teacher_math_qwq / trainer_state.json
sedrickkeh's picture
End of training
7e02779 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.99492385786802,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0040609137055837565,
"grad_norm": 8.62887018244275,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.3281,
"step": 1
},
{
"epoch": 0.008121827411167513,
"grad_norm": 8.528693763624386,
"learning_rate": 6.504065040650407e-07,
"loss": 1.3167,
"step": 2
},
{
"epoch": 0.012182741116751269,
"grad_norm": 8.64295988544749,
"learning_rate": 9.75609756097561e-07,
"loss": 1.3497,
"step": 3
},
{
"epoch": 0.016243654822335026,
"grad_norm": 8.513854480399086,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.3226,
"step": 4
},
{
"epoch": 0.02030456852791878,
"grad_norm": 8.034222432123684,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.285,
"step": 5
},
{
"epoch": 0.024365482233502538,
"grad_norm": 7.956353805331463,
"learning_rate": 1.951219512195122e-06,
"loss": 1.313,
"step": 6
},
{
"epoch": 0.028426395939086295,
"grad_norm": 6.639649080704839,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.277,
"step": 7
},
{
"epoch": 0.03248730964467005,
"grad_norm": 6.145878036438217,
"learning_rate": 2.601626016260163e-06,
"loss": 1.2577,
"step": 8
},
{
"epoch": 0.03654822335025381,
"grad_norm": 3.7121117474970133,
"learning_rate": 2.926829268292683e-06,
"loss": 1.242,
"step": 9
},
{
"epoch": 0.04060913705583756,
"grad_norm": 3.0026204471431077,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.2168,
"step": 10
},
{
"epoch": 0.04467005076142132,
"grad_norm": 2.769041332882583,
"learning_rate": 3.577235772357724e-06,
"loss": 1.2155,
"step": 11
},
{
"epoch": 0.048730964467005075,
"grad_norm": 6.377649257998999,
"learning_rate": 3.902439024390244e-06,
"loss": 1.206,
"step": 12
},
{
"epoch": 0.05279187817258883,
"grad_norm": 7.028820624654692,
"learning_rate": 4.227642276422765e-06,
"loss": 1.224,
"step": 13
},
{
"epoch": 0.05685279187817259,
"grad_norm": 7.169815209444628,
"learning_rate": 4.552845528455285e-06,
"loss": 1.2076,
"step": 14
},
{
"epoch": 0.06091370558375635,
"grad_norm": 6.903469103811548,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.1946,
"step": 15
},
{
"epoch": 0.0649746192893401,
"grad_norm": 5.51194757755764,
"learning_rate": 5.203252032520326e-06,
"loss": 1.1696,
"step": 16
},
{
"epoch": 0.06903553299492386,
"grad_norm": 4.905523952903594,
"learning_rate": 5.528455284552846e-06,
"loss": 1.1389,
"step": 17
},
{
"epoch": 0.07309644670050762,
"grad_norm": 3.771508935085431,
"learning_rate": 5.853658536585366e-06,
"loss": 1.0961,
"step": 18
},
{
"epoch": 0.07715736040609138,
"grad_norm": 2.5744042874872193,
"learning_rate": 6.178861788617887e-06,
"loss": 1.0917,
"step": 19
},
{
"epoch": 0.08121827411167512,
"grad_norm": 1.8431005762273183,
"learning_rate": 6.504065040650407e-06,
"loss": 1.1078,
"step": 20
},
{
"epoch": 0.08527918781725888,
"grad_norm": 1.5364887515121395,
"learning_rate": 6.829268292682928e-06,
"loss": 1.0766,
"step": 21
},
{
"epoch": 0.08934010152284264,
"grad_norm": 1.766853743034859,
"learning_rate": 7.154471544715448e-06,
"loss": 1.0758,
"step": 22
},
{
"epoch": 0.09340101522842639,
"grad_norm": 1.6796973091236893,
"learning_rate": 7.4796747967479676e-06,
"loss": 1.0812,
"step": 23
},
{
"epoch": 0.09746192893401015,
"grad_norm": 1.4139633784286207,
"learning_rate": 7.804878048780489e-06,
"loss": 1.0809,
"step": 24
},
{
"epoch": 0.10152284263959391,
"grad_norm": 1.1019008459828188,
"learning_rate": 8.130081300813009e-06,
"loss": 1.08,
"step": 25
},
{
"epoch": 0.10558375634517767,
"grad_norm": 0.8894528191997672,
"learning_rate": 8.45528455284553e-06,
"loss": 1.0642,
"step": 26
},
{
"epoch": 0.10964467005076142,
"grad_norm": 0.9719521032491878,
"learning_rate": 8.78048780487805e-06,
"loss": 1.0312,
"step": 27
},
{
"epoch": 0.11370558375634518,
"grad_norm": 1.107583812116328,
"learning_rate": 9.10569105691057e-06,
"loss": 1.021,
"step": 28
},
{
"epoch": 0.11776649746192894,
"grad_norm": 0.9134551268878944,
"learning_rate": 9.43089430894309e-06,
"loss": 1.0505,
"step": 29
},
{
"epoch": 0.1218274111675127,
"grad_norm": 0.7210846664331737,
"learning_rate": 9.756097560975611e-06,
"loss": 0.9874,
"step": 30
},
{
"epoch": 0.12588832487309645,
"grad_norm": 0.7659361480497492,
"learning_rate": 1.008130081300813e-05,
"loss": 1.0069,
"step": 31
},
{
"epoch": 0.1299492385786802,
"grad_norm": 0.7438429636075108,
"learning_rate": 1.0406504065040652e-05,
"loss": 1.0112,
"step": 32
},
{
"epoch": 0.13401015228426397,
"grad_norm": 0.6602902126941446,
"learning_rate": 1.0731707317073172e-05,
"loss": 1.0098,
"step": 33
},
{
"epoch": 0.13807106598984772,
"grad_norm": 0.7053271024764887,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.9973,
"step": 34
},
{
"epoch": 0.14213197969543148,
"grad_norm": 0.7850807386281604,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.9651,
"step": 35
},
{
"epoch": 0.14619289340101524,
"grad_norm": 0.8422666355462011,
"learning_rate": 1.1707317073170731e-05,
"loss": 1.0283,
"step": 36
},
{
"epoch": 0.150253807106599,
"grad_norm": 0.7041794430867934,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.9736,
"step": 37
},
{
"epoch": 0.15431472081218275,
"grad_norm": 0.7150981346028934,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.9735,
"step": 38
},
{
"epoch": 0.15837563451776648,
"grad_norm": 0.5678935577448442,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.9805,
"step": 39
},
{
"epoch": 0.16243654822335024,
"grad_norm": 0.7160267447779376,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.9921,
"step": 40
},
{
"epoch": 0.166497461928934,
"grad_norm": 0.6097502302876489,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.9298,
"step": 41
},
{
"epoch": 0.17055837563451776,
"grad_norm": 0.6849819481605205,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.9676,
"step": 42
},
{
"epoch": 0.1746192893401015,
"grad_norm": 0.6372314391490987,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.9593,
"step": 43
},
{
"epoch": 0.17868020304568527,
"grad_norm": 0.6949260989786994,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.9642,
"step": 44
},
{
"epoch": 0.18274111675126903,
"grad_norm": 0.6660665121083698,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.9533,
"step": 45
},
{
"epoch": 0.18680203045685279,
"grad_norm": 0.4770367257773198,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.9431,
"step": 46
},
{
"epoch": 0.19086294416243654,
"grad_norm": 0.7214063100874251,
"learning_rate": 1.528455284552846e-05,
"loss": 0.9418,
"step": 47
},
{
"epoch": 0.1949238578680203,
"grad_norm": 0.6206595122857601,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.9835,
"step": 48
},
{
"epoch": 0.19898477157360406,
"grad_norm": 0.5819135578361109,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.9219,
"step": 49
},
{
"epoch": 0.20304568527918782,
"grad_norm": 0.6984130437647903,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.9447,
"step": 50
},
{
"epoch": 0.20710659898477157,
"grad_norm": 0.599926623605935,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.9499,
"step": 51
},
{
"epoch": 0.21116751269035533,
"grad_norm": 0.829014022306582,
"learning_rate": 1.691056910569106e-05,
"loss": 0.9425,
"step": 52
},
{
"epoch": 0.2152284263959391,
"grad_norm": 1.343436172185058,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.9445,
"step": 53
},
{
"epoch": 0.21928934010152284,
"grad_norm": 0.9026823421939604,
"learning_rate": 1.75609756097561e-05,
"loss": 0.9366,
"step": 54
},
{
"epoch": 0.2233502538071066,
"grad_norm": 1.0967883445136937,
"learning_rate": 1.788617886178862e-05,
"loss": 0.9341,
"step": 55
},
{
"epoch": 0.22741116751269036,
"grad_norm": 1.1948480173664822,
"learning_rate": 1.821138211382114e-05,
"loss": 0.9294,
"step": 56
},
{
"epoch": 0.23147208121827412,
"grad_norm": 0.8493475703298556,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.9487,
"step": 57
},
{
"epoch": 0.23553299492385787,
"grad_norm": 1.3492368231774985,
"learning_rate": 1.886178861788618e-05,
"loss": 0.9282,
"step": 58
},
{
"epoch": 0.23959390862944163,
"grad_norm": 1.3087960937509129,
"learning_rate": 1.91869918699187e-05,
"loss": 0.9388,
"step": 59
},
{
"epoch": 0.2436548223350254,
"grad_norm": 1.1617787103967965,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.947,
"step": 60
},
{
"epoch": 0.24771573604060915,
"grad_norm": 1.1354482550335967,
"learning_rate": 1.983739837398374e-05,
"loss": 0.9391,
"step": 61
},
{
"epoch": 0.2517766497461929,
"grad_norm": 1.3242366978315538,
"learning_rate": 2.016260162601626e-05,
"loss": 0.9107,
"step": 62
},
{
"epoch": 0.25583756345177666,
"grad_norm": 1.1395001987487705,
"learning_rate": 2.048780487804878e-05,
"loss": 0.9504,
"step": 63
},
{
"epoch": 0.2598984771573604,
"grad_norm": 0.9928039484149661,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.9206,
"step": 64
},
{
"epoch": 0.2639593908629442,
"grad_norm": 0.9320816198662614,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.9186,
"step": 65
},
{
"epoch": 0.26802030456852793,
"grad_norm": 1.2766110623793032,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.9144,
"step": 66
},
{
"epoch": 0.2720812182741117,
"grad_norm": 0.839313541450921,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.9133,
"step": 67
},
{
"epoch": 0.27614213197969545,
"grad_norm": 1.282384554332314,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.9207,
"step": 68
},
{
"epoch": 0.2802030456852792,
"grad_norm": 0.8111838756570011,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.8936,
"step": 69
},
{
"epoch": 0.28426395939086296,
"grad_norm": 0.9622977658184423,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.8851,
"step": 70
},
{
"epoch": 0.2883248730964467,
"grad_norm": 0.7758344644787591,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.9121,
"step": 71
},
{
"epoch": 0.2923857868020305,
"grad_norm": 0.7562640940886655,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.9339,
"step": 72
},
{
"epoch": 0.29644670050761424,
"grad_norm": 0.8905332558044384,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.9306,
"step": 73
},
{
"epoch": 0.300507614213198,
"grad_norm": 1.2270641158421352,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.9235,
"step": 74
},
{
"epoch": 0.30456852791878175,
"grad_norm": 1.0656954222404864,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.9238,
"step": 75
},
{
"epoch": 0.3086294416243655,
"grad_norm": 1.2196245463670141,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.9497,
"step": 76
},
{
"epoch": 0.31269035532994927,
"grad_norm": 1.150393707175204,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.9216,
"step": 77
},
{
"epoch": 0.31675126903553297,
"grad_norm": 1.090410484121022,
"learning_rate": 2.536585365853659e-05,
"loss": 0.918,
"step": 78
},
{
"epoch": 0.3208121827411167,
"grad_norm": 1.235719658980269,
"learning_rate": 2.569105691056911e-05,
"loss": 0.9219,
"step": 79
},
{
"epoch": 0.3248730964467005,
"grad_norm": 1.2523455830778798,
"learning_rate": 2.601626016260163e-05,
"loss": 0.9255,
"step": 80
},
{
"epoch": 0.32893401015228424,
"grad_norm": 1.209276984792975,
"learning_rate": 2.634146341463415e-05,
"loss": 0.93,
"step": 81
},
{
"epoch": 0.332994923857868,
"grad_norm": 0.9630847912963698,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.9115,
"step": 82
},
{
"epoch": 0.33705583756345175,
"grad_norm": 1.4372311838774192,
"learning_rate": 2.699186991869919e-05,
"loss": 0.8968,
"step": 83
},
{
"epoch": 0.3411167512690355,
"grad_norm": 1.0825911155684622,
"learning_rate": 2.731707317073171e-05,
"loss": 0.9312,
"step": 84
},
{
"epoch": 0.34517766497461927,
"grad_norm": 1.2645456054818507,
"learning_rate": 2.764227642276423e-05,
"loss": 0.9099,
"step": 85
},
{
"epoch": 0.349238578680203,
"grad_norm": 1.371893094379104,
"learning_rate": 2.796747967479675e-05,
"loss": 0.909,
"step": 86
},
{
"epoch": 0.3532994923857868,
"grad_norm": 0.9776119228492275,
"learning_rate": 2.829268292682927e-05,
"loss": 0.9037,
"step": 87
},
{
"epoch": 0.35736040609137054,
"grad_norm": 1.1683176873484633,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.9062,
"step": 88
},
{
"epoch": 0.3614213197969543,
"grad_norm": 1.832036981881903,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.9205,
"step": 89
},
{
"epoch": 0.36548223350253806,
"grad_norm": 1.1194694253811004,
"learning_rate": 2.926829268292683e-05,
"loss": 0.9003,
"step": 90
},
{
"epoch": 0.3695431472081218,
"grad_norm": 1.9711060947222045,
"learning_rate": 2.959349593495935e-05,
"loss": 0.9194,
"step": 91
},
{
"epoch": 0.37360406091370557,
"grad_norm": 1.6491355556209406,
"learning_rate": 2.991869918699187e-05,
"loss": 0.9349,
"step": 92
},
{
"epoch": 0.37766497461928933,
"grad_norm": 1.4354683835503652,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.9328,
"step": 93
},
{
"epoch": 0.3817258883248731,
"grad_norm": 1.3309451738696487,
"learning_rate": 3.056910569105692e-05,
"loss": 0.9161,
"step": 94
},
{
