-07,
"loss": 0.0004,
"step": 1351
},
{
"epoch": 5.102079395085066,
"grad_norm": 0.19351264834403992,
"learning_rate": 3.3630990882603215e-07,
"loss": 0.0004,
"step": 1352
},
{
"epoch": 5.10586011342155,
"grad_norm": 0.0876229852437973,
"learning_rate": 3.3356568099349287e-07,
"loss": 0.0003,
"step": 1353
},
{
"epoch": 5.109640831758034,
"grad_norm": 0.015671370550990105,
"learning_rate": 3.3083189473157584e-07,
"loss": 0.0001,
"step": 1354
},
{
"epoch": 5.113421550094518,
"grad_norm": 0.05094285309314728,
"learning_rate": 3.281085632163211e-07,
"loss": 0.0002,
"step": 1355
},
{
"epoch": 5.117202268431002,
"grad_norm": 0.004448906984180212,
"learning_rate": 3.253956995733798e-07,
"loss": 0.0001,
"step": 1356
},
{
"epoch": 5.120982986767486,
"grad_norm": 0.06320744007825851,
"learning_rate": 3.2269331687794695e-07,
"loss": 0.0003,
"step": 1357
},
{
"epoch": 5.12476370510397,
"grad_norm": 0.11461395025253296,
"learning_rate": 3.200014281547076e-07,
"loss": 0.0005,
"step": 1358
},
{
"epoch": 5.128544423440454,
"grad_norm": 0.04418833553791046,
"learning_rate": 3.1732004637776594e-07,
"loss": 0.0004,
"step": 1359
},
{
"epoch": 5.1323251417769375,
"grad_norm": 0.04186620935797691,
"learning_rate": 3.1464918447058623e-07,
"loss": 0.0003,
"step": 1360
},
{
"epoch": 5.136105860113422,
"grad_norm": 0.0854787528514862,
"learning_rate": 3.1198885530593025e-07,
"loss": 0.0004,
"step": 1361
},
{
"epoch": 5.1398865784499055,
"grad_norm": 0.048746850341558456,
"learning_rate": 3.0933907170579646e-07,
"loss": 0.0003,
"step": 1362
},
{
"epoch": 5.143667296786389,
"grad_norm": 0.02848944626748562,
"learning_rate": 3.0669984644135455e-07,
"loss": 0.0001,
"step": 1363
},
{
"epoch": 5.147448015122873,
"grad_norm": 0.1088767722249031,
"learning_rate": 3.040711922328871e-07,
"loss": 0.0003,
"step": 1364
},
{
"epoch": 5.151228733459357,
"grad_norm": 0.23314891755580902,
"learning_rate": 3.014531217497282e-07,
"loss": 0.001,
"step": 1365
},
{
"epoch": 5.155009451795841,
"grad_norm": 0.010488437488675117,
"learning_rate": 2.9884564761020083e-07,
"loss": 0.0001,
"step": 1366
},
{
"epoch": 5.158790170132325,
"grad_norm": 0.0435272715985775,
"learning_rate": 2.9624878238155605e-07,
"loss": 0.0003,
"step": 1367
},
{
"epoch": 5.162570888468809,
"grad_norm": 0.16945038735866547,
"learning_rate": 2.9366253857991326e-07,
"loss": 0.001,
"step": 1368
},
{
"epoch": 5.166351606805293,
"grad_norm": 0.22480547428131104,
"learning_rate": 2.910869286702023e-07,
"loss": 0.0017,
"step": 1369
},
{
"epoch": 5.170132325141777,
"grad_norm": 0.016588395461440086,
"learning_rate": 2.8852196506609644e-07,
"loss": 0.0002,
"step": 1370
},
{
"epoch": 5.173913043478261,
"grad_norm": 0.12868349254131317,
"learning_rate": 2.8596766012996047e-07,
"loss": 0.0005,
"step": 1371
},
{
"epoch": 5.177693761814745,
"grad_norm": 0.008798662573099136,
"learning_rate": 2.8342402617278606e-07,
"loss": 0.0001,
"step": 1372
},
{
"epoch": 5.181474480151229,
"grad_norm": 0.013135658577084541,
"learning_rate": 2.808910754541336e-07,
"loss": 0.0001,
"step": 1373
},
{
"epoch": 5.185255198487713,
"grad_norm": 0.08154350519180298,
"learning_rate": 2.78368820182075e-07,
"loss": 0.0003,
"step": 1374
},
{
"epoch": 5.189035916824197,
"grad_norm": 0.2533663511276245,
"learning_rate": 2.7585727251313196e-07,
"loss": 0.0019,
"step": 1375
},
{
"epoch": 5.19281663516068,
"grad_norm": 0.023373350501060486,
"learning_rate": 2.733564445522191e-07,
"loss": 0.0002,
"step": 1376
},
{
"epoch": 5.196597353497165,
"grad_norm": 0.004523281939327717,
"learning_rate": 2.7086634835258416e-07,
"loss": 0.0001,
"step": 1377
},
{
"epoch": 5.200378071833648,
"grad_norm": 0.07378626614809036,
"learning_rate": 2.683869959157534e-07,
"loss": 0.0002,
"step": 1378
},
{
"epoch": 5.2041587901701325,
"grad_norm": 0.15514302253723145,
"learning_rate": 2.6591839919146963e-07,
"loss": 0.0008,
"step": 1379
},
{
"epoch": 5.207939508506616,
"grad_norm": 0.045652490109205246,
"learning_rate": 2.6346057007763603e-07,
"loss": 0.0001,
"step": 1380
},
{
"epoch": 5.2117202268431,
"grad_norm": 0.2619495391845703,
"learning_rate": 2.610135204202599e-07,