"epoch": 0.38578680203045684,
"grad_norm": 1.5811084801844135,
"learning_rate": 3.089430894308943e-05,
"loss": 0.9188,
"step": 95
},
{
"epoch": 0.3898477157360406,
"grad_norm": 1.176548775430425,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.9051,
"step": 96
},
{
"epoch": 0.39390862944162436,
"grad_norm": 1.2122534887391603,
"learning_rate": 3.154471544715447e-05,
"loss": 0.8937,
"step": 97
},
{
"epoch": 0.3979695431472081,
"grad_norm": 1.12501344622318,
"learning_rate": 3.186991869918699e-05,
"loss": 0.8748,
"step": 98
},
{
"epoch": 0.4020304568527919,
"grad_norm": 1.5974353003522328,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.9434,
"step": 99
},
{
"epoch": 0.40609137055837563,
"grad_norm": 1.274490925200243,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.8936,
"step": 100
},
{
"epoch": 0.4101522842639594,
"grad_norm": 1.3596273471267648,
"learning_rate": 3.284552845528456e-05,
"loss": 0.9214,
"step": 101
},
{
"epoch": 0.41421319796954315,
"grad_norm": 1.1164242189125648,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.9074,
"step": 102
},
{
"epoch": 0.4182741116751269,
"grad_norm": 1.7027932599729587,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.9167,
"step": 103
},
{
"epoch": 0.42233502538071066,
"grad_norm": 1.3498510889282662,
"learning_rate": 3.382113821138212e-05,
"loss": 0.8972,
"step": 104
},
{
"epoch": 0.4263959390862944,
"grad_norm": 1.2966464251641212,
"learning_rate": 3.414634146341463e-05,
"loss": 0.8847,
"step": 105
},
{
"epoch": 0.4304568527918782,
"grad_norm": 1.3361713238901625,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.8965,
"step": 106
},
{
"epoch": 0.43451776649746193,
"grad_norm": 1.3007432538326449,
"learning_rate": 3.479674796747968e-05,
"loss": 0.9055,
"step": 107
},
{
"epoch": 0.4385786802030457,
"grad_norm": 1.3616170274936827,
"learning_rate": 3.51219512195122e-05,
"loss": 0.885,
"step": 108
},
{
"epoch": 0.44263959390862945,
"grad_norm": 1.1151752619167619,
"learning_rate": 3.544715447154472e-05,
"loss": 0.9021,
"step": 109
},
{
"epoch": 0.4467005076142132,
"grad_norm": 1.4081969056893318,
"learning_rate": 3.577235772357724e-05,
"loss": 0.9121,
"step": 110
},
{
"epoch": 0.45076142131979696,
"grad_norm": 1.3636162242197252,
"learning_rate": 3.609756097560976e-05,
"loss": 0.9209,
"step": 111
},
{
"epoch": 0.4548223350253807,
"grad_norm": 1.3529060508094846,
"learning_rate": 3.642276422764228e-05,
"loss": 0.9087,
"step": 112
},
{
"epoch": 0.4588832487309645,
"grad_norm": 1.0863826570098778,
"learning_rate": 3.67479674796748e-05,
"loss": 0.8899,
"step": 113
},
{
"epoch": 0.46294416243654823,
"grad_norm": 1.8508252011914397,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.9056,
"step": 114
},
{
"epoch": 0.467005076142132,
"grad_norm": 1.0882435369896517,
"learning_rate": 3.739837398373984e-05,
"loss": 0.8992,
"step": 115
},
{
"epoch": 0.47106598984771575,
"grad_norm": 1.548103089374484,
"learning_rate": 3.772357723577236e-05,
"loss": 0.8992,
"step": 116
},
{
"epoch": 0.4751269035532995,
"grad_norm": 2.105575269063434,
"learning_rate": 3.804878048780488e-05,
"loss": 0.8969,
"step": 117
},
{
"epoch": 0.47918781725888326,
"grad_norm": 1.1172540075433763,
"learning_rate": 3.83739837398374e-05,
"loss": 0.8839,
"step": 118
},
{
"epoch": 0.483248730964467,
"grad_norm": 2.096224983174092,
"learning_rate": 3.869918699186992e-05,
"loss": 0.9189,
"step": 119
},
{
"epoch": 0.4873096446700508,
"grad_norm": 1.6004033011411418,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.8718,
"step": 120
},
{
"epoch": 0.49137055837563454,
"grad_norm": 1.9947362072601837,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.918,
"step": 121
},
{
"epoch": 0.4954314720812183,
"grad_norm": 1.3759430526372507,
"learning_rate": 3.967479674796748e-05,
"loss": 0.8972,
"step": 122
},
{
"epoch": 0.49949238578680205,
"grad_norm": 1.9098124889098713,
"learning_rate": 4e-05,
"loss": 0.9019,
"step": 123
},
{
"epoch": 0.5035532994923858,
"grad_norm": 1.082614628150109,
"learning_rate": 3.999991946137476e-05,
"loss": 0.9028,
"step": 124
},
{
"epoch": 0.5076142131979695,
"grad_norm": 2.559425102628464,
"learning_rate": 3.999967784614766e-05,
"loss": 0.9214,
"step": 125
},
{
"epoch": 0.5116751269035533,
"grad_norm": 2.2365217011388614,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.9241,
"step": 126
},
{
"epoch": 0.515736040609137,
"grad_norm": 1.8063156574474821,
"learning_rate": 3.999871139496895e-05,
"loss": 0.9111,
"step": 127
},
{
"epoch": 0.5197969543147208,
"grad_norm": 1.7083344676113184,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.8995,
"step": 128
},
{
"epoch": 0.5238578680203045,
"grad_norm": 1.729611466678804,
"learning_rate": 3.999710067759846e-05,
"loss": 0.8946,
"step": 129
},
{
"epoch": 0.5279187817258884,
"grad_norm": 1.4869825271287,
"learning_rate": 3.999605373449617e-05,
"loss": 0.9068,
"step": 130
},
{
"epoch": 0.531979695431472,
"grad_norm": 1.3561492611073167,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.9025,
"step": 131
},
{
"epoch": 0.5360406091370559,
"grad_norm": 1.0385530287552618,
"learning_rate": 3.999347672161713e-05,
"loss": 0.8899,
"step": 132
},
{
"epoch": 0.5401015228426396,
"grad_norm": 1.5019294934265033,
"learning_rate": 3.999194667259528e-05,
"loss": 0.8857,
"step": 133
},
{
"epoch": 0.5441624365482234,
"grad_norm": 1.4418973667053565,
"learning_rate": 3.999025561118334e-05,
"loss": 0.8958,
"step": 134
},
{
"epoch": 0.5482233502538071,
"grad_norm": 1.2572340067708199,
"learning_rate": 3.998840355100086e-05,
"loss": 0.9194,
"step": 135
},
{
"epoch": 0.5522842639593909,
"grad_norm": 1.0609439598407637,
"learning_rate": 3.998639050696409e-05,
"loss": 0.8905,
"step": 136
},
{
"epoch": 0.5563451776649746,
"grad_norm": 1.828464746740132,
"learning_rate": 3.998421649528582e-05,
"loss": 0.9098,
"step": 137
},
{
"epoch": 0.5604060913705584,
"grad_norm": 1.3123199894067794,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.8963,
"step": 138
},
{
"epoch": 0.5644670050761421,
"grad_norm": 1.5113191363186305,
"learning_rate": 3.997938564033779e-05,
"loss": 0.9081,
"step": 139
},
{
"epoch": 0.5685279187817259,
"grad_norm": 1.4594144525399735,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.904,
"step": 140
},
{
"epoch": 0.5725888324873096,
"grad_norm": 1.0739239964028016,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.8955,
"step": 141
},
{
"epoch": 0.5766497461928934,
"grad_norm": 1.1044482971871308,
"learning_rate": 3.997093258045973e-05,
"loss": 0.9116,
"step": 142
},
{
"epoch": 0.5807106598984771,
"grad_norm": 1.3541917407610233,
"learning_rate": 3.996779317598936e-05,
"loss": 0.872,
"step": 143
},
{
"epoch": 0.584771573604061,
"grad_norm": 1.4788266851172207,
"learning_rate": 3.996449295365782e-05,
"loss": 0.9205,
"step": 144
},
{
"epoch": 0.5888324873096447,
"grad_norm": 1.3629106749425273,
"learning_rate": 3.996103194004467e-05,
"loss": 0.8713,
"step": 145
},
{
"epoch": 0.5928934010152285,
"grad_norm": 2.4427358502687513,
"learning_rate": 3.995741016302441e-05,
"loss": 0.8939,
"step": 146
},
{
"epoch": 0.5969543147208122,
"grad_norm": 1.6797175024449704,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.8754,
"step": 147
},
{
"epoch": 0.601015228426396,
"grad_norm": 2.656246713221402,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.9144,
"step": 148
},
{
"epoch": 0.6050761421319797,
"grad_norm": 2.2256619791977332,
"learning_rate": 3.994558054968643e-05,
"loss": 0.8902,
"step": 149
},
{
"epoch": 0.6091370558375635,
"grad_norm": 1.902247994330618,
"learning_rate": 3.994131602367481e-05,
"loss": 0.887,
"step": 150
},
{
"epoch": 0.6131979695431472,
"grad_norm": 1.4339019895673746,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.9245,
"step": 151
},
{
"epoch": 0.617258883248731,
"grad_norm": 2.006879516844076,
"learning_rate": 3.993230519343752e-05,
"loss": 0.8953,
"step": 152
},
{
"epoch": 0.6213197969543147,
"grad_norm": 1.4766747212288462,
"learning_rate": 3.992755896178383e-05,
"loss": 0.9287,
"step": 153
},
{
"epoch": 0.6253807106598985,
"grad_norm": 1.6612016246841708,
"learning_rate": 3.992265223630981e-05,
"loss": 0.9113,
"step": 154
},
{
"epoch": 0.6294416243654822,
"grad_norm": 1.2524581232021137,
"learning_rate": 3.991758505653355e-05,
"loss": 0.9208,
"step": 155
},
{
"epoch": 0.6335025380710659,
"grad_norm": 0.8947457667466222,
"learning_rate": 3.991235746326543e-05,
"loss": 0.8935,
"step": 156
},
{
"epoch": 0.6375634517766497,
"grad_norm": 1.1872808865574065,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.8804,
"step": 157
},
{
"epoch": 0.6416243654822334,
"grad_norm": 1.1607741588867717,
"learning_rate": 3.990142120595444e-05,
"loss": 0.8717,
"step": 158
},
{
"epoch": 0.6456852791878173,
"grad_norm": 1.1301289500098948,
"learning_rate": 3.98957126299907e-05,
"loss": 0.8694,
"step": 159
},
{
"epoch": 0.649746192893401,
"grad_norm": 1.1062587867402132,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.9092,
"step": 160
},
{
"epoch": 0.6538071065989848,
"grad_norm": 1.4735621770055154,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.8964,
"step": 161
},
{
"epoch": 0.6578680203045685,
"grad_norm": 0.7922607397953237,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.896,
"step": 162
},
{
"epoch": 0.6619289340101523,
"grad_norm": 0.8564877101370162,
"learning_rate": 3.98712764319087e-05,
"loss": 0.8746,
"step": 163
},
{
"epoch": 0.665989847715736,
"grad_norm": 1.0659363574068204,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.9058,
"step": 164
},
{
"epoch": 0.6700507614213198,
"grad_norm": 1.1373363162559778,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.8672,
"step": 165
},
{
"epoch": 0.6741116751269035,
"grad_norm": 0.8084407152789492,
"learning_rate": 3.985126869010129e-05,
"loss": 0.8885,
"step": 166
},
{
"epoch": 0.6781725888324873,
"grad_norm": 1.1688586467424646,
"learning_rate": 3.984427961114833e-05,
"loss": 0.8966,
"step": 167
},
{
"epoch": 0.682233502538071,
"grad_norm": 0.9545097994276284,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.9262,
"step": 168
},
{
"epoch": 0.6862944162436548,
"grad_norm": 0.8306555701034513,
"learning_rate": 3.982982204151901e-05,
"loss": 0.8884,
"step": 169
},
{
"epoch": 0.6903553299492385,
"grad_norm": 0.896887761734214,
"learning_rate": 3.982235366728193e-05,
"loss": 0.8773,
"step": 170
},
{
"epoch": 0.6944162436548224,
"grad_norm": 1.0750114859521285,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.8884,
"step": 171
},
{
"epoch": 0.698477157360406,
"grad_norm": 1.398388939662489,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.8754,
"step": 172
},
{
"epoch": 0.7025380710659899,
"grad_norm": 0.9809999765020677,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.882,
"step": 173
},
{
"epoch": 0.7065989847715736,
"grad_norm": 1.2150932671667536,
"learning_rate": 3.979088432599627e-05,
"loss": 0.8752,
"step": 174
},
{
"epoch": 0.7106598984771574,
"grad_norm": 0.7965308868517167,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.857,
"step": 175
},
{
"epoch": 0.7147208121827411,
"grad_norm": 0.7367530890948703,
"learning_rate": 3.977419304032111e-05,
"loss": 0.8654,
"step": 176
},
{
"epoch": 0.7187817258883249,
"grad_norm": 0.8664659272194589,
"learning_rate": 3.976560847560697e-05,
"loss": 0.8746,
"step": 177
},
{
"epoch": 0.7228426395939086,
"grad_norm": 0.9052609132468968,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.8682,
"step": 178
},
{
"epoch": 0.7269035532994924,
"grad_norm": 0.8704482585060862,
"learning_rate": 3.974796184811956e-05,
"loss": 0.888,
"step": 179
},
{
"epoch": 0.7309644670050761,
"grad_norm": 0.7867736193455253,
"learning_rate": 3.973889992746979e-05,
"loss": 0.9054,
"step": 180
},
{
"epoch": 0.7350253807106599,
"grad_norm": 1.1466452926946205,
"learning_rate": 3.972967903243361e-05,
"loss": 0.8982,
"step": 181
},
{
"epoch": 0.7390862944162436,
"grad_norm": 1.3303110617626006,
"learning_rate": 3.972029923727486e-05,
"loss": 0.8733,
"step": 182
},
{
"epoch": 0.7431472081218274,
"grad_norm": 1.1064217635357834,
"learning_rate": 3.971076061753709e-05,
"loss": 0.876,
"step": 183
},
{
"epoch": 0.7472081218274111,
"grad_norm": 0.8736652299666252,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.9041,
"step": 184
},
{
"epoch": 0.751269035532995,
"grad_norm": 0.6874008145028504,
"learning_rate": 3.969120721289402e-05,
"loss": 0.8986,
"step": 185
},
{
"epoch": 0.7553299492385787,
"grad_norm": 0.5366622765527447,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.8768,
"step": 186
},
{
"epoch": 0.7593908629441625,
"grad_norm": 0.7019165424708133,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.8826,
"step": 187
},
{
"epoch": 0.7634517766497462,
"grad_norm": 1.0041895606249729,
"learning_rate": 3.966068788369422e-05,
"loss": 0.9054,
"step": 188
},
{
"epoch": 0.76751269035533,
"grad_norm": 1.12437982338163,
"learning_rate": 3.965019797448622e-05,
"loss": 0.9073,
"step": 189
},
{
"epoch": 0.7715736040609137,
"grad_norm": 1.0582060197050378,
"learning_rate": 3.963954980528515e-05,