"loss": 0.0002,
"step": 1381
},
{
"epoch": 5.215500945179584,
"grad_norm": 0.1071959137916565,
"learning_rate": 2.585772620133947e-07,
"loss": 0.0008,
"step": 1382
},
{
"epoch": 5.219281663516068,
"grad_norm": 0.11351490020751953,
"learning_rate": 2.561518065990834e-07,
"loss": 0.0006,
"step": 1383
},
{
"epoch": 5.223062381852552,
"grad_norm": 0.09148130565881729,
"learning_rate": 2.5373716586730047e-07,
"loss": 0.0004,
"step": 1384
},
{
"epoch": 5.226843100189036,
"grad_norm": 0.11707410216331482,
"learning_rate": 2.513333514558994e-07,
"loss": 0.0013,
"step": 1385
},
{
"epoch": 5.23062381852552,
"grad_norm": 0.006176994182169437,
"learning_rate": 2.4894037495055036e-07,
"loss": 0.0001,
"step": 1386
},
{
"epoch": 5.234404536862003,
"grad_norm": 0.020725781098008156,
"learning_rate": 2.465582478846917e-07,
"loss": 0.0001,
"step": 1387
},
{
"epoch": 5.238185255198488,
"grad_norm": 0.04577802121639252,
"learning_rate": 2.4418698173946873e-07,
"loss": 0.0003,
"step": 1388
},
{
"epoch": 5.241965973534971,
"grad_norm": 0.22789855301380157,
"learning_rate": 2.418265879436804e-07,
"loss": 0.0011,
"step": 1389
},
{
"epoch": 5.245746691871456,
"grad_norm": 0.22857272624969482,
"learning_rate": 2.39477077873724e-07,
"loss": 0.0021,
"step": 1390
},
{
"epoch": 5.249527410207939,
"grad_norm": 0.025963351130485535,
"learning_rate": 2.3713846285354226e-07,
"loss": 0.0002,
"step": 1391
},
{
"epoch": 5.253308128544424,
"grad_norm": 0.0036598974838852882,
"learning_rate": 2.3481075415456494e-07,
"loss": 0.0001,
"step": 1392
},
{
"epoch": 5.257088846880907,
"grad_norm": 0.1250731348991394,
"learning_rate": 2.3249396299565685e-07,
"loss": 0.0006,
"step": 1393
},
{
"epoch": 5.260869565217392,
"grad_norm": 0.05748940631747246,
"learning_rate": 2.3018810054306505e-07,
"loss": 0.0004,
"step": 1394
},
{
"epoch": 5.264650283553875,
"grad_norm": 0.10143335908651352,
"learning_rate": 2.2789317791036186e-07,
"loss": 0.0009,
"step": 1395
},
{
"epoch": 5.2684310018903595,
"grad_norm": 0.07542428374290466,
"learning_rate": 2.2560920615839339e-07,
"loss": 0.0003,
"step": 1396
},
{
"epoch": 5.272211720226843,
"grad_norm": 1.3619216680526733,
"learning_rate": 2.2333619629522519e-07,
"loss": 0.0019,
"step": 1397
},
{
"epoch": 5.2759924385633274,
"grad_norm": 0.06844630837440491,
"learning_rate": 2.2107415927609177e-07,
"loss": 0.0003,
"step": 1398
},
{
"epoch": 5.279773156899811,
"grad_norm": 0.0675884336233139,
"learning_rate": 2.1882310600333968e-07,
"loss": 0.0005,
"step": 1399
},
{
"epoch": 5.2835538752362945,
"grad_norm": 0.13881497085094452,
"learning_rate": 2.1658304732637719e-07,
"loss": 0.0005,
"step": 1400
},
{
"epoch": 5.287334593572779,
"grad_norm": 0.0871807262301445,
"learning_rate": 2.1435399404162393e-07,
"loss": 0.0006,
"step": 1401
},
{
"epoch": 5.291115311909262,
"grad_norm": 0.12734076380729675,
"learning_rate": 2.1213595689245386e-07,
"loss": 0.0004,
"step": 1402
},
{
"epoch": 5.294896030245747,
"grad_norm": 0.15326640009880066,
"learning_rate": 2.0992894656914898e-07,
"loss": 0.0008,
"step": 1403
},
{
"epoch": 5.29867674858223,
"grad_norm": 0.007928919047117233,
"learning_rate": 2.0773297370884376e-07,
"loss": 0.0001,
"step": 1404
},
{
"epoch": 5.302457466918715,
"grad_norm": 0.03422323986887932,
"learning_rate": 2.0554804889547587e-07,
"loss": 0.0003,
"step": 1405
},
{
"epoch": 5.306238185255198,
"grad_norm": 0.05949372425675392,
"learning_rate": 2.0337418265973435e-07,
"loss": 0.0002,
"step": 1406
},
{
"epoch": 5.310018903591683,
"grad_norm": 0.005510756745934486,
"learning_rate": 2.0121138547901014e-07,
"loss": 0.0001,
"step": 1407
},
{
"epoch": 5.313799621928166,
"grad_norm": 0.152805358171463,
"learning_rate": 1.990596677773435e-07,
"loss": 0.0004,
"step": 1408
},
{
"epoch": 5.317580340264651,
"grad_norm": 0.09017285704612732,
"learning_rate": 1.9691903992537508e-07,
"loss": 0.0004,
"step": 1409
},
{
"epoch": 5.321361058601134,
"grad_norm": 0.04021616280078888,
"learning_rate": 1.947895122402968e-07,
"loss": 0.0002,
"step": 1410
},
{
"epoch": 5.325141776937619,
"grad_norm": 0.0911293551325798,
"learning_rate": 1.9267109498579962e-07,