"loss": 0.9264,
"step": 190
},
{
"epoch": 0.7756345177664975,
"grad_norm": 1.365876984339088,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.876,
"step": 191
},
{
"epoch": 0.7796954314720812,
"grad_norm": 0.8184426932769878,
"learning_rate": 3.961777903121329e-05,
"loss": 0.8864,
"step": 192
},
{
"epoch": 0.783756345177665,
"grad_norm": 0.6577528037342761,
"learning_rate": 3.960665660168131e-05,
"loss": 0.8727,
"step": 193
},
{
"epoch": 0.7878172588832487,
"grad_norm": 0.843479299371653,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.8894,
"step": 194
},
{
"epoch": 0.7918781725888325,
"grad_norm": 1.3447235795085584,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.8784,
"step": 195
},
{
"epoch": 0.7959390862944162,
"grad_norm": 1.0975742696354134,
"learning_rate": 3.957234222185657e-05,
"loss": 0.8975,
"step": 196
},
{
"epoch": 0.8,
"grad_norm": 0.7498545128384897,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.8948,
"step": 197
},
{
"epoch": 0.8040609137055837,
"grad_norm": 0.7117837345568039,
"learning_rate": 3.954867765033605e-05,
"loss": 0.894,
"step": 198
},
{
"epoch": 0.8081218274111676,
"grad_norm": 0.6808730780261149,
"learning_rate": 3.953660915306728e-05,
"loss": 0.8652,
"step": 199
},
{
"epoch": 0.8121827411167513,
"grad_norm": 0.8124656346643899,
"learning_rate": 3.952438331063419e-05,
"loss": 0.8833,
"step": 200
},
{
"epoch": 0.8162436548223351,
"grad_norm": 0.9400758170270892,
"learning_rate": 3.951200022150205e-05,
"loss": 0.8854,
"step": 201
},
{
"epoch": 0.8203045685279188,
"grad_norm": 1.1049216056060052,
"learning_rate": 3.949945998540253e-05,
"loss": 0.874,
"step": 202
},
{
"epoch": 0.8243654822335026,
"grad_norm": 1.1048812029285373,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.8589,
"step": 203
},
{
"epoch": 0.8284263959390863,
"grad_norm": 1.1303699376953664,
"learning_rate": 3.947390847755559e-05,
"loss": 0.8798,
"step": 204
},
{
"epoch": 0.8324873096446701,
"grad_norm": 0.8848178315244256,
"learning_rate": 3.946089741159648e-05,
"loss": 0.8922,
"step": 205
},
{
"epoch": 0.8365482233502538,
"grad_norm": 0.7503975377719104,
"learning_rate": 3.944772961024501e-05,
"loss": 0.8841,
"step": 206
},
{
"epoch": 0.8406091370558376,
"grad_norm": 0.5971919276044977,
"learning_rate": 3.943440517955285e-05,
"loss": 0.8832,
"step": 207
},
{
"epoch": 0.8446700507614213,
"grad_norm": 0.5979145635380856,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.8713,
"step": 208
},
{
"epoch": 0.848730964467005,
"grad_norm": 0.6732671100669172,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.849,
"step": 209
},
{
"epoch": 0.8527918781725888,
"grad_norm": 0.6174177952103478,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.8839,
"step": 210
},
{
"epoch": 0.8568527918781725,
"grad_norm": 0.7612283827735356,
"learning_rate": 3.937954332854371e-05,
"loss": 0.9072,
"step": 211
},
{
"epoch": 0.8609137055837564,
"grad_norm": 0.6636998492164042,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.8758,
"step": 212
},
{
"epoch": 0.86497461928934,
"grad_norm": 0.5883924354128413,
"learning_rate": 3.935117547697387e-05,
"loss": 0.8653,
"step": 213
},
{
"epoch": 0.8690355329949239,
"grad_norm": 0.635271334829329,
"learning_rate": 3.933675771619675e-05,
"loss": 0.8697,
"step": 214
},
{
"epoch": 0.8730964467005076,
"grad_norm": 0.6520681868719912,
"learning_rate": 3.932218421983131e-05,
"loss": 0.8724,
"step": 215
},
{
"epoch": 0.8771573604060914,
"grad_norm": 0.8942318463482811,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.869,
"step": 216
},
{
"epoch": 0.8812182741116751,
"grad_norm": 1.0723982134418812,
"learning_rate": 3.929257049108054e-05,
"loss": 0.8595,
"step": 217
},
{
"epoch": 0.8852791878172589,
"grad_norm": 0.9862484313163941,
"learning_rate": 3.927753049720011e-05,
"loss": 0.8539,
"step": 218
},
{
"epoch": 0.8893401015228426,
"grad_norm": 1.1745551086540567,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.8627,
"step": 219
},
{
"epoch": 0.8934010152284264,
"grad_norm": 0.9315083146027145,
"learning_rate": 3.92469848560784e-05,
"loss": 0.8772,
"step": 220
},
{
"epoch": 0.8974619289340101,
"grad_norm": 0.6765390361868112,
"learning_rate": 3.923147945484751e-05,
"loss": 0.8615,
"step": 221
},
{
"epoch": 0.9015228426395939,
"grad_norm": 0.6855267721984877,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.8901,
"step": 222
},
{
"epoch": 0.9055837563451776,
"grad_norm": 0.7531347860395485,
"learning_rate": 3.920000411543654e-05,
"loss": 0.8586,
"step": 223
},
{
"epoch": 0.9096446700507614,
"grad_norm": 0.7263419004831295,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.8704,
"step": 224
},
{
"epoch": 0.9137055837563451,
"grad_norm": 0.7549773029380547,
"learning_rate": 3.916791024049648e-05,
"loss": 0.88,
"step": 225
},
{
"epoch": 0.917766497461929,
"grad_norm": 0.8335613254633946,
"learning_rate": 3.91516316745245e-05,
"loss": 0.8728,
"step": 226
},
{
"epoch": 0.9218274111675127,
"grad_norm": 0.9835568669580276,
"learning_rate": 3.913519886394389e-05,
"loss": 0.8687,
"step": 227
},
{
"epoch": 0.9258883248730965,
"grad_norm": 1.2322022298768474,
"learning_rate": 3.911861194110225e-05,
"loss": 0.8779,
"step": 228
},
{
"epoch": 0.9299492385786802,
"grad_norm": 0.7423357752207448,
"learning_rate": 3.910187103958837e-05,
"loss": 0.8625,
"step": 229
},
{
"epoch": 0.934010152284264,
"grad_norm": 0.5737924971648999,
"learning_rate": 3.908497629423117e-05,
"loss": 0.8586,
"step": 230
},
{
"epoch": 0.9380710659898477,
"grad_norm": 0.8786906523233956,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.8788,
"step": 231
},
{
"epoch": 0.9421319796954315,
"grad_norm": 0.98193260573937,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.847,
"step": 232
},
{
"epoch": 0.9461928934010152,
"grad_norm": 0.7913892945281108,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.8747,
"step": 233
},
{
"epoch": 0.950253807106599,
"grad_norm": 0.7577447884683057,
"learning_rate": 3.901586161429081e-05,
"loss": 0.8759,
"step": 234
},
{
"epoch": 0.9543147208121827,
"grad_norm": 0.6759329196235779,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.8721,
"step": 235
},
{
"epoch": 0.9583756345177665,
"grad_norm": 0.5931549570210413,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.8727,
"step": 236
},
{
"epoch": 0.9624365482233502,
"grad_norm": 0.5756110565640363,
"learning_rate": 3.896241703466667e-05,
"loss": 0.8743,
"step": 237
},
{
"epoch": 0.966497461928934,
"grad_norm": 0.6529948707200075,
"learning_rate": 3.894429654085585e-05,
"loss": 0.8696,
"step": 238
},
{
"epoch": 0.9705583756345177,
"grad_norm": 0.5895747055634767,
"learning_rate": 3.892602347228505e-05,
"loss": 0.8534,
"step": 239
},
{
"epoch": 0.9746192893401016,
"grad_norm": 0.5410817057573484,
"learning_rate": 3.890759797612307e-05,
"loss": 0.8662,
"step": 240
},
{
"epoch": 0.9786802030456853,
"grad_norm": 0.6646080065582237,
"learning_rate": 3.888902020076632e-05,
"loss": 0.8743,
"step": 241
},
{
"epoch": 0.9827411167512691,
"grad_norm": 0.7829869432980208,
"learning_rate": 3.887029029583764e-05,
"loss": 0.8689,
"step": 242
},
{
"epoch": 0.9868020304568528,
"grad_norm": 1.0596085542111642,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.8683,
"step": 243
},
{
"epoch": 0.9908629441624366,
"grad_norm": 0.911377363067641,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.8889,
"step": 244
},
{
"epoch": 0.9949238578680203,
"grad_norm": 1.2488366056373077,
"learning_rate": 3.881318931821972e-05,
"loss": 0.8702,
"step": 245
},
{
"epoch": 0.9989847715736041,
"grad_norm": 0.7195805367184755,
"learning_rate": 3.879385241571817e-05,
"loss": 0.8583,
"step": 246
},
{
"epoch": 1.0030456852791878,
"grad_norm": 1.2718149151821105,
"learning_rate": 3.8774364150112955e-05,
"loss": 1.5016,
"step": 247
},
{
"epoch": 1.0071065989847716,
"grad_norm": 1.2228390639433937,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.841,
"step": 248
},
{
"epoch": 1.0111675126903554,
"grad_norm": 0.6706593945110678,
"learning_rate": 3.873493415863256e-05,
"loss": 0.8529,
"step": 249
},
{
"epoch": 1.015228426395939,
"grad_norm": 1.1194033577934361,
"learning_rate": 3.871499275032111e-05,
"loss": 0.8444,
"step": 250
},
{
"epoch": 1.0192893401015228,
"grad_norm": 0.9046523505591721,
"learning_rate": 3.869490061403091e-05,
"loss": 0.8034,
"step": 251
},
{
"epoch": 1.0233502538071066,
"grad_norm": 0.9231951220105618,
"learning_rate": 3.867465791158124e-05,
"loss": 0.8495,
"step": 252
},
{
"epoch": 1.0274111675126905,
"grad_norm": 1.1108905063748793,
"learning_rate": 3.865426480600407e-05,
"loss": 0.8886,
"step": 253
},
{
"epoch": 1.031472081218274,
"grad_norm": 1.069777537016959,
"learning_rate": 3.863372146154264e-05,
"loss": 0.8365,
"step": 254
},
{
"epoch": 1.0355329949238579,
"grad_norm": 0.9934931436409761,
"learning_rate": 3.861302804365024e-05,
"loss": 0.8315,
"step": 255
},
{
"epoch": 1.0395939086294417,
"grad_norm": 0.8067455786134673,
"learning_rate": 3.85921847189888e-05,
"loss": 0.8332,
"step": 256
},
{
"epoch": 1.0436548223350255,
"grad_norm": 0.7489125843311554,
"learning_rate": 3.85711916554276e-05,
"loss": 0.8285,
"step": 257
},
{
"epoch": 1.047715736040609,
"grad_norm": 0.7299894747921222,
"learning_rate": 3.85500490220419e-05,
"loss": 0.86,
"step": 258
},
{
"epoch": 1.051776649746193,
"grad_norm": 0.7066648467209312,
"learning_rate": 3.852875698911154e-05,
"loss": 0.8253,
"step": 259
},
{
"epoch": 1.0558375634517767,
"grad_norm": 0.7587559320224089,
"learning_rate": 3.850731572811963e-05,
"loss": 0.8312,
"step": 260
},
{
"epoch": 1.0598984771573603,
"grad_norm": 0.961795434361132,
"learning_rate": 3.848572541175116e-05,
"loss": 0.8458,
"step": 261
},
{
"epoch": 1.063959390862944,
"grad_norm": 1.22462106606504,
"learning_rate": 3.846398621389154e-05,
"loss": 0.8625,
"step": 262
},
{
"epoch": 1.068020304568528,
"grad_norm": 0.6507218661867931,
"learning_rate": 3.84420983096253e-05,
"loss": 0.8236,
"step": 263
},
{
"epoch": 1.0720812182741117,
"grad_norm": 1.0756974743905785,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.8525,
"step": 264
},
{
"epoch": 1.0761421319796955,
"grad_norm": 1.4751285556259122,
"learning_rate": 3.839787708819787e-05,
"loss": 0.8535,
"step": 265
},
{
"epoch": 1.0802030456852791,
"grad_norm": 0.5078725621691786,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.8238,
"step": 266
},
{
"epoch": 1.084263959390863,
"grad_norm": 1.4024156360173399,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.8461,
"step": 267
},
{
"epoch": 1.0883248730964468,
"grad_norm": 0.7028446018672242,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.8437,
"step": 268
},
{
"epoch": 1.0923857868020304,
"grad_norm": 0.7439443483427532,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.8528,
"step": 269
},
{
"epoch": 1.0964467005076142,
"grad_norm": 0.9142959001379658,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.8118,
"step": 270
},
{
"epoch": 1.100507614213198,
"grad_norm": 0.8140387316882918,
"learning_rate": 3.826166304956594e-05,
"loss": 0.8265,
"step": 271
},
{
"epoch": 1.1045685279187818,
"grad_norm": 0.967359818436721,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.8326,
"step": 272
},
{
"epoch": 1.1086294416243654,
"grad_norm": 0.9352312088978563,
"learning_rate": 3.821507978767479e-05,
"loss": 0.8442,
"step": 273
},
{
"epoch": 1.1126903553299492,
"grad_norm": 0.7817192574090527,
"learning_rate": 3.819156801001693e-05,
"loss": 0.863,
"step": 274
},
{
"epoch": 1.116751269035533,
"grad_norm": 0.6735200770668788,
"learning_rate": 3.816790971997121e-05,
"loss": 0.8266,
"step": 275
},
{
"epoch": 1.1208121827411168,
"grad_norm": 0.8605010951047443,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.8526,
"step": 276
},
{
"epoch": 1.1248730964467004,
"grad_norm": 1.0227467527268905,
"learning_rate": 3.81201543660571e-05,
"loss": 0.8444,
"step": 277
},
{
"epoch": 1.1289340101522842,
"grad_norm": 0.7987787785636662,
"learning_rate": 3.809605768680377e-05,
"loss": 0.8391,
"step": 278
},
{
"epoch": 1.132994923857868,
"grad_norm": 0.6177334752184757,
"learning_rate": 3.807181526438958e-05,
"loss": 0.8547,
"step": 279
},
{
"epoch": 1.1370558375634519,
"grad_norm": 0.8315819027056006,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.8563,
"step": 280
},
{
"epoch": 1.1411167512690354,
"grad_norm": 0.7573786183934449,
"learning_rate": 3.802289397223145e-05,
"loss": 0.8316,
"step": 281
},
{
"epoch": 1.1451776649746193,
"grad_norm": 0.5962393811766117,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.8489,
"step": 282
},
{
"epoch": 1.149238578680203,
"grad_norm": 0.5879813743061921,
"learning_rate": 3.797339206560096e-05,
"loss": 0.8619,
"step": 283
},
{
"epoch": 1.1532994923857869,
"grad_norm": 0.5811752126956523,
"learning_rate": 3.794842387948027e-05,
"loss": 0.8428,
"step": 284
},
{
"epoch": 1.1573604060913705,
"grad_norm": 0.5899020662123778,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.8457,
"step": 285
},
{
"epoch": 1.1614213197969543,
"grad_norm": 0.6207512309428056,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.8353,
"step": 286
},
{
"epoch": 1.165482233502538,
"grad_norm": 0.5339900916047498,
"learning_rate": 3.787265280646825e-05,
"loss": 0.8552,
"step": 287
},
{
"epoch": 1.169543147208122,
"grad_norm": 0.6444599393545426,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.8438,
"step": 288
},
{
"epoch": 1.1736040609137055,
"grad_norm": 0.5433275865967176,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.8159,
"step": 289
},
{
"epoch": 1.1776649746192893,
"grad_norm": 0.653360301844724,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.8549,
"step": 290
},
{
"epoch": 1.1817258883248731,
"grad_norm": 0.6912622334819194,
"learning_rate": 3.776961046830306e-05,
"loss": 0.8542,
"step": 291
},
{
"epoch": 1.185786802030457,
"grad_norm": 0.6630563047663287,
"learning_rate": 3.774349157719215e-05,
"loss": 0.834,
"step": 292
},
{
"epoch": 1.1898477157360405,
"grad_norm": 0.8737868046028318,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.8324,
"step": 293
},
{
"epoch": 1.1939086294416243,
"grad_norm": 0.7347359129644747,
"learning_rate": 3.769082529555359e-05,
"loss": 0.859,
"step": 294
},
{
"epoch": 1.1979695431472082,
"grad_norm": 0.6248164010893552,
"learning_rate": 3.766427832919294e-05,
"loss": 0.7832,
"step": 295
},
{
"epoch": 1.202030456852792,
"grad_norm": 0.6606792337066806,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.8124,
"step": 296
},
{
"epoch": 1.2060913705583756,
"grad_norm": 0.7038523403331013,
"learning_rate": 3.761075781441526e-05,
"loss": 0.8307,
"step": 297
},
{
"epoch": 1.2101522842639594,
"grad_norm": 0.6283740105271951,
"learning_rate": 3.75837846970451e-05,
"loss": 0.8561,
"step": 298
},
{
"epoch": 1.2142131979695432,
"grad_norm": 0.6254476058893971,
"learning_rate": 3.755666996229032e-05,
"loss": 0.8336,
"step": 299
},
{
"epoch": 1.218274111675127,
"grad_norm": 0.6323427666126807,
"learning_rate": 3.752941382852927e-05,
"loss": 0.8484,
"step": 300
},
{
"epoch": 1.2223350253807106,
"grad_norm": 0.7335318018317287,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.8155,
"step": 301
},
{
"epoch": 1.2263959390862944,
"grad_norm": 0.4934412190629873,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.862,
"step": 302
},
{
"epoch": 1.2304568527918782,
"grad_norm": 0.5155083972739437,
"learning_rate": 3.744679923406351e-05,
"loss": 0.8339,
"step": 303
},
{
"epoch": 1.2345177664974618,
"grad_norm": 0.5858260171823958,
"learning_rate": 3.741897971081043e-05,
"loss": 0.8541,
"step": 304
},
{
"epoch": 1.2385786802030456,
"grad_norm": 0.40878814245310663,
"learning_rate": 3.739101989748946e-05,
"loss": 0.8254,
"step": 305
},
{
"epoch": 1.2426395939086294,
"grad_norm": 0.4948015703887049,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.8469,
"step": 306
},
{
"epoch": 1.2467005076142132,
"grad_norm": 0.48431379760652044,
"learning_rate": 3.73346803025098e-05,
"loss": 0.8293,
"step": 307
},
{
"epoch": 1.250761421319797,
"grad_norm": 0.7345247518550934,
"learning_rate": 3.730630097460247e-05,
"loss": 0.834,
"step": 308
},
{
"epoch": 1.2548223350253807,
"grad_norm": 0.8095303549251766,
"learning_rate": 3.727778226412628e-05,
"loss": 0.8589,
"step": 309
},
{
"epoch": 1.2588832487309645,
"grad_norm": 0.6266535188783025,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.8294,
"step": 310
},
{
"epoch": 1.2629441624365483,
"grad_norm": 0.5606213487931734,
"learning_rate": 3.722032761533114e-05,
"loss": 0.8506,
"step": 311
},
{
"epoch": 1.2670050761421319,
"grad_norm": 0.5235648877677851,
"learning_rate": 3.719139213974403e-05,
"loss": 0.8421,
"step": 312
},
{
"epoch": 1.2710659898477157,
"grad_norm": 0.5338534116249731,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.8385,
"step": 313
},
{
"epoch": 1.2751269035532995,
"grad_norm": 0.5155584520437608,
"learning_rate": 3.713310605140055e-05,
"loss": 0.8352,
"step": 314
},
{
"epoch": 1.2791878172588833,
"grad_norm": 0.560262301579279,
"learning_rate": 3.710375590807233e-05,
"loss": 0.8286,
"step": 315
},
{
"epoch": 1.2832487309644671,
"grad_norm": 0.6767012487164431,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.8606,
"step": 316
},
{
"epoch": 1.2873096446700507,
"grad_norm": 0.6323284092160929,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.8223,
"step": 317
},
{
"epoch": 1.2913705583756345,
"grad_norm": 0.4760479512094556,
"learning_rate": 3.701487992136854e-05,
"loss": 0.802,
"step": 318
},
{
"epoch": 1.2954314720812183,
"grad_norm": 0.4406940891484498,
"learning_rate": 3.69849802022222e-05,
"loss": 0.8303,
"step": 319
},
{
"epoch": 1.299492385786802,
"grad_norm": 0.4176284457984641,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.8253,
"step": 320
},
{
"epoch": 1.3035532994923857,
"grad_norm": 0.6144711037846401,
"learning_rate": 3.692477062175289e-05,
"loss": 0.8234,
"step": 321
},
{
"epoch": 1.3076142131979696,
"grad_norm": 0.6489588668539846,
"learning_rate": 3.689446124534958e-05,
"loss": 0.8109,
"step": 322
},
{
"epoch": 1.3116751269035534,
"grad_norm": 0.4867211040381677,
"learning_rate": 3.686401580327799e-05,
"loss": 0.8459,
"step": 323
},
{
"epoch": 1.3157360406091372,
"grad_norm": 0.5157527312911335,
"learning_rate": 3.683343454074149e-05,
"loss": 0.8283,
"step": 324
},
{
"epoch": 1.3197969543147208,
"grad_norm": 0.43348577693844054,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.8102,
"step": 325
},
{
"epoch": 1.3238578680203046,
"grad_norm": 0.42506958213886786,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.8282,
"step": 326
},
{
"epoch": 1.3279187817258884,
"grad_norm": 0.5618717340049045,
"learning_rate": 3.674087829877297e-05,
"loss": 0.8293,
"step": 327
},
{
"epoch": 1.331979695431472,
"grad_norm": 0.5825627504697245,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.843,
"step": 328
},
{
"epoch": 1.3360406091370558,
"grad_norm": 0.6532594098269143,
"learning_rate": 3.667849957966501e-05,
"loss": 0.8447,
"step": 329
},
{
"epoch": 1.3401015228426396,
"grad_norm": 0.6288100030104357,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.8266,
"step": 330
},
{
"epoch": 1.3441624365482234,
"grad_norm": 0.5270083032515198,
"learning_rate": 3.661558355626795e-05,
"loss": 0.8523,
"step": 331
},
{
"epoch": 1.3482233502538072,
"grad_norm": 0.5100819666791492,
"learning_rate": 3.658392468818163e-05,
"loss": 0.8494,
"step": 332
},
{
"epoch": 1.3522842639593908,
"grad_norm": 0.5717607811074165,
"learning_rate": 3.655213225544574e-05,
"loss": 0.8437,
"step": 333
},
{
"epoch": 1.3563451776649746,
"grad_norm": 0.5472972569864112,
"learning_rate": 3.652020651411218e-05,
"loss": 0.87,
"step": 334
},
{
"epoch": 1.3604060913705585,
"grad_norm": 0.5932671337266103,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.8496,
"step": 335
},
{
"epoch": 1.364467005076142,
"grad_norm": 0.6243921876876665,
"learning_rate": 3.645595613522574e-05,
"loss": 0.8363,
"step": 336
},
{
"epoch": 1.3685279187817259,
"grad_norm": 0.6036716093983425,
"learning_rate": 3.642363201513657e-05,
"loss": 0.8276,
"step": 337
},
{
"epoch": 1.3725888324873097,
"grad_norm": 0.48087432063851737,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.8484,
"step": 338
},
{
"epoch": 1.3766497461928933,
"grad_norm": 0.4494954915301649,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.8069,
"step": 339
},
{
"epoch": 1.380710659898477,
"grad_norm": 0.4499646351199436,
"learning_rate": 3.632586705948318e-05,
"loss": 0.8316,
"step": 340
},
{
"epoch": 1.384771573604061,
"grad_norm": 0.4720861417103376,
"learning_rate": 3.629301541734311e-05,
"loss": 0.824,
"step": 341
},
{
"epoch": 1.3888324873096447,
"grad_norm": 0.5384355848996392,
"learning_rate": 3.626003255349676e-05,
"loss": 0.8609,
"step": 342
},
{
"epoch": 1.3928934010152285,
"grad_norm": 0.5743578246745845,
"learning_rate": 3.622691873358357e-05,
"loss": 0.8279,
"step": 343
},
{
"epoch": 1.396954314720812,
"grad_norm": 0.40898866517926397,
"learning_rate": 3.61936742242977e-05,
"loss": 0.8324,
"step": 344
},
{
"epoch": 1.401015228426396,
"grad_norm": 0.5232633950374341,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.8272,
"step": 345
},
{
"epoch": 1.4050761421319797,
"grad_norm": 0.4387075801608212,
"learning_rate": 3.612679420964516e-05,
"loss": 0.8586,
"step": 346
},
{
"epoch": 1.4091370558375633,
"grad_norm": 0.4111109623813353,
"learning_rate": 3.609315924292092e-05,
"loss": 0.8773,
"step": 347
},
{
"epoch": 1.4131979695431471,
"grad_norm": 0.419413559657943,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.8407,
"step": 348
},
{
"epoch": 1.417258883248731,
"grad_norm": 0.5860828029960355,
"learning_rate": 3.602550074513133e-05,
"loss": 0.8367,
"step": 349
},
{
"epoch": 1.4213197969543148,
"grad_norm": 0.6514078271211347,
"learning_rate": 3.599147775897822e-05,
"loss": 0.796,
"step": 350
},
{
"epoch": 1.4253807106598986,
"grad_norm": 0.721387803404927,
"learning_rate": 3.595732597966167e-05,
"loss": 0.8467,
"step": 351
},
{
"epoch": 1.4294416243654822,
"grad_norm": 0.835493892770028,
"learning_rate": 3.592304568223542e-05,
"loss": 0.8333,
"step": 352
},
{
"epoch": 1.433502538071066,
"grad_norm": 0.8003063387064198,
"learning_rate": 3.588863714278826e-05,
"loss": 0.8331,
"step": 353
},
{
"epoch": 1.4375634517766498,
"grad_norm": 0.7636733471270608,
"learning_rate": 3.585410063844186e-05,
"loss": 0.7958,
"step": 354
},
{
"epoch": 1.4416243654822334,
"grad_norm": 0.5519704825353865,
"learning_rate": 3.581943644734846e-05,
"loss": 0.8434,
"step": 355
},
{
"epoch": 1.4456852791878172,
"grad_norm": 0.4065090146578767,
"learning_rate": 3.578464484868869e-05,
"loss": 0.8227,
"step": 356
},
{
"epoch": 1.449746192893401,
"grad_norm": 0.7514051814118571,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.8232,
"step": 357
},
{
"epoch": 1.4538071065989848,
"grad_norm": 0.8293720347046424,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.8515,
"step": 358
},
{
"epoch": 1.4578680203045686,
"grad_norm": 0.49087240025620166,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.8314,
"step": 359
},
{
"epoch": 1.4619289340101522,
"grad_norm": 0.489734961840385,
"learning_rate": 3.564420999786543e-05,
"loss": 0.8503,
"step": 360
},
{
"epoch": 1.465989847715736,
"grad_norm": 0.9007290974785335,
"learning_rate": 3.560878558491842e-05,
"loss": 0.8377,
"step": 361
},
{
"epoch": 1.4700507614213199,
"grad_norm": 1.0330195191711173,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.8424,
"step": 362
},
{
"epoch": 1.4741116751269034,
"grad_norm": 0.7420925711864492,
"learning_rate": 3.553755991230039e-05,
"loss": 0.8244,
"step": 363
},
{
"epoch": 1.4781725888324873,
"grad_norm": 0.5819889550053469,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.8274,
"step": 364
},
{
"epoch": 1.482233502538071,
"grad_norm": 0.5377088402951758,
"learning_rate": 3.546583369120419e-05,
"loss": 0.8354,
"step": 365
},
{
"epoch": 1.4862944162436549,
"grad_norm": 0.6410653258797205,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.8679,
"step": 366
},
{
"epoch": 1.4903553299492387,
"grad_norm": 0.6341081249286764,
"learning_rate": 3.539360923231766e-05,
"loss": 0.8258,
"step": 367
},
{
"epoch": 1.4944162436548223,
"grad_norm": 0.5893440514843585,
"learning_rate": 3.535731089018394e-05,
"loss": 0.844,
"step": 368
},
{
"epoch": 1.498477157360406,
"grad_norm": 0.49173933113343044,
"learning_rate": 3.532088886237956e-05,
"loss": 0.8373,
"step": 369
},
{
"epoch": 1.50253807106599,
"grad_norm": 0.43745055806759486,
"learning_rate": 3.528434344224253e-05,
"loss": 0.8273,
"step": 370
},
{
"epoch": 1.5065989847715735,
"grad_norm": 0.4212287691935103,
"learning_rate": 3.524767492410464e-05,
"loss": 0.8364,
"step": 371
},
{
"epoch": 1.5106598984771573,
"grad_norm": 0.4390928272699896,
"learning_rate": 3.521088360328908e-05,
"loss": 0.8342,
"step": 372
},
{
"epoch": 1.5147208121827411,
"grad_norm": 0.411013391753576,
"learning_rate": 3.517396977610811e-05,
"loss": 0.8261,
"step": 373
},
{
"epoch": 1.5187817258883247,
"grad_norm": 0.49000931323135516,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.8442,
"step": 374
},
{
"epoch": 1.5228426395939088,
"grad_norm": 0.47799622425356125,
"learning_rate": 3.509977579282971e-05,
"loss": 0.8352,
"step": 375
},
{
"epoch": 1.5269035532994923,
"grad_norm": 0.4732295258064042,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.8607,
"step": 376
},
{
"epoch": 1.5309644670050762,
"grad_norm": 0.48009589490827587,
"learning_rate": 3.502509536445719e-05,
"loss": 0.8386,
"step": 377
},
{
"epoch": 1.53502538071066,
"grad_norm": 0.44051482756806337,
"learning_rate": 3.498757348458147e-05,
"loss": 0.8307,
"step": 378
},
{
"epoch": 1.5390862944162436,
"grad_norm": 0.5846582764740224,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.8346,
"step": 379
},
{
"epoch": 1.5431472081218274,
"grad_norm": 0.4983006044925057,
"learning_rate": 3.491216790442899e-05,
"loss": 0.8367,
"step": 380
},
{
"epoch": 1.5472081218274112,
"grad_norm": 0.5747074346507772,
"learning_rate": 3.487428481145839e-05,