"loss": 0.0003,
"step": 1411
},
{
"epoch": 5.328922495274102,
"grad_norm": 0.06532394886016846,
"learning_rate": 1.9056379837202626e-07,
"loss": 0.0004,
"step": 1412
},
{
"epoch": 5.332703213610586,
"grad_norm": 0.05046873539686203,
"learning_rate": 1.8846763255552098e-07,
"loss": 0.0003,
"step": 1413
},
{
"epoch": 5.33648393194707,
"grad_norm": 0.05093162879347801,
"learning_rate": 1.8638260763918198e-07,
"loss": 0.0002,
"step": 1414
},
{
"epoch": 5.340264650283554,
"grad_norm": 0.12098737806081772,
"learning_rate": 1.8430873367220908e-07,
"loss": 0.0007,
"step": 1415
},
{
"epoch": 5.344045368620038,
"grad_norm": 0.2110583633184433,
"learning_rate": 1.8224602065006035e-07,
"loss": 0.0034,
"step": 1416
},
{
"epoch": 5.3478260869565215,
"grad_norm": 0.2050369828939438,
"learning_rate": 1.8019447851440165e-07,
"loss": 0.001,
"step": 1417
},
{
"epoch": 5.351606805293006,
"grad_norm": 0.16130445897579193,
"learning_rate": 1.781541171530554e-07,
"loss": 0.0038,
"step": 1418
},
{
"epoch": 5.355387523629489,
"grad_norm": 0.256651908159256,
"learning_rate": 1.7612494639995926e-07,
"loss": 0.0011,
"step": 1419
},
{
"epoch": 5.359168241965974,
"grad_norm": 0.047191645950078964,
"learning_rate": 1.7410697603511383e-07,
"loss": 0.0002,
"step": 1420
},
{
"epoch": 5.362948960302457,
"grad_norm": 0.05085434764623642,
"learning_rate": 1.721002157845375e-07,
"loss": 0.0003,
"step": 1421
},
{
"epoch": 5.366729678638942,
"grad_norm": 0.05291781574487686,
"learning_rate": 1.7010467532021902e-07,
"loss": 0.0003,
"step": 1422
},
{
"epoch": 5.370510396975425,
"grad_norm": 0.05128391087055206,
"learning_rate": 1.6812036426007177e-07,
"loss": 0.0004,
"step": 1423
},
{
"epoch": 5.374291115311909,
"grad_norm": 0.07259636372327805,
"learning_rate": 1.661472921678864e-07,
"loss": 0.0004,
"step": 1424
},
{
"epoch": 5.378071833648393,
"grad_norm": 0.08863488584756851,
"learning_rate": 1.64185468553284e-07,
"loss": 0.0003,
"step": 1425
},
{
"epoch": 5.381852551984877,
"grad_norm": 0.13265292346477509,
"learning_rate": 1.6223490287167354e-07,
"loss": 0.0007,
"step": 1426
},
{
"epoch": 5.385633270321361,
"grad_norm": 0.013009740971028805,
"learning_rate": 1.602956045242024e-07,
"loss": 0.0001,
"step": 1427
},
{
"epoch": 5.389413988657845,
"grad_norm": 0.1399068683385849,
"learning_rate": 1.5836758285771303e-07,
"loss": 0.0007,
"step": 1428
},
{
"epoch": 5.393194706994329,
"grad_norm": 0.07947681099176407,
"learning_rate": 1.5645084716469778e-07,
"loss": 0.0003,
"step": 1429
},
{
"epoch": 5.396975425330813,
"grad_norm": 0.07316634058952332,
"learning_rate": 1.5454540668325442e-07,
"loss": 0.0002,
"step": 1430
},
{
"epoch": 5.400756143667297,
"grad_norm": 0.01046953909099102,
"learning_rate": 1.526512705970401e-07,
"loss": 0.0001,
"step": 1431
},
{
"epoch": 5.404536862003781,
"grad_norm": 0.033824142068624496,
"learning_rate": 1.507684480352292e-07,
"loss": 0.0003,
"step": 1432
},
{
"epoch": 5.408317580340265,
"grad_norm": 0.02064157836139202,
"learning_rate": 1.488969480724678e-07,
"loss": 0.0002,
"step": 1433
},
{
"epoch": 5.4120982986767485,
"grad_norm": 0.2741069197654724,
"learning_rate": 1.4703677972882996e-07,
"loss": 0.0004,
"step": 1434
},
{
"epoch": 5.415879017013232,
"grad_norm": 0.12840867042541504,
"learning_rate": 1.4518795196977575e-07,
"loss": 0.0008,
"step": 1435
},
{
"epoch": 5.4196597353497165,
"grad_norm": 0.010716261342167854,
"learning_rate": 1.4335047370610616e-07,
"loss": 0.0001,
"step": 1436
},
{
"epoch": 5.4234404536862,
"grad_norm": 0.15260237455368042,
"learning_rate": 1.415243537939215e-07,
"loss": 0.0008,
"step": 1437
},
{
"epoch": 5.427221172022684,
"grad_norm": 0.15434612333774567,
"learning_rate": 1.3970960103457722e-07,
"loss": 0.0008,
"step": 1438
},
{
"epoch": 5.431001890359168,
"grad_norm": 0.040472716093063354,
"learning_rate": 1.37906224174644e-07,
"loss": 0.0002,
"step": 1439
},
{
"epoch": 5.434782608695652,
"grad_norm": 0.022796135395765305,
"learning_rate": 1.3611423190586308e-07,
"loss": 0.0001,
"step": 1440
},
{
"epoch": 5.438563327032136,
"grad_norm": 0.580222487449646,
"learning_rate": 1.3433363286510514e-07,
"loss": 0.0029,
"step": 1441
},
{
"epoch": 5.44234404536862,
"grad_norm": 0.0772964134812355,
"learning_rate": 1.325644356343289e-07,
"loss": 0.0004,
"step": 1442
},
{
"epoch": 5.446124763705104,
"grad_norm": 0.1101996898651123,
"learning_rate": 1.3080664874054127e-07,
"loss": 0.0007,
"step": 1443
},
{
"epoch": 5.449905482041588,
"grad_norm": 0.02217869460582733,
"learning_rate": 1.2906028065575255e-07,
"loss": 0.0002,
"step": 1444
},
{
"epoch": 5.453686200378072,
"grad_norm": 0.05168196186423302,
"learning_rate": 1.2732533979693902e-07,
"loss": 0.0002,
"step": 1445
},
{
"epoch": 5.457466918714556,
"grad_norm": 0.07490702718496323,
"learning_rate": 1.256018345260013e-07,
"loss": 0.0003,
"step": 1446
},
{
"epoch": 5.46124763705104,
"grad_norm": 0.10955027490854263,
"learning_rate": 1.238897731497224e-07,
"loss": 0.0009,
"step": 1447
},
{
"epoch": 5.465028355387524,
"grad_norm": 0.1142568439245224,
"learning_rate": 1.2218916391973118e-07,
"loss": 0.0004,
"step": 1448
},
{
"epoch": 5.468809073724008,
"grad_norm": 0.05572046339511871,
"learning_rate": 1.2050001503245894e-07,
"loss": 0.0003,
"step": 1449
},
{
"epoch": 5.472589792060491,
"grad_norm": 0.007800830993801355,
"learning_rate": 1.1882233462910259e-07,
"loss": 0.0001,
"step": 1450
},
{
"epoch": 5.4763705103969755,
"grad_norm": 0.08063488453626633,
"learning_rate": 1.1715613079558319e-07,
"loss": 0.0004,
"step": 1451
},
{
"epoch": 5.480151228733459,
"grad_norm": 0.11554457992315292,
"learning_rate": 1.1550141156250944e-07,
"loss": 0.0005,
"step": 1452
},
{
"epoch": 5.4839319470699435,
"grad_norm": 0.08450254052877426,
"learning_rate": 1.1385818490513733e-07,
"loss": 0.0003,
"step": 1453
},
{
"epoch": 5.487712665406427,
"grad_norm": 0.08703256398439407,
"learning_rate": 1.1222645874333082e-07,
"loss": 0.0006,
"step": 1454
},
{
"epoch": 5.491493383742911,
"grad_norm": 0.10387970507144928,
"learning_rate": 1.1060624094152678e-07,
"loss": 0.0005,
"step": 1455
},
{
"epoch": 5.495274102079395,
"grad_norm": 0.09885484725236893,
"learning_rate": 1.0899753930869395e-07,
"loss": 0.0005,
"step": 1456
},
{
"epoch": 5.499054820415879,
"grad_norm": 0.03747022524476051,
"learning_rate": 1.0740036159829714e-07,
"loss": 0.0003,
"step": 1457
},
{
"epoch": 5.502835538752363,
"grad_norm": 0.13394807279109955,
"learning_rate": 1.0581471550825812e-07,
"loss": 0.0004,
"step": 1458
},
{
"epoch": 5.506616257088847,
"grad_norm": 0.06739701330661774,
"learning_rate": 1.0424060868092195e-07,
"loss": 0.0005,
"step": 1459
},
{
"epoch": 5.510396975425331,
"grad_norm": 0.21297559142112732,
"learning_rate": 1.0267804870301546e-07,
"loss": 0.0008,
"step": 1460
},
{
"epoch": 5.514177693761814,
"grad_norm": 0.0547952838242054,
"learning_rate": 1.011270431056141e-07,
"loss": 0.0002,
"step": 1461
},
{
"epoch": 5.517958412098299,
"grad_norm": 0.05023105815052986,
"learning_rate": 9.958759936410572e-08,
"loss": 0.0003,
"step": 1462
},
{
"epoch": 5.521739130434782,
"grad_norm": 0.026179540902376175,
"learning_rate": 9.805972489815103e-08,
"loss": 0.0002,
"step": 1463
},
{
"epoch": 5.525519848771267,
"grad_norm": 0.07386493682861328,
"learning_rate": 9.654342707165254e-08,
"loss": 0.0003,
"step": 1464
},
{
"epoch": 5.52930056710775,
"grad_norm": 0.037049535661935806,
"learning_rate": 9.503871319271552e-08,
"loss": 0.0003,
"step": 1465
},
{
"epoch": 5.533081285444235,
"grad_norm": 0.010233812965452671,
"learning_rate": 9.354559051361457e-08,
"loss": 0.0001,
"step": 1466
},
{
"epoch": 5.536862003780718,
"grad_norm": 0.018706122413277626,
"learning_rate": 9.20640662307576e-08,
"loss": 0.0001,
"step": 1467
},
{
"epoch": 5.5406427221172025,
"grad_norm": 0.09481865912675858,
"learning_rate": 9.059414748465278e-08,
"loss": 0.0004,
"step": 1468
},
{
"epoch": 5.544423440453686,
"grad_norm": 0.08331181108951569,
"learning_rate": 8.913584135987197e-08,
"loss": 0.0006,
"step": 1469
},
{
"epoch": 5.5482041587901705,
"grad_norm": 0.1462664157152176,
"learning_rate": 8.768915488501728e-08,
"loss": 0.0009,
"step": 1470
},
{
"epoch": 5.551984877126654,
"grad_norm": 0.02431604079902172,
"learning_rate": 8.625409503268956e-08,