"loss": 0.8557,
"step": 381
},
{
"epoch": 1.5512690355329948,
"grad_norm": 0.47143284734196855,
"learning_rate": 3.483628192304278e-05,
"loss": 0.8253,
"step": 382
},
{
"epoch": 1.5553299492385788,
"grad_norm": 0.4736548286621583,
"learning_rate": 3.479815954525219e-05,
"loss": 0.8382,
"step": 383
},
{
"epoch": 1.5593908629441624,
"grad_norm": 0.5561693427903793,
"learning_rate": 3.475991798511899e-05,
"loss": 0.8064,
"step": 384
},
{
"epoch": 1.5634517766497462,
"grad_norm": 0.5391802839736144,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.8641,
"step": 385
},
{
"epoch": 1.56751269035533,
"grad_norm": 0.7309692409651939,
"learning_rate": 3.468307855075128e-05,
"loss": 0.8143,
"step": 386
},
{
"epoch": 1.5715736040609136,
"grad_norm": 0.6600441036636492,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.8337,
"step": 387
},
{
"epoch": 1.5756345177664974,
"grad_norm": 0.5717141236160842,
"learning_rate": 3.460576609535163e-05,
"loss": 0.8494,
"step": 388
},
{
"epoch": 1.5796954314720812,
"grad_norm": 0.6238447468803623,
"learning_rate": 3.456693326250006e-05,
"loss": 0.8285,
"step": 389
},
{
"epoch": 1.5837563451776648,
"grad_norm": 0.591732320416858,
"learning_rate": 3.452798310957058e-05,
"loss": 0.8515,
"step": 390
},
{
"epoch": 1.5878172588832489,
"grad_norm": 0.5936459476711158,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.8202,
"step": 391
},
{
"epoch": 1.5918781725888325,
"grad_norm": 0.37671136708665454,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.8284,
"step": 392
},
{
"epoch": 1.5959390862944163,
"grad_norm": 0.43963529330306667,
"learning_rate": 3.441043187201574e-05,
"loss": 0.8435,
"step": 393
},
{
"epoch": 1.6,
"grad_norm": 0.3993677880177503,
"learning_rate": 3.437101558517728e-05,
"loss": 0.8414,
"step": 394
},
{
"epoch": 1.6040609137055837,
"grad_norm": 0.4185773945006266,
"learning_rate": 3.433148355615496e-05,
"loss": 0.854,
"step": 395
},
{
"epoch": 1.6081218274111675,
"grad_norm": 0.3563179666586982,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.8554,
"step": 396
},
{
"epoch": 1.6121827411167513,
"grad_norm": 0.42707544455112123,
"learning_rate": 3.425207354603043e-05,
"loss": 0.835,
"step": 397
},
{
"epoch": 1.616243654822335,
"grad_norm": 0.3505764756546602,
"learning_rate": 3.421219620448553e-05,
"loss": 0.8205,
"step": 398
},
{
"epoch": 1.620304568527919,
"grad_norm": 0.41681274804425994,
"learning_rate": 3.417220439986623e-05,
"loss": 0.8329,
"step": 399
},
{
"epoch": 1.6243654822335025,
"grad_norm": 0.3987237346071938,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.8576,
"step": 400
},
{
"epoch": 1.6284263959390863,
"grad_norm": 0.4342758632109407,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.8236,
"step": 401
},
{
"epoch": 1.6324873096446701,
"grad_norm": 0.40246207331357586,
"learning_rate": 3.405154543304065e-05,
"loss": 0.8402,
"step": 402
},
{
"epoch": 1.6365482233502537,
"grad_norm": 0.36968385873105397,
"learning_rate": 3.401109900618843e-05,
"loss": 0.8133,
"step": 403
},
{
"epoch": 1.6406091370558376,
"grad_norm": 0.470626091829263,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.822,
"step": 404
},
{
"epoch": 1.6446700507614214,
"grad_norm": 0.42274267758007883,
"learning_rate": 3.392986794874714e-05,
"loss": 0.8433,
"step": 405
},
{
"epoch": 1.648730964467005,
"grad_norm": 0.42581483050039326,
"learning_rate": 3.388908397238184e-05,
"loss": 0.843,
"step": 406
},
{
"epoch": 1.652791878172589,
"grad_norm": 0.3811781652219886,
"learning_rate": 3.384818813524362e-05,
"loss": 0.8234,
"step": 407
},
{
"epoch": 1.6568527918781726,
"grad_norm": 0.4378560649814688,
"learning_rate": 3.380718076670195e-05,
"loss": 0.8711,
"step": 408
},
{
"epoch": 1.6609137055837564,
"grad_norm": 0.4380933961236448,
"learning_rate": 3.376606219702454e-05,
"loss": 0.8444,
"step": 409
},
{
"epoch": 1.6649746192893402,
"grad_norm": 0.4784341986704259,
"learning_rate": 3.372483275737468e-05,
"loss": 0.818,
"step": 410
},
{
"epoch": 1.6690355329949238,
"grad_norm": 0.40186769373815034,
"learning_rate": 3.368349277980861e-05,
"loss": 0.8346,
"step": 411
},
{
"epoch": 1.6730964467005076,
"grad_norm": 0.47286683395878815,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.8399,
"step": 412
},
{
"epoch": 1.6771573604060914,
"grad_norm": 0.4215315216900389,
"learning_rate": 3.360048254360144e-05,
"loss": 0.829,
"step": 413
},
{
"epoch": 1.681218274111675,
"grad_norm": 0.4158266141298097,
"learning_rate": 3.355881295351336e-05,
"loss": 0.8126,
"step": 414
},
{
"epoch": 1.6852791878172588,
"grad_norm": 0.4529742870865966,
"learning_rate": 3.351703416260975e-05,
"loss": 0.8411,
"step": 415
},
{
"epoch": 1.6893401015228426,
"grad_norm": 0.49074538259207634,
"learning_rate": 3.347514650737126e-05,
"loss": 0.8421,
"step": 416
},
{
"epoch": 1.6934010152284262,
"grad_norm": 0.4579236760357225,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.8288,
"step": 417
},
{
"epoch": 1.6974619289340103,
"grad_norm": 0.4311924007579793,
"learning_rate": 3.339104595419334e-05,
"loss": 0.8647,
"step": 418
},
{
"epoch": 1.7015228426395939,
"grad_norm": 0.41180356661698275,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.8128,
"step": 419
},
{
"epoch": 1.7055837563451777,
"grad_norm": 0.45089835442765447,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.856,
"step": 420
},
{
"epoch": 1.7096446700507615,
"grad_norm": 0.4666342651590086,
"learning_rate": 3.326408710419996e-05,
"loss": 0.824,
"step": 421
},
{
"epoch": 1.713705583756345,
"grad_norm": 0.5582725502158407,
"learning_rate": 3.322155337795454e-05,
"loss": 0.8427,
"step": 422
},
{
"epoch": 1.7177664974619289,
"grad_norm": 0.5236657610137029,
"learning_rate": 3.317891316713587e-05,
"loss": 0.8519,
"step": 423
},
{
"epoch": 1.7218274111675127,
"grad_norm": 0.40545397300113567,
"learning_rate": 3.313616681516231e-05,
"loss": 0.8466,
"step": 424
},
{
"epoch": 1.7258883248730963,
"grad_norm": 0.5927784064084088,
"learning_rate": 3.309331466630713e-05,
"loss": 0.8392,
"step": 425
},
{
"epoch": 1.7299492385786803,
"grad_norm": 0.5733524087638878,
"learning_rate": 3.305035706569563e-05,
"loss": 0.8473,
"step": 426
},
{
"epoch": 1.734010152284264,
"grad_norm": 0.47082937583539364,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.8025,
"step": 427
},
{
"epoch": 1.7380710659898477,
"grad_norm": 0.5535127223714374,
"learning_rate": 3.296412689394864e-05,
"loss": 0.842,
"step": 428
},
{
"epoch": 1.7421319796954315,
"grad_norm": 0.5873732480320274,
"learning_rate": 3.292085501729909e-05,
"loss": 0.8306,
"step": 429
},
{
"epoch": 1.7461928934010151,
"grad_norm": 0.5252482534668429,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.8562,
"step": 430
},
{
"epoch": 1.750253807106599,
"grad_norm": 0.528621381905603,
"learning_rate": 3.283399942497381e-05,
"loss": 0.8182,
"step": 431
},
{
"epoch": 1.7543147208121828,
"grad_norm": 0.4430118134283268,
"learning_rate": 3.279041640882108e-05,
"loss": 0.8382,
"step": 432
},
{
"epoch": 1.7583756345177664,
"grad_norm": 0.3794230007735607,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.837,
"step": 433
},
{
"epoch": 1.7624365482233504,
"grad_norm": 0.40453993920595976,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.823,
"step": 434
},
{
"epoch": 1.766497461928934,
"grad_norm": 0.4338084789288484,
"learning_rate": 3.265905069502244e-05,
"loss": 0.8119,
"step": 435
},
{
"epoch": 1.7705583756345178,
"grad_norm": 0.4351239967530858,
"learning_rate": 3.261505774420016e-05,
"loss": 0.8422,
"step": 436
},
{
"epoch": 1.7746192893401016,
"grad_norm": 0.36111813045701585,
"learning_rate": 3.257096319343707e-05,
"loss": 0.8305,
"step": 437
},
{
"epoch": 1.7786802030456852,
"grad_norm": 0.462148299460509,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.8363,
"step": 438
},
{
"epoch": 1.782741116751269,
"grad_norm": 0.69550528185203,
"learning_rate": 3.248247071342966e-05,
"loss": 0.8135,
"step": 439
},
{
"epoch": 1.7868020304568528,
"grad_norm": 0.5376681744893077,
"learning_rate": 3.243807349689161e-05,
"loss": 0.8319,
"step": 440
},
{
"epoch": 1.7908629441624364,
"grad_norm": 0.5451580257880063,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.8399,
"step": 441
},
{
"epoch": 1.7949238578680204,
"grad_norm": 0.4163395496544942,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.8389,
"step": 442
},
{
"epoch": 1.798984771573604,
"grad_norm": 0.5208179848639264,
"learning_rate": 3.230428223438075e-05,
"loss": 0.8267,
"step": 443
},
{
"epoch": 1.8030456852791878,
"grad_norm": 0.41962299024851224,
"learning_rate": 3.225948647317459e-05,
"loss": 0.8178,
"step": 444
},
{
"epoch": 1.8071065989847717,
"grad_norm": 0.5468355109349753,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.8445,
"step": 445
},
{
"epoch": 1.8111675126903553,
"grad_norm": 0.39121359491390423,
"learning_rate": 3.216959910368034e-05,
"loss": 0.8527,
"step": 446
},
{
"epoch": 1.815228426395939,
"grad_norm": 0.5495738573910421,
"learning_rate": 3.212450821933277e-05,
"loss": 0.8016,
"step": 447
},
{
"epoch": 1.8192893401015229,
"grad_norm": 0.5259377037578279,
"learning_rate": 3.207931968586281e-05,
"loss": 0.8482,
"step": 448
},
{
"epoch": 1.8233502538071065,
"grad_norm": 0.42176938334536146,
"learning_rate": 3.203403386721272e-05,
"loss": 0.8455,
"step": 449
},
{
"epoch": 1.8274111675126905,
"grad_norm": 0.6240398919356389,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.8517,
"step": 450
},
{
"epoch": 1.831472081218274,
"grad_norm": 0.4540136297979541,
"learning_rate": 3.194317183405573e-05,
"loss": 0.8154,
"step": 451
},
{
"epoch": 1.835532994923858,
"grad_norm": 0.5657467074646221,
"learning_rate": 3.189759635133914e-05,
"loss": 0.8365,
"step": 452
},
{
"epoch": 1.8395939086294417,
"grad_norm": 0.5031354632398635,
"learning_rate": 3.185192504701718e-05,
"loss": 0.7963,
"step": 453
},
{
"epoch": 1.8436548223350253,
"grad_norm": 0.45275222086451605,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.8039,
"step": 454
},
{
"epoch": 1.8477157360406091,
"grad_norm": 0.44591852875018373,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.8236,
"step": 455
},
{
"epoch": 1.851776649746193,
"grad_norm": 0.3622969054011584,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.8287,
"step": 456
},
{
"epoch": 1.8558375634517765,
"grad_norm": 0.5602662584307379,
"learning_rate": 3.166828898179731e-05,
"loss": 0.8136,
"step": 457
},
{
"epoch": 1.8598984771573606,
"grad_norm": 0.6134443242294224,
"learning_rate": 3.162214410223536e-05,
"loss": 0.835,
"step": 458
},
{
"epoch": 1.8639593908629442,
"grad_norm": 0.5200614850287897,
"learning_rate": 3.157590561952257e-05,
"loss": 0.8339,
"step": 459
},
{
"epoch": 1.868020304568528,
"grad_norm": 0.4335544913469135,
"learning_rate": 3.152957390605732e-05,
"loss": 0.8398,
"step": 460
},
{
"epoch": 1.8720812182741118,
"grad_norm": 0.372840474288923,
"learning_rate": 3.148314933498886e-05,
"loss": 0.8472,
"step": 461
},
{
"epoch": 1.8761421319796954,
"grad_norm": 0.4706597738146551,
"learning_rate": 3.143663228021431e-05,
"loss": 0.8015,
"step": 462
},
{
"epoch": 1.8802030456852792,
"grad_norm": 0.5741709424555732,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.8454,
"step": 463
},
{
"epoch": 1.884263959390863,
"grad_norm": 0.39803678460941144,
"learning_rate": 3.134332221885661e-05,
"loss": 0.8348,
"step": 464
},
{
"epoch": 1.8883248730964466,
"grad_norm": 0.41271462321957536,
"learning_rate": 3.129652996377987e-05,
"loss": 0.8504,
"step": 465
},
{
"epoch": 1.8923857868020306,
"grad_norm": 0.5564905292094973,
"learning_rate": 3.12496467280038e-05,
"loss": 0.849,
"step": 466
},
{
"epoch": 1.8964467005076142,
"grad_norm": 0.4961563840640516,
"learning_rate": 3.120267288911952e-05,
"loss": 0.8492,
"step": 467
},
{
"epoch": 1.9005076142131978,
"grad_norm": 0.5531674725037493,
"learning_rate": 3.11556088254479e-05,
"loss": 0.827,
"step": 468
},
{
"epoch": 1.9045685279187818,
"grad_norm": 0.43399507010793253,
"learning_rate": 3.11084549160364e-05,
"loss": 0.8574,
"step": 469
},
{
"epoch": 1.9086294416243654,
"grad_norm": 0.4029498159888287,
"learning_rate": 3.106121154065615e-05,
"loss": 0.8298,
"step": 470
},
{
"epoch": 1.9126903553299492,
"grad_norm": 0.488464255946036,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.8627,
"step": 471
},
{
"epoch": 1.916751269035533,
"grad_norm": 0.40353283850652966,
"learning_rate": 3.096645791467348e-05,
"loss": 0.8124,
"step": 472
},
{
"epoch": 1.9208121827411166,
"grad_norm": 0.33219837444905637,
"learning_rate": 3.091894842720373e-05,
"loss": 0.8261,
"step": 473
},
{
"epoch": 1.9248730964467005,
"grad_norm": 0.4683100154613775,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.8602,
"step": 474
},
{
"epoch": 1.9289340101522843,
"grad_norm": 0.4454596129143337,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.8393,
"step": 475
},
{
"epoch": 1.9329949238578679,
"grad_norm": 0.44938194473030474,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.8192,
"step": 476
},
{