"loss": 0.0002,
"step": 1471
},
{
"epoch": 5.5557655954631375,
"grad_norm": 0.12077602744102478,
"learning_rate": 8.483066871945083e-08,
"loss": 0.0008,
"step": 1472
},
{
"epoch": 5.559546313799622,
"grad_norm": 0.13491597771644592,
"learning_rate": 8.341888280579386e-08,
"loss": 0.0006,
"step": 1473
},
{
"epoch": 5.563327032136106,
"grad_norm": 0.056098464876413345,
"learning_rate": 8.201874409610732e-08,
"loss": 0.0003,
"step": 1474
},
{
"epoch": 5.56710775047259,
"grad_norm": 0.006893400568515062,
"learning_rate": 8.063025933864515e-08,
"loss": 0.0001,
"step": 1475
},
{
"epoch": 5.570888468809073,
"grad_norm": 0.09452251344919205,
"learning_rate": 7.92534352254909e-08,
"loss": 0.0003,
"step": 1476
},
{
"epoch": 5.574669187145558,
"grad_norm": 0.08272690325975418,
"learning_rate": 7.788827839252888e-08,
"loss": 0.0005,
"step": 1477
},
{
"epoch": 5.578449905482041,
"grad_norm": 0.1099521815776825,
"learning_rate": 7.653479541941039e-08,
"loss": 0.0003,
"step": 1478
},
{
"epoch": 5.582230623818526,
"grad_norm": 0.049844082444906235,
"learning_rate": 7.519299282952086e-08,
"loss": 0.0002,
"step": 1479
},
{
"epoch": 5.586011342155009,
"grad_norm": 0.007461511995643377,
"learning_rate": 7.38628770899516e-08,
"loss": 0.0001,
"step": 1480
},
{
"epoch": 5.589792060491494,
"grad_norm": 0.29780179262161255,
"learning_rate": 7.254445461146509e-08,
"loss": 0.0012,
"step": 1481
},
{
"epoch": 5.593572778827977,
"grad_norm": 0.06711945682764053,
"learning_rate": 7.123773174846694e-08,
"loss": 0.0004,
"step": 1482
},
{
"epoch": 5.597353497164462,
"grad_norm": 0.0450141467154026,
"learning_rate": 6.994271479897313e-08,
"loss": 0.0003,
"step": 1483
},
{
"epoch": 5.601134215500945,
"grad_norm": 0.06763432174921036,
"learning_rate": 6.865941000458148e-08,
"loss": 0.0003,
"step": 1484
},
{
"epoch": 5.60491493383743,
"grad_norm": 0.09284917265176773,
"learning_rate": 6.738782355044048e-08,
"loss": 0.0006,
"step": 1485
},
{
"epoch": 5.608695652173913,
"grad_norm": 0.14375682175159454,
"learning_rate": 6.612796156521912e-08,
"loss": 0.0015,
"step": 1486
},
{
"epoch": 5.612476370510397,
"grad_norm": 0.15720754861831665,
"learning_rate": 6.487983012107873e-08,
"loss": 0.0003,
"step": 1487
},
{
"epoch": 5.616257088846881,
"grad_norm": 0.03545217588543892,
"learning_rate": 6.364343523364264e-08,
"loss": 0.0002,
"step": 1488
},
{
"epoch": 5.6200378071833645,
"grad_norm": 0.04067006707191467,
"learning_rate": 6.241878286196684e-08,
"loss": 0.0003,
"step": 1489
},
{
"epoch": 5.623818525519849,
"grad_norm": 0.030347026884555817,
"learning_rate": 6.120587890851293e-08,
"loss": 0.0002,
"step": 1490
},
{
"epoch": 5.6275992438563325,
"grad_norm": 0.28576311469078064,
"learning_rate": 6.000472921911776e-08,
"loss": 0.0005,
"step": 1491
},
{
"epoch": 5.631379962192817,
"grad_norm": 0.15578241646289825,
"learning_rate": 5.8815339582966316e-08,
"loss": 0.0009,
"step": 1492
},
{
"epoch": 5.6351606805293,
"grad_norm": 0.026040202006697655,
"learning_rate": 5.763771573256416e-08,
"loss": 0.0002,
"step": 1493
},
{
"epoch": 5.638941398865785,
"grad_norm": 0.02388918586075306,
"learning_rate": 5.6471863343708636e-08,
"loss": 0.0001,
"step": 1494
},
{
"epoch": 5.642722117202268,
"grad_norm": 0.01957562379539013,
"learning_rate": 5.531778803546217e-08,
"loss": 0.0001,
"step": 1495
},
{
"epoch": 5.646502835538753,
"grad_norm": 0.26871076226234436,
"learning_rate": 5.4175495370125674e-08,
"loss": 0.0005,
"step": 1496
},
{
"epoch": 5.650283553875236,
"grad_norm": 0.07626042515039444,
"learning_rate": 5.304499085321074e-08,
"loss": 0.0003,
"step": 1497
},
{
"epoch": 5.65406427221172,
"grad_norm": 0.0436607301235199,
"learning_rate": 5.192627993341359e-08,
"loss": 0.0002,
"step": 1498
},
{
"epoch": 5.657844990548204,
"grad_norm": 0.09298070520162582,
"learning_rate": 5.081936800258897e-08,
"loss": 0.0004,
"step": 1499
},
{
"epoch": 5.661625708884688,
"grad_norm": 0.13054236769676208,
"learning_rate": 4.972426039572431e-08,
"loss": 0.0004,
"step": 1500
},
{
"epoch": 5.665406427221172,
"grad_norm": 0.276819109916687,
"learning_rate": 4.864096239091287e-08,
"loss": 0.0015,
"step": 1501
},
{
"epoch": 5.669187145557656,
"grad_norm": 0.06801024079322815,
"learning_rate": 4.7569479209329804e-08,
"loss": 0.0003,
"step": 1502
},
{
"epoch": 5.67296786389414,
"grad_norm": 0.15092165768146515,
"learning_rate": 4.65098160152061e-08,
"loss": 0.0009,
"step": 1503
},
{
"epoch": 5.676748582230624,
"grad_norm": 0.3054540753364563,
"learning_rate": 4.5461977915803866e-08,
"loss": 0.0016,
"step": 1504
},
{
"epoch": 5.680529300567108,
"grad_norm": 0.05866989865899086,
"learning_rate": 4.442596996139192e-08,
"loss": 0.0001,
"step": 1505
},
{
"epoch": 5.6843100189035916,
"grad_norm": 0.10321120917797089,
"learning_rate": 4.3401797145221356e-08,
"loss": 0.0004,
"step": 1506
},
{
"epoch": 5.688090737240076,
"grad_norm": 0.03960734233260155,
"learning_rate": 4.238946440350111e-08,
"loss": 0.0002,
"step": 1507
},
{
"epoch": 5.6918714555765595,
"grad_norm": 0.04474017396569252,
"learning_rate": 4.1388976615374664e-08,
"loss": 0.0002,
"step": 1508
},
{
"epoch": 5.695652173913043,
"grad_norm": 0.19345064461231232,
"learning_rate": 4.040033860289644e-08,
"loss": 0.0011,
"step": 1509
},
{
"epoch": 5.699432892249527,
"grad_norm": 0.107937291264534,
"learning_rate": 3.942355513100793e-08,
"loss": 0.0005,
"step": 1510
},
{
"epoch": 5.703213610586012,
"grad_norm": 0.07085900008678436,
"learning_rate": 3.8458630907516336e-08,
"loss": 0.0003,
"step": 1511
},
{
"epoch": 5.706994328922495,
"grad_norm": 0.27383333444595337,
"learning_rate": 3.7505570583069304e-08,
"loss": 0.001,
"step": 1512
},
{
"epoch": 5.710775047258979,
"grad_norm": 0.10581612586975098,
"learning_rate": 3.656437875113522e-08,
"loss": 0.0006,
"step": 1513
},
{
"epoch": 5.714555765595463,
"grad_norm": 0.10589946061372757,
"learning_rate": 3.563505994797961e-08,
"loss": 0.0004,
"step": 1514
},
{
"epoch": 5.718336483931947,
"grad_norm": 0.09371243417263031,
"learning_rate": 3.4717618652642947e-08,
"loss": 0.0004,
"step": 1515
},
{
"epoch": 5.722117202268431,
"grad_norm": 0.15393085777759552,
"learning_rate": 3.381205928692094e-08,
"loss": 0.0008,
"step": 1516
},
{
"epoch": 5.725897920604915,
"grad_norm": 0.05155206471681595,
"learning_rate": 3.2918386215339824e-08,
"loss": 0.0002,
"step": 1517
},
{
"epoch": 5.729678638941399,
"grad_norm": 0.10959815979003906,
"learning_rate": 3.203660374513945e-08,
"loss": 0.0003,
"step": 1518
},
{
"epoch": 5.733459357277883,
"grad_norm": 0.11393946409225464,
"learning_rate": 3.1166716126249664e-08,
"loss": 0.0006,
"step": 1519
},
{
"epoch": 5.737240075614367,
"grad_norm": 0.11420772224664688,
"learning_rate": 3.030872755127034e-08,
"loss": 0.0003,
"step": 1520
},
{
"epoch": 5.741020793950851,
"grad_norm": 0.17956776916980743,
"learning_rate": 2.9462642155451972e-08,
"loss": 0.0005,
"step": 1521
},
{
"epoch": 5.744801512287335,
"grad_norm": 0.04493578523397446,
"learning_rate": 2.8628464016675073e-08,
"loss": 0.0002,
"step": 1522
},
{
"epoch": 5.748582230623819,
"grad_norm": 0.07825593650341034,
"learning_rate": 2.7806197155431093e-08,
"loss": 0.0003,
"step": 1523
},
{
"epoch": 5.752362948960302,
"grad_norm": 0.10764266550540924,
"learning_rate": 2.6995845534802113e-08,
"loss": 0.0007,
"step": 1524
},
{
"epoch": 5.7561436672967865,
"grad_norm": 0.017182912677526474,
"learning_rate": 2.6197413060442268e-08,
"loss": 0.0001,
"step": 1525
},
{
"epoch": 5.75992438563327,
"grad_norm": 0.05833111330866814,
"learning_rate": 2.5410903580559698e-08,
"loss": 0.0004,
"step": 1526
},
{
"epoch": 5.763705103969754,
"grad_norm": 0.23259037733078003,
"learning_rate": 2.4636320885896294e-08,
"loss": 0.002,
"step": 1527
},
{
"epoch": 5.767485822306238,
"grad_norm": 0.025183355435729027,
"learning_rate": 2.387366870971103e-08,
"loss": 0.0002,
"step": 1528
},
{
"epoch": 5.771266540642722,
"grad_norm": 0.004049487411975861,
"learning_rate": 2.3122950727761383e-08,
"loss": 0.0001,
"step": 1529
},
{
"epoch": 5.775047258979206,
"grad_norm": 0.21495746076107025,
"learning_rate": 2.2384170558284456e-08,
"loss": 0.0007,
"step": 1530
},
{
"epoch": 5.77882797731569,
"grad_norm": 0.04991794005036354,
"learning_rate": 2.165733176198198e-08,
"loss": 0.0002,
"step": 1531
},