"epoch": 1.937055837563452,
"grad_norm": 0.3816472694149887,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.8487,
"step": 477
},
{
"epoch": 1.9411167512690355,
"grad_norm": 0.3778446539669071,
"learning_rate": 3.06800895716334e-05,
"loss": 0.8223,
"step": 478
},
{
"epoch": 1.9451776649746193,
"grad_norm": 0.4571312100432892,
"learning_rate": 3.063205821011029e-05,
"loss": 0.8272,
"step": 479
},
{
"epoch": 1.9492385786802031,
"grad_norm": 0.38238231314243787,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.818,
"step": 480
},
{
"epoch": 1.9532994923857867,
"grad_norm": 0.4292022249425022,
"learning_rate": 3.053573898718618e-05,
"loss": 0.8332,
"step": 481
},
{
"epoch": 1.9573604060913705,
"grad_norm": 0.4222897716962284,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.8419,
"step": 482
},
{
"epoch": 1.9614213197969543,
"grad_norm": 0.5095951074691307,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.837,
"step": 483
},
{
"epoch": 1.965482233502538,
"grad_norm": 0.3934653880091596,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.8305,
"step": 484
},
{
"epoch": 1.969543147208122,
"grad_norm": 0.4236604675793409,
"learning_rate": 3.034208541656255e-05,
"loss": 0.8445,
"step": 485
},
{
"epoch": 1.9736040609137055,
"grad_norm": 0.42933859663999613,
"learning_rate": 3.029346281309218e-05,
"loss": 0.8269,
"step": 486
},
{
"epoch": 1.9776649746192894,
"grad_norm": 0.4005271032147896,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.8401,
"step": 487
},
{
"epoch": 1.9817258883248732,
"grad_norm": 0.4568015114196979,
"learning_rate": 3.019596929201569e-05,
"loss": 0.8448,
"step": 488
},
{
"epoch": 1.9857868020304568,
"grad_norm": 0.4501516144300372,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.8276,
"step": 489
},
{
"epoch": 1.9898477157360406,
"grad_norm": 0.4562428309574424,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.8377,
"step": 490
},
{
"epoch": 1.9939086294416244,
"grad_norm": 0.3718375146687905,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.8304,
"step": 491
},
{
"epoch": 1.997969543147208,
"grad_norm": 0.3442612163540122,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.836,
"step": 492
},
{
"epoch": 2.002030456852792,
"grad_norm": 0.64891394208042,
"learning_rate": 2.995080534235264e-05,
"loss": 1.405,
"step": 493
},
{
"epoch": 2.0060913705583756,
"grad_norm": 0.8652463421629549,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.8131,
"step": 494
},
{
"epoch": 2.010152284263959,
"grad_norm": 0.9348779273473778,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.7784,
"step": 495
},
{
"epoch": 2.0142131979695432,
"grad_norm": 0.6438173289423972,
"learning_rate": 2.980274210295326e-05,
"loss": 0.8134,
"step": 496
},
{
"epoch": 2.018274111675127,
"grad_norm": 0.5277100925611112,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.7794,
"step": 497
},
{
"epoch": 2.022335025380711,
"grad_norm": 0.49448138330918556,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.7722,
"step": 498
},
{
"epoch": 2.0263959390862945,
"grad_norm": 0.43315896786743613,
"learning_rate": 2.965396831793362e-05,
"loss": 0.7732,
"step": 499
},
{
"epoch": 2.030456852791878,
"grad_norm": 0.6072775668922821,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.8156,
"step": 500
},
{
"epoch": 2.034517766497462,
"grad_norm": 0.48662773123353975,
"learning_rate": 2.955439637024526e-05,
"loss": 0.7899,
"step": 501
},
{
"epoch": 2.0385786802030457,
"grad_norm": 0.5358077374013118,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.8169,
"step": 502
},
{
"epoch": 2.0426395939086293,
"grad_norm": 0.4538550735736086,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.7856,
"step": 503
},
{
"epoch": 2.0467005076142133,
"grad_norm": 0.40460715149019766,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.7905,
"step": 504
},
{
"epoch": 2.050761421319797,
"grad_norm": 0.5198564106693823,
"learning_rate": 2.93543322968538e-05,
"loss": 0.7688,
"step": 505
},
{
"epoch": 2.054822335025381,
"grad_norm": 0.429772606498242,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.7712,
"step": 506
},
{
"epoch": 2.0588832487309645,
"grad_norm": 0.46558086367866325,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.772,
"step": 507
},
{
"epoch": 2.062944162436548,
"grad_norm": 0.5051467286549566,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.7949,
"step": 508
},
{
"epoch": 2.067005076142132,
"grad_norm": 0.3657620548936342,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.769,
"step": 509
},
{
"epoch": 2.0710659898477157,
"grad_norm": 0.43776290565561105,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.7881,
"step": 510
},
{
"epoch": 2.0751269035532993,
"grad_norm": 0.44747061775202096,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.7857,
"step": 511
},
{
"epoch": 2.0791878172588834,
"grad_norm": 0.3783287428915489,
"learning_rate": 2.900133526131588e-05,
"loss": 0.7828,
"step": 512
},
{
"epoch": 2.083248730964467,
"grad_norm": 0.35726606313433734,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.7903,
"step": 513
},
{
"epoch": 2.087309644670051,
"grad_norm": 0.37007618024489103,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.7934,
"step": 514
},
{
"epoch": 2.0913705583756346,
"grad_norm": 0.39161406647004077,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.7873,
"step": 515
},
{
"epoch": 2.095431472081218,
"grad_norm": 0.4115642656408347,
"learning_rate": 2.879801882631847e-05,
"loss": 0.7881,
"step": 516
},
{
"epoch": 2.099492385786802,
"grad_norm": 0.4671805677365532,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.7885,
"step": 517
},
{
"epoch": 2.103553299492386,
"grad_norm": 0.4904240725705561,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.7814,
"step": 518
},
{
"epoch": 2.1076142131979694,
"grad_norm": 0.4139929184686476,
"learning_rate": 2.864478605989494e-05,
"loss": 0.7931,
"step": 519
},
{
"epoch": 2.1116751269035534,
"grad_norm": 0.48786396999136983,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.7622,
"step": 520
},
{
"epoch": 2.115736040609137,
"grad_norm": 0.4422834030841654,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.7857,
"step": 521
},
{
"epoch": 2.1197969543147206,
"grad_norm": 0.4276585611401357,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.7993,
"step": 522
},
{
"epoch": 2.1238578680203046,
"grad_norm": 0.5074429805485987,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.7915,
"step": 523
},
{
"epoch": 2.127918781725888,
"grad_norm": 0.4057127068652121,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.7939,
"step": 524
},
{
"epoch": 2.1319796954314723,
"grad_norm": 0.4872289463867049,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.8059,
"step": 525
},
{
"epoch": 2.136040609137056,
"grad_norm": 0.4919609224055489,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.7835,
"step": 526
},
{
"epoch": 2.1401015228426394,
"grad_norm": 0.45746916285393524,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.792,
"step": 527
},
{
"epoch": 2.1441624365482235,
"grad_norm": 0.43800494914342597,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.7871,
"step": 528
},
{
"epoch": 2.148223350253807,
"grad_norm": 0.3779645547189469,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.7677,
"step": 529
},
{
"epoch": 2.152284263959391,
"grad_norm": 0.48737288282761904,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.7948,
"step": 530
},
{
"epoch": 2.1563451776649747,
"grad_norm": 0.40719384181057056,
"learning_rate": 2.802570062143278e-05,
"loss": 0.797,
"step": 531
},
{
"epoch": 2.1604060913705583,
"grad_norm": 0.48830468818476336,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.8137,
"step": 532
},
{
"epoch": 2.1644670050761423,
"grad_norm": 0.5520460121499337,
"learning_rate": 2.792159532078314e-05,
"loss": 0.7772,
"step": 533
},
{
"epoch": 2.168527918781726,
"grad_norm": 0.31765489806843067,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.7569,
"step": 534
},
{
"epoch": 2.1725888324873095,
"grad_norm": 0.4522380557543925,
"learning_rate": 2.781723482288857e-05,
"loss": 0.8046,
"step": 535
},
{
"epoch": 2.1766497461928935,
"grad_norm": 0.4599354074079185,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.7781,
"step": 536
},
{
"epoch": 2.180710659898477,
"grad_norm": 0.4129825199733175,
"learning_rate": 2.771262248976272e-05,
"loss": 0.7696,
"step": 537
},
{
"epoch": 2.1847715736040607,
"grad_norm": 0.3597972772000314,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.7945,
"step": 538
},
{
"epoch": 2.1888324873096447,
"grad_norm": 0.448372440211052,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.7656,
"step": 539
},
{
"epoch": 2.1928934010152283,
"grad_norm": 0.4629531913543579,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.7759,
"step": 540
},
{
"epoch": 2.1969543147208124,
"grad_norm": 0.41594750482355514,
"learning_rate": 2.7502655806328e-05,
"loss": 0.7739,
"step": 541
},
{
"epoch": 2.201015228426396,
"grad_norm": 0.6099230511312791,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.7973,
"step": 542
},
{
"epoch": 2.2050761421319796,
"grad_norm": 0.6270226494530453,
"learning_rate": 2.739730822017673e-05,
"loss": 0.8124,
"step": 543
},
{
"epoch": 2.2091370558375636,
"grad_norm": 0.48848635278516567,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.8037,
"step": 544
},
{
"epoch": 2.213197969543147,
"grad_norm": 0.3593375178240309,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.791,
"step": 545
},
{
"epoch": 2.2172588832487308,
"grad_norm": 0.47039650273919076,
"learning_rate": 2.723884107774236e-05,
"loss": 0.775,
"step": 546
},
{
"epoch": 2.221319796954315,
"grad_norm": 0.468698088818375,
"learning_rate": 2.718590152796239e-05,
"loss": 0.7908,
"step": 547
},
{
"epoch": 2.2253807106598984,
"grad_norm": 0.4314485372282206,
"learning_rate": 2.71329041039194e-05,
"loss": 0.8069,
"step": 548
},
{
"epoch": 2.2294416243654824,
"grad_norm": 0.3756302978314692,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.781,
"step": 549
},
{
"epoch": 2.233502538071066,
"grad_norm": 0.364665646824605,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.8081,
"step": 550
},
{
"epoch": 2.2375634517766496,
"grad_norm": 0.4373311003115894,
"learning_rate": 2.697356885686189e-05,
"loss": 0.7913,
"step": 551
},
{
"epoch": 2.2416243654822336,
"grad_norm": 0.43919524890125977,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.7769,
"step": 552
},
{
"epoch": 2.2456852791878172,
"grad_norm": 0.3751290179278852,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.7905,
"step": 553
},
{
"epoch": 2.249746192893401,
"grad_norm": 0.3460448698631867,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.7637,
"step": 554
},
{
"epoch": 2.253807106598985,
"grad_norm": 0.4676584148997351,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.7876,
"step": 555
},
{
"epoch": 2.2578680203045685,
"grad_norm": 0.44025520680191255,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.813,
"step": 556
},
{
"epoch": 2.261928934010152,
"grad_norm": 0.41621443270651415,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.7825,
"step": 557
},
{
"epoch": 2.265989847715736,
"grad_norm": 0.3701080537144901,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.7827,
"step": 558
},
{
"epoch": 2.2700507614213197,
"grad_norm": 0.42554138077738807,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.7648,
"step": 559
},
{
"epoch": 2.2741116751269037,
"grad_norm": 0.3010582955673964,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.8058,
"step": 560
},
{
"epoch": 2.2781725888324873,
"grad_norm": 0.35203361672088324,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.7888,
"step": 561
},
{
"epoch": 2.282233502538071,
"grad_norm": 0.4084088179119911,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.8021,
"step": 562
},
{
"epoch": 2.286294416243655,
"grad_norm": 0.3442625949102979,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.7909,
"step": 563
},
{
"epoch": 2.2903553299492385,
"grad_norm": 0.3398821977797976,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.7851,
"step": 564
},
{
"epoch": 2.2944162436548226,
"grad_norm": 0.412777572413902,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.7859,
"step": 565
},
{
"epoch": 2.298477157360406,
"grad_norm": 0.45102079570324355,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.7897,
"step": 566
},
{
"epoch": 2.3025380710659897,
"grad_norm": 0.3542313092630835,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.7714,
"step": 567
},
{
"epoch": 2.3065989847715738,
"grad_norm": 0.43307413741713047,
"learning_rate": 2.606146264569603e-05,
"loss": 0.7802,
"step": 568
},
{
"epoch": 2.3106598984771574,
"grad_norm": 0.38621936570142634,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.7946,
"step": 569
},
{
"epoch": 2.314720812182741,
"grad_norm": 0.36107902630036987,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.7846,
"step": 570
},
{
"epoch": 2.318781725888325,
"grad_norm": 0.39381941502999723,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.7975,
"step": 571
},
{
"epoch": 2.3228426395939086,
"grad_norm": 0.33146036036906096,
"learning_rate": 2.584472010606015e-05,
"loss": 0.7859,
"step": 572
},
{
"epoch": 2.326903553299492,
"grad_norm": 0.353608761683652,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.7729,
"step": 573
},
{
"epoch": 2.330964467005076,
"grad_norm": 0.3677979026424005,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.8157,
"step": 574
},
{
"epoch": 2.33502538071066,
"grad_norm": 0.3136935391293619,
"learning_rate": 2.568166741322048e-05,
"loss": 0.7688,
"step": 575
},
{
"epoch": 2.339086294416244,
"grad_norm": 0.5186534474774878,
"learning_rate": 2.56272244128509e-05,
"loss": 0.8109,
"step": 576
},
{
"epoch": 2.3431472081218274,
"grad_norm": 0.3422352191041233,
"learning_rate": 2.55727360915895e-05,
"loss": 0.7693,
"step": 577
},
{
"epoch": 2.347208121827411,
"grad_norm": 0.4021137219933842,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.815,
"step": 578
},
{
"epoch": 2.351269035532995,
"grad_norm": 0.37025113684125394,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.7471,
"step": 579
},
{
"epoch": 2.3553299492385786,
"grad_norm": 0.3161917054488647,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.7852,
"step": 580
},
{
"epoch": 2.3593908629441627,
"grad_norm": 0.3940494026071141,
"learning_rate": 2.535433837985559e-05,
"loss": 0.7647,
"step": 581
},
{
"epoch": 2.3634517766497463,
"grad_norm": 0.4049994808454731,
"learning_rate": 2.529963004393324e-05,
"loss": 0.782,
"step": 582
},
{
"epoch": 2.36751269035533,
"grad_norm": 0.34393786441543645,
"learning_rate": 2.524487902551908e-05,
"loss": 0.8022,
"step": 583
},
{
"epoch": 2.371573604060914,
"grad_norm": 0.3458739188615557,
"learning_rate": 2.519008576557029e-05,
"loss": 0.7867,
"step": 584
},
{
"epoch": 2.3756345177664975,
"grad_norm": 0.3512886583377272,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.7959,
"step": 585
},
{
"epoch": 2.379695431472081,
"grad_norm": 0.3166567008381363,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.7828,
"step": 586
},
{
"epoch": 2.383756345177665,
"grad_norm": 0.37827872090233694,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.7899,
"step": 587
},
{
"epoch": 2.3878172588832487,
"grad_norm": 0.38341693540910676,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.8023,
"step": 588
},
{
"epoch": 2.3918781725888323,
"grad_norm": 0.314871268529053,
"learning_rate": 2.491550129992345e-05,
"loss": 0.7816,
"step": 589
},
{
"epoch": 2.3959390862944163,
"grad_norm": 0.4337890281152915,
"learning_rate": 2.486046386967024e-05,
"loss": 0.777,
"step": 590
},
{
"epoch": 2.4,
"grad_norm": 0.4251034334619548,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.8113,
"step": 591
},
{
"epoch": 2.404060913705584,
"grad_norm": 0.2994041294404965,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.7872,
"step": 592
},
{
"epoch": 2.4081218274111675,
"grad_norm": 0.49840203513811554,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.8043,
"step": 593
},
{
"epoch": 2.412182741116751,
"grad_norm": 0.4059710203578824,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.7784,
"step": 594
},
{
"epoch": 2.416243654822335,
"grad_norm": 0.39640518858341955,
"learning_rate": 2.458469841202444e-05,
"loss": 0.7958,
"step": 595
},
{
"epoch": 2.4203045685279188,
"grad_norm": 0.3167486556837784,
"learning_rate": 2.452943276859226e-05,
"loss": 0.7967,
"step": 596
},
{
"epoch": 2.4243654822335023,
"grad_norm": 0.3878734460827972,
"learning_rate": 2.447413064573125e-05,
"loss": 0.7868,
"step": 597
},
{
"epoch": 2.4284263959390864,
"grad_norm": 0.2793482242945048,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.764,
"step": 598
},
{
"epoch": 2.43248730964467,
"grad_norm": 0.3565673803423223,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.7855,
"step": 599
},
{
"epoch": 2.436548223350254,
"grad_norm": 0.39096967168186086,
"learning_rate": 2.430800985597963e-05,
"loss": 0.7914,
"step": 600
},
{
"epoch": 2.4406091370558376,
"grad_norm": 0.3356770336056809,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.7753,
"step": 601
},
{
"epoch": 2.444670050761421,
"grad_norm": 0.3693679483682359,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.7773,
"step": 602
},
{
"epoch": 2.4487309644670052,
"grad_norm": 0.3352969801111382,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.7768,
"step": 603
},
{
"epoch": 2.452791878172589,
"grad_norm": 0.33402183067598984,
"learning_rate": 2.408603181104957e-05,
"loss": 0.7884,
"step": 604
},
{
"epoch": 2.4568527918781724,
"grad_norm": 0.3145526563253167,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.7789,
"step": 605
},
{
"epoch": 2.4609137055837564,
"grad_norm": 0.3255196063630698,
"learning_rate": 2.397484355008521e-05,
"loss": 0.7916,
"step": 606
},
{
"epoch": 2.46497461928934,
"grad_norm": 0.29204519991654526,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.7758,
"step": 607
},
{
"epoch": 2.4690355329949236,
"grad_norm": 0.2891030572484462,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.7629,
"step": 608
},
{
"epoch": 2.4730964467005077,
"grad_norm": 0.2923567174037668,
"learning_rate": 2.380782218329337e-05,
"loss": 0.7888,
"step": 609
},
{
"epoch": 2.4771573604060912,
"grad_norm": 0.3306525917922542,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.7978,
"step": 610
},
{
"epoch": 2.4812182741116753,
"grad_norm": 0.3773588068876117,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.7915,
"step": 611
},
{
"epoch": 2.485279187817259,
"grad_norm": 0.31683732276895005,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.7982,
"step": 612
},
{
"epoch": 2.4893401015228425,
"grad_norm": 0.38581970723619197,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.7964,
"step": 613
},
{
"epoch": 2.4934010152284265,
"grad_norm": 0.2866990737962306,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.7873,
"step": 614
},
{
"epoch": 2.49746192893401,
"grad_norm": 0.34632372526373356,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.7884,
"step": 615
},
{
"epoch": 2.501522842639594,
"grad_norm": 0.30665637390151196,
"learning_rate": 2.341705325992901e-05,
"loss": 0.7907,
"step": 616
},
{
"epoch": 2.5055837563451777,
"grad_norm": 0.39478841436867845,
"learning_rate": 2.336111544604222e-05,
"loss": 0.8006,
"step": 617
},
{
"epoch": 2.5096446700507613,
"grad_norm": 0.2727421043324723,
"learning_rate": 2.33051505621937e-05,
"loss": 0.8001,
"step": 618
},
{
"epoch": 2.5137055837563453,
"grad_norm": 0.3642318685099745,
"learning_rate": 2.324915905911693e-05,
"loss": 0.7765,
"step": 619
},
{
"epoch": 2.517766497461929,
"grad_norm": 0.2883606617502641,
"learning_rate": 2.319314138775977e-05,
"loss": 0.7885,
"step": 620
},
{
"epoch": 2.5218274111675125,
"grad_norm": 0.39038816390432624,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.7987,
"step": 621
},
{
"epoch": 2.5258883248730966,
"grad_norm": 0.3499604329365916,
"learning_rate": 2.308102934504593e-05,
"loss": 0.8071,
"step": 622
},
{
"epoch": 2.52994923857868,
"grad_norm": 0.3472031129926152,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.787,
"step": 623
},
{
"epoch": 2.5340101522842637,
"grad_norm": 0.41039609992932435,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.7771,
"step": 624
},
{
"epoch": 2.5380710659898478,
"grad_norm": 0.3257746967506483,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.7932,
"step": 625
},
{
"epoch": 2.5421319796954314,
"grad_norm": 0.3789015257069384,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.7851,
"step": 626
},
{
"epoch": 2.546192893401015,
"grad_norm": 0.43971258275737324,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.8072,
"step": 627
},
{
"epoch": 2.550253807106599,
"grad_norm": 0.3987210296702516,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.7721,
"step": 628
},
{
"epoch": 2.5543147208121826,
"grad_norm": 0.37554551707357203,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.7816,
"step": 629
},
{
"epoch": 2.5583756345177666,
"grad_norm": 0.3250075504151975,
"learning_rate": 2.26316247733093e-05,
"loss": 0.8002,
"step": 630
},
{
"epoch": 2.56243654822335,
"grad_norm": 0.2974967391492129,
"learning_rate": 2.257534907110328e-05,
"loss": 0.7643,
"step": 631
},
{
"epoch": 2.5664974619289342,
"grad_norm": 0.32417206916043345,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.7651,
"step": 632
},
{
"epoch": 2.570558375634518,
"grad_norm": 0.30435586977894385,
"learning_rate": 2.246273589557294e-05,
"loss": 0.7877,
"step": 633
},
{
"epoch": 2.5746192893401014,
"grad_norm": 0.33126346298012493,
"learning_rate": 2.240639932921966e-05,
"loss": 0.7651,
"step": 634
},
{
"epoch": 2.5786802030456855,
"grad_norm": 0.3796617549151834,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.7972,
"step": 635
},
{
"epoch": 2.582741116751269,
"grad_norm": 0.33986951853650826,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.7634,
"step": 636
},
{
"epoch": 2.5868020304568526,
"grad_norm": 0.32371264761925017,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.7908,
"step": 637
},
{
"epoch": 2.5908629441624367,
"grad_norm": 0.3148593795022237,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.7803,
"step": 638
},
{
"epoch": 2.5949238578680203,
"grad_norm": 0.34282558227552595,
"learning_rate": 2.212443486523819e-05,
"loss": 0.7741,
"step": 639
},
{
"epoch": 2.598984771573604,
"grad_norm": 0.30869686016726433,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.7962,
"step": 640
},
{
"epoch": 2.603045685279188,
"grad_norm": 0.28972449545870527,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.7698,
"step": 641
},
{
"epoch": 2.6071065989847715,
"grad_norm": 0.33798493260143075,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.7951,
"step": 642
},
{
"epoch": 2.611167512690355,
"grad_norm": 0.28133817024005686,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.7883,
"step": 643
},
{
"epoch": 2.615228426395939,
"grad_norm": 0.35740552787840285,
"learning_rate": 2.184204266048803e-05,
"loss": 0.7842,
"step": 644
},
{
"epoch": 2.6192893401015227,
"grad_norm": 0.3285179257035973,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.8005,
"step": 645
},
{
"epoch": 2.6233502538071067,
"grad_norm": 0.33793604823410167,
"learning_rate": 2.17289787446841e-05,
"loss": 0.7917,
"step": 646
},
{
"epoch": 2.6274111675126903,
"grad_norm": 0.3398528301474974,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.7862,
"step": 647
},
{
"epoch": 2.6314720812182744,
"grad_norm": 0.35028892036669446,
"learning_rate": 2.161585912916385e-05,
"loss": 0.8009,
"step": 648
},
{
"epoch": 2.635532994923858,
"grad_norm": 0.3323206026801166,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.7761,
"step": 649
},
{
"epoch": 2.6395939086294415,
"grad_norm": 0.36641983645762743,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.7743,
"step": 650
},
{
"epoch": 2.6436548223350256,
"grad_norm": 0.32408973495890764,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.8109,
"step": 651
},
{
"epoch": 2.647715736040609,
"grad_norm": 0.3261536162295097,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.7714,
"step": 652
},
{
"epoch": 2.6517766497461928,
"grad_norm": 0.36292631470694736,
"learning_rate": 2.133284032321232e-05,
"loss": 0.7974,
"step": 653
},
{
"epoch": 2.655837563451777,
"grad_norm": 0.37019989732408076,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.7822,
"step": 654
},
{
"epoch": 2.6598984771573604,
"grad_norm": 0.3487374975364812,
"learning_rate": 2.121955446741306e-05,
"loss": 0.8187,
"step": 655
},
{
"epoch": 2.663959390862944,
"grad_norm": 0.31130310561499763,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.7814,
"step": 656
},
{
"epoch": 2.668020304568528,
"grad_norm": 0.3432290413777575,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.7984,
"step": 657
},
{
"epoch": 2.6720812182741116,
"grad_norm": 0.2784292520484534,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.7724,
"step": 658
},
{
"epoch": 2.676142131979695,
"grad_norm": 0.29809379730160107,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.7777,
"step": 659
},
{
"epoch": 2.6802030456852792,
"grad_norm": 0.314447025149746,
"learning_rate": 2.093617592756158e-05,
"loss": 0.7991,
"step": 660
},
{
"epoch": 2.684263959390863,
"grad_norm": 0.28844677713981254,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.7672,
"step": 661
},
{
"epoch": 2.688324873096447,
"grad_norm": 0.33337354917144957,
"learning_rate": 2.082276853708978e-05,
"loss": 0.8011,
"step": 662
},
{
"epoch": 2.6923857868020304,
"grad_norm": 0.29090897293063595,
"learning_rate": 2.076605467380071e-05,
"loss": 0.7754,
"step": 663
},
{
"epoch": 2.6964467005076145,
"grad_norm": 0.32825311041463845,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.7856,
"step": 664
},
{
"epoch": 2.700507614213198,
"grad_norm": 0.3024193894680331,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.8121,
"step": 665
},
{
"epoch": 2.7045685279187817,