{
"epoch": 5.782608695652174,
"grad_norm": 0.03515658527612686,
"learning_rate": 2.0942437842001727e-08,
"loss": 0.0002,
"step": 1532
},
{
"epoch": 5.786389413988658,
"grad_norm": 0.03204856067895889,
"learning_rate": 2.0239492243919467e-08,
"loss": 0.0002,
"step": 1533
},
{
"epoch": 5.790170132325142,
"grad_norm": 0.03522752225399017,
"learning_rate": 1.9548498355725086e-08,
"loss": 0.0003,
"step": 1534
},
{
"epoch": 5.793950850661625,
"grad_norm": 0.006235829554498196,
"learning_rate": 1.8869459507803722e-08,
"loss": 0.0001,
"step": 1535
},
{
"epoch": 5.79773156899811,
"grad_norm": 0.053368937224149704,
"learning_rate": 1.8202378972921043e-08,
"loss": 0.0003,
"step": 1536
},
{
"epoch": 5.801512287334593,
"grad_norm": 0.15392562747001648,
"learning_rate": 1.7547259966207708e-08,
"loss": 0.0007,
"step": 1537
},
{
"epoch": 5.805293005671078,
"grad_norm": 0.17654727399349213,
"learning_rate": 1.6904105645142443e-08,
"loss": 0.0009,
"step": 1538
},
{
"epoch": 5.809073724007561,
"grad_norm": 0.05364728718996048,
"learning_rate": 1.6272919109538986e-08,
"loss": 0.0003,
"step": 1539
},
{
"epoch": 5.812854442344046,
"grad_norm": 0.08575626462697983,
"learning_rate": 1.565370340152833e-08,
"loss": 0.0012,
"step": 1540
},
{
"epoch": 5.816635160680529,
"grad_norm": 0.05414729565382004,
"learning_rate": 1.504646150554706e-08,
"loss": 0.0002,
"step": 1541
},
{
"epoch": 5.8204158790170135,
"grad_norm": 0.11814671009778976,
"learning_rate": 1.4451196348320706e-08,
"loss": 0.0005,
"step": 1542
},
{
"epoch": 5.824196597353497,
"grad_norm": 0.07417936623096466,
"learning_rate": 1.3867910798850692e-08,
"loss": 0.0002,
"step": 1543
},
{
"epoch": 5.8279773156899815,
"grad_norm": 0.05950590968132019,
"learning_rate": 1.3296607668400186e-08,
"loss": 0.0004,
"step": 1544
},
{
"epoch": 5.831758034026465,
"grad_norm": 0.23362042009830475,
"learning_rate": 1.2737289710481048e-08,
"loss": 0.0032,
"step": 1545
},
{
"epoch": 5.8355387523629485,
"grad_norm": 0.028809770941734314,
"learning_rate": 1.2189959620839686e-08,
"loss": 0.0002,
"step": 1546
},
{
"epoch": 5.839319470699433,
"grad_norm": 0.12129279226064682,
"learning_rate": 1.1654620037445109e-08,
"loss": 0.0006,
"step": 1547
},
{
"epoch": 5.843100189035916,
"grad_norm": 0.0542544461786747,
"learning_rate": 1.1131273540474497e-08,
"loss": 0.0002,
"step": 1548
},
{
"epoch": 5.846880907372401,
"grad_norm": 0.3382869064807892,
"learning_rate": 1.0619922652303494e-08,
"loss": 0.0023,
"step": 1549
},
{
"epoch": 5.850661625708884,
"grad_norm": 0.191293865442276,
"learning_rate": 1.0120569837491212e-08,
"loss": 0.0024,
"step": 1550
},
{
"epoch": 5.854442344045369,
"grad_norm": 0.10547634959220886,
"learning_rate": 9.633217502770242e-09,
"loss": 0.0006,
"step": 1551
},
{
"epoch": 5.858223062381852,
"grad_norm": 0.12045734375715256,
"learning_rate": 9.157867997034165e-09,
"loss": 0.0003,
"step": 1552
},
{
"epoch": 5.862003780718337,
"grad_norm": 0.013879460282623768,
"learning_rate": 8.694523611326445e-09,
"loss": 0.0001,
"step": 1553
},
{
"epoch": 5.86578449905482,
"grad_norm": 0.20656751096248627,
"learning_rate": 8.243186578829609e-09,
"loss": 0.0009,
"step": 1554
},
{
"epoch": 5.869565217391305,
"grad_norm": 0.013404203578829765,
"learning_rate": 7.803859074854425e-09,
"loss": 0.0001,
"step": 1555
},
{
"epoch": 5.873345935727788,
"grad_norm": 0.0296834297478199,
"learning_rate": 7.376543216828791e-09,
"loss": 0.0002,
"step": 1556
},
{
"epoch": 5.877126654064273,
"grad_norm": 0.01866457611322403,
"learning_rate": 6.9612410642888595e-09,
"loss": 0.0002,
"step": 1557
},
{
"epoch": 5.880907372400756,
"grad_norm": 0.1199808269739151,
"learning_rate": 6.557954618867102e-09,
"loss": 0.0006,
"step": 1558
},
{
"epoch": 5.8846880907372405,
"grad_norm": 0.05731095373630524,
"learning_rate": 6.16668582428509e-09,
"loss": 0.0003,
"step": 1559
},
{
"epoch": 5.888468809073724,
"grad_norm": 0.10954930633306503,
"learning_rate": 5.7874365663421195e-09,
"loss": 0.0004,
"step": 1560
},
{
"epoch": 5.892249527410208,
"grad_norm": 0.04187054932117462,
"learning_rate": 5.420208672907157e-09,
"loss": 0.0002,
"step": 1561
},
{
"epoch": 5.896030245746692,
"grad_norm": 0.031233927235007286,
"learning_rate": 5.065003913909406e-09,
"loss": 0.0003,
"step": 1562
},
{
"epoch": 5.8998109640831755,
"grad_norm": 0.16604606807231903,
"learning_rate": 4.7218240013308105e-09,
"loss": 0.0004,
"step": 1563
},
{
"epoch": 5.90359168241966,
"grad_norm": 0.23016543686389923,
"learning_rate": 4.390670589196622e-09,
"loss": 0.0009,
"step": 1564
},
{
"epoch": 5.9073724007561434,
"grad_norm": 0.059649109840393066,
"learning_rate": 4.071545273567623e-09,
"loss": 0.0003,
"step": 1565
},
{
"epoch": 5.911153119092628,
"grad_norm": 0.16850413382053375,
"learning_rate": 3.764449592533748e-09,
"loss": 0.0003,
"step": 1566
},
{
"epoch": 5.914933837429111,
"grad_norm": 0.1676778495311737,
"learning_rate": 3.4693850262046415e-09,
"loss": 0.002,
"step": 1567
},
{
"epoch": 5.918714555765596,
"grad_norm": 0.025193141773343086,
"learning_rate": 3.1863529967041117e-09,
"loss": 0.0002,
"step": 1568
},
{
"epoch": 5.922495274102079,
"grad_norm": 0.023572875186800957,
"learning_rate": 2.9153548681626344e-09,
"loss": 0.0001,
"step": 1569
},
{
"epoch": 5.926275992438564,
"grad_norm": 0.2784389853477478,
"learning_rate": 2.65639194671069e-09,
"loss": 0.0011,
"step": 1570
},
{
"epoch": 5.930056710775047,
"grad_norm": 0.09678535163402557,
"learning_rate": 2.409465480472939e-09,
"loss": 0.0002,
"step": 1571
},
{
"epoch": 5.933837429111531,
"grad_norm": 0.9893901348114014,
"learning_rate": 2.1745766595612785e-09,
"loss": 0.001,
"step": 1572
},
{
"epoch": 5.937618147448015,
"grad_norm": 0.05149627849459648,
"learning_rate": 1.951726616070404e-09,
"loss": 0.0002,
"step": 1573
},
{
"epoch": 5.941398865784499,
"grad_norm": 0.12159543484449387,
"learning_rate": 1.740916424071426e-09,
"loss": 0.0004,
"step": 1574
},
{
"epoch": 5.945179584120983,
"grad_norm": 0.006072481628507376,
"learning_rate": 1.5421470996065946e-09,
"loss": 0.0001,
"step": 1575
},
{
"epoch": 5.948960302457467,
"grad_norm": 0.16599299013614655,
"learning_rate": 1.3554196006854148e-09,
"loss": 0.0008,
"step": 1576
},
{
"epoch": 5.952741020793951,
"grad_norm": 0.1662135273218155,
"learning_rate": 1.1807348272790952e-09,
"loss": 0.001,
"step": 1577
},
{
"epoch": 5.956521739130435,
"grad_norm": 0.1766807585954666,
"learning_rate": 1.018093621316385e-09,
"loss": 0.001,
"step": 1578
},
{
"epoch": 5.960302457466919,
"grad_norm": 0.17354196310043335,
"learning_rate": 8.674967666791323e-10,
"loss": 0.0011,
"step": 1579
},
{
"epoch": 5.9640831758034025,
"grad_norm": 0.26174816489219666,
"learning_rate": 7.289449892000644e-10,
"loss": 0.0009,
"step": 1580
},
{
"epoch": 5.967863894139887,
"grad_norm": 0.0724356546998024,
"learning_rate": 6.02438956657514e-10,
"loss": 0.0003,
"step": 1581
},
{
"epoch": 5.9716446124763705,
"grad_norm": 0.03557472303509712,
"learning_rate": 4.87979278772921e-10,
"loss": 0.0002,
"step": 1582
},
{
"epoch": 5.975425330812854,
"grad_norm": 0.009614085778594017,
"learning_rate": 3.855665072080572e-10,
"loss": 0.0001,
"step": 1583
},
{
"epoch": 5.979206049149338,
"grad_norm": 0.02289929799735546,
"learning_rate": 2.952011355619733e-10,
"loss": 0.0002,
"step": 1584
},
{
"epoch": 5.982986767485822,
"grad_norm": 0.08921874314546585,
"learning_rate": 2.1688359936905547e-10,
"loss": 0.0005,
"step": 1585
},
{
"epoch": 5.986767485822306,
"grad_norm": 0.12357887625694275,
"learning_rate": 1.506142760965279e-10,
"loss": 0.0003,
"step": 1586
},
{
"epoch": 5.99054820415879,
"grad_norm": 0.10850539058446884,
"learning_rate": 9.639348514278724e-11,
"loss": 0.0007,
"step": 1587
},
{
"epoch": 5.994328922495274,
"grad_norm": 0.047880396246910095,
"learning_rate": 5.422148783629233e-11,
"loss": 0.0002,
"step": 1588
},
{
"epoch": 5.998109640831758,
"grad_norm": 0.11887647956609726,
"learning_rate": 2.4098487433066308e-11,
"loss": 0.0041,
"step": 1589
},
{
"epoch": 6.0,
"grad_norm": 0.02721889689564705,
"learning_rate": 6.024629117529212e-12,
"loss": 0.0002,
"step": 1590
}
],
"logging_steps": 1,
"max_steps": 1590,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.792993297129308e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}