"grad_norm": 0.2784730254181954,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.7642,
"step": 666
},
{
"epoch": 2.7086294416243657,
"grad_norm": 0.3321101625599293,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.7761,
"step": 667
},
{
"epoch": 2.7126903553299493,
"grad_norm": 0.29321779886563354,
"learning_rate": 2.048240194884496e-05,
"loss": 0.7839,
"step": 668
},
{
"epoch": 2.716751269035533,
"grad_norm": 0.43605725460431627,
"learning_rate": 2.042565792045695e-05,
"loss": 0.7753,
"step": 669
},
{
"epoch": 2.720812182741117,
"grad_norm": 0.2849203692091733,
"learning_rate": 2.036891046387857e-05,
"loss": 0.7718,
"step": 670
},
{
"epoch": 2.7248730964467005,
"grad_norm": 0.3646339044233378,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.7885,
"step": 671
},
{
"epoch": 2.728934010152284,
"grad_norm": 0.4205278755093403,
"learning_rate": 2.025540709431948e-05,
"loss": 0.7924,
"step": 672
},
{
"epoch": 2.732994923857868,
"grad_norm": 0.30850312866758717,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.7789,
"step": 673
},
{
"epoch": 2.7370558375634517,
"grad_norm": 0.3915107267999157,
"learning_rate": 2.014189549672245e-05,
"loss": 0.7811,
"step": 674
},
{
"epoch": 2.7411167512690353,
"grad_norm": 0.3323152447853919,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.7981,
"step": 675
},
{
"epoch": 2.7451776649746193,
"grad_norm": 0.3081867744420649,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.7772,
"step": 676
},
{
"epoch": 2.749238578680203,
"grad_norm": 0.36836206498360724,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.784,
"step": 677
},
{
"epoch": 2.7532994923857865,
"grad_norm": 0.342196622234712,
"learning_rate": 1.991486224484123e-05,
"loss": 0.7785,
"step": 678
},
{
"epoch": 2.7573604060913706,
"grad_norm": 0.3367752226504114,
"learning_rate": 1.985810450327756e-05,
"loss": 0.7658,
"step": 679
},
{
"epoch": 2.761421319796954,
"grad_norm": 0.36320326684524723,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.7847,
"step": 680
},
{
"epoch": 2.765482233502538,
"grad_norm": 0.3194637387581025,
"learning_rate": 1.974459290568053e-05,
"loss": 0.8106,
"step": 681
},
{
"epoch": 2.769543147208122,
"grad_norm": 0.3892996166428429,
"learning_rate": 1.968783996385397e-05,
"loss": 0.7764,
"step": 682
},
{
"epoch": 2.773604060913706,
"grad_norm": 0.3057574427412112,
"learning_rate": 1.963108953612143e-05,
"loss": 0.7843,
"step": 683
},
{
"epoch": 2.7776649746192894,
"grad_norm": 0.4016761209641784,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.7932,
"step": 684
},
{
"epoch": 2.781725888324873,
"grad_norm": 0.322622055083241,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.7798,
"step": 685
},
{
"epoch": 2.785786802030457,
"grad_norm": 0.40234657520074923,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.8261,
"step": 686
},
{
"epoch": 2.7898477157360406,
"grad_norm": 0.4234600208046727,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.8103,
"step": 687
},
{
"epoch": 2.793908629441624,
"grad_norm": 0.35789927893759477,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.7491,
"step": 688
},
{
"epoch": 2.7979695431472082,
"grad_norm": 0.41315271083978455,
"learning_rate": 1.92906653591874e-05,
"loss": 0.7774,
"step": 689
},
{
"epoch": 2.802030456852792,
"grad_norm": 0.3684100960951132,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.8035,
"step": 690
},
{
"epoch": 2.8060913705583754,
"grad_norm": 0.4227462927920012,
"learning_rate": 1.917723146291022e-05,
"loss": 0.7674,
"step": 691
},
{
"epoch": 2.8101522842639595,
"grad_norm": 0.3084964299555269,
"learning_rate": 1.912052422608584e-05,
"loss": 0.766,
"step": 692
},
{
"epoch": 2.814213197969543,
"grad_norm": 0.4152962504576633,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.7962,
"step": 693
},
{
"epoch": 2.8182741116751266,
"grad_norm": 0.3241429092791355,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.7822,
"step": 694
},
{
"epoch": 2.8223350253807107,
"grad_norm": 0.40096597609423773,
"learning_rate": 1.895044684123479e-05,
"loss": 0.8005,
"step": 695
},
{
"epoch": 2.8263959390862943,
"grad_norm": 0.362653277090994,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.7943,
"step": 696
},
{
"epoch": 2.8304568527918783,
"grad_norm": 0.3525812853214671,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.7748,
"step": 697
},
{
"epoch": 2.834517766497462,
"grad_norm": 0.3247871605038389,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.7664,
"step": 698
},
{
"epoch": 2.838578680203046,
"grad_norm": 0.3288431037591601,
"learning_rate": 1.872379746550743e-05,
"loss": 0.7801,
"step": 699
},
{
"epoch": 2.8426395939086295,
"grad_norm": 0.3286665751725121,
"learning_rate": 1.866715967678769e-05,
"loss": 0.7976,
"step": 700
},
{
"epoch": 2.846700507614213,
"grad_norm": 0.3600684897254582,
"learning_rate": 1.861053262258067e-05,
"loss": 0.777,
"step": 701
},
{
"epoch": 2.850761421319797,
"grad_norm": 0.3239039887067652,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.7958,
"step": 702
},
{
"epoch": 2.8548223350253807,
"grad_norm": 0.2739302973678819,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.7884,
"step": 703
},
{
"epoch": 2.8588832487309643,
"grad_norm": 0.3126432071370836,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.7995,
"step": 704
},
{
"epoch": 2.8629441624365484,
"grad_norm": 0.3478690354942657,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.7757,
"step": 705
},
{
"epoch": 2.867005076142132,
"grad_norm": 0.2867621498391961,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.7805,
"step": 706
},
{
"epoch": 2.8710659898477155,
"grad_norm": 0.34410969176451933,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.7905,
"step": 707
},
{
"epoch": 2.8751269035532996,
"grad_norm": 0.35899648607635093,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.8038,
"step": 708
},
{
"epoch": 2.879187817258883,
"grad_norm": 0.33216399555339865,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.7686,
"step": 709
},
{
"epoch": 2.8832487309644668,
"grad_norm": 0.3215952334885353,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.7916,
"step": 710
},
{
"epoch": 2.887309644670051,
"grad_norm": 0.3211710199800581,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.7788,
"step": 711
},
{
"epoch": 2.8913705583756344,
"grad_norm": 0.30048135851688895,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.8097,
"step": 712
},
{
"epoch": 2.8954314720812184,
"grad_norm": 0.32880550154574145,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.786,
"step": 713
},
{
"epoch": 2.899492385786802,
"grad_norm": 0.3142321285198872,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.7918,
"step": 714
},
{
"epoch": 2.903553299492386,
"grad_norm": 0.35774097679425537,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.781,
"step": 715
},
{
"epoch": 2.9076142131979696,
"grad_norm": 0.28967080931043626,
"learning_rate": 1.776272483901182e-05,
"loss": 0.7835,
"step": 716
},
{
"epoch": 2.9116751269035532,
"grad_norm": 0.3274965021895058,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.7849,
"step": 717
},
{
"epoch": 2.9157360406091373,
"grad_norm": 0.4001142306005694,
"learning_rate": 1.764995661794301e-05,
"loss": 0.7799,
"step": 718
},
{
"epoch": 2.919796954314721,
"grad_norm": 0.2898541351185558,
"learning_rate": 1.759360067078035e-05,
"loss": 0.7777,
"step": 719
},
{
"epoch": 2.9238578680203045,
"grad_norm": 0.38831169491429907,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.787,
"step": 720
},
{
"epoch": 2.9279187817258885,
"grad_norm": 0.33206179946457476,
"learning_rate": 1.748094737261012e-05,
"loss": 0.8013,
"step": 721
},
{
"epoch": 2.931979695431472,
"grad_norm": 0.30537339104276,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.7945,
"step": 722
},
{
"epoch": 2.9360406091370557,
"grad_norm": 0.3208894660631289,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.7557,
"step": 723
},
{
"epoch": 2.9401015228426397,
"grad_norm": 0.29765219320445946,
"learning_rate": 1.731212071922883e-05,
"loss": 0.779,
"step": 724
},
{
"epoch": 2.9441624365482233,
"grad_norm": 0.29651569524467913,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.7647,
"step": 725
},
{
"epoch": 2.948223350253807,
"grad_norm": 0.29164399705633753,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.7764,
"step": 726
},
{
"epoch": 2.952284263959391,
"grad_norm": 0.26751152924186367,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.7993,
"step": 727
},
{
"epoch": 2.9563451776649745,
"grad_norm": 0.3278592842330587,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.7677,
"step": 728
},
{
"epoch": 2.960406091370558,
"grad_norm": 0.32065302487819675,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.807,
"step": 729
},
{
"epoch": 2.964467005076142,
"grad_norm": 0.3454848496680953,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.8048,
"step": 730
},
{
"epoch": 2.9685279187817257,
"grad_norm": 0.29033536100637847,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.802,
"step": 731
},
{
"epoch": 2.9725888324873098,
"grad_norm": 0.35081973599917765,
"learning_rate": 1.686290200071915e-05,
"loss": 0.7896,
"step": 732
},
{
"epoch": 2.9766497461928934,
"grad_norm": 0.3422142854536168,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.7884,
"step": 733
},
{
"epoch": 2.9807106598984774,
"grad_norm": 0.3039406476626592,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.7859,
"step": 734
},
{
"epoch": 2.984771573604061,
"grad_norm": 0.3045718685073243,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.7882,
"step": 735
},
{
"epoch": 2.9888324873096446,
"grad_norm": 0.2535105823290247,
"learning_rate": 1.663888455395778e-05,
"loss": 0.7918,
"step": 736
},
{
"epoch": 2.9928934010152286,
"grad_norm": 0.32120562574837785,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.8062,
"step": 737
},
{
"epoch": 2.996954314720812,
"grad_norm": 0.26905677090440744,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.7644,
"step": 738
},
{
"epoch": 3.001015228426396,
"grad_norm": 0.5272561843723065,
"learning_rate": 1.6471154124022818e-05,
"loss": 1.3683,
"step": 739
},
{
"epoch": 3.00507614213198,
"grad_norm": 0.3995125412114343,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.7224,
"step": 740
},
{
"epoch": 3.0091370558375634,
"grad_norm": 0.3938366508268558,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.7477,
"step": 741
},
{
"epoch": 3.013197969543147,
"grad_norm": 0.4473634225355428,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.7484,
"step": 742
},
{
"epoch": 3.017258883248731,
"grad_norm": 0.36658505587549567,
"learning_rate": 1.624791353909428e-05,
"loss": 0.7439,
"step": 743
},
{
"epoch": 3.0213197969543146,
"grad_norm": 0.44252399981787904,
"learning_rate": 1.619217781670663e-05,
"loss": 0.7382,
"step": 744
},
{
"epoch": 3.0253807106598987,
"grad_norm": 0.35827662892550266,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.7344,
"step": 745
},
{
"epoch": 3.0294416243654823,
"grad_norm": 0.4185665617633392,
"learning_rate": 1.608079882360134e-05,
"loss": 0.7137,
"step": 746
},
{
"epoch": 3.033502538071066,
"grad_norm": 0.4876804065724073,
"learning_rate": 1.60251564499148e-05,
"loss": 0.7441,
"step": 747
},
{
"epoch": 3.03756345177665,
"grad_norm": 0.3303488082709638,
"learning_rate": 1.596954608907176e-05,
"loss": 0.7335,
"step": 748
},
{
"epoch": 3.0416243654822335,
"grad_norm": 0.43963751974485077,
"learning_rate": 1.591396818895043e-05,
"loss": 0.7557,
"step": 749
},
{
"epoch": 3.045685279187817,
"grad_norm": 0.3220084387379337,
"learning_rate": 1.585842319716759e-05,
"loss": 0.7458,
"step": 750
},
{
"epoch": 3.049746192893401,
"grad_norm": 0.34193528234557186,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.7319,
"step": 751
},
{
"epoch": 3.0538071065989847,
"grad_norm": 0.32045802025445913,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.729,
"step": 752
},
{
"epoch": 3.0578680203045687,
"grad_norm": 0.3279893882544517,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.7341,
"step": 753
},
{
"epoch": 3.0619289340101523,
"grad_norm": 0.29691347373923505,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.722,
"step": 754
},
{
"epoch": 3.065989847715736,
"grad_norm": 0.3274278485365102,
"learning_rate": 1.558120751116291e-05,
"loss": 0.7485,
"step": 755
},
{
"epoch": 3.07005076142132,
"grad_norm": 0.2876091199522131,
"learning_rate": 1.552586935426876e-05,
"loss": 0.7401,
"step": 756
},
{
"epoch": 3.0741116751269035,
"grad_norm": 0.34117776094647645,
"learning_rate": 1.547056723140774e-05,
"loss": 0.7245,
"step": 757
},
{
"epoch": 3.078172588832487,
"grad_norm": 0.361802081945869,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.7416,
"step": 758
},
{
"epoch": 3.082233502538071,
"grad_norm": 0.3426236885673128,
"learning_rate": 1.536007286907411e-05,
"loss": 0.7599,
"step": 759
},
{
"epoch": 3.0862944162436547,
"grad_norm": 0.4852890605345928,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.7414,
"step": 760
},
{
"epoch": 3.090355329949239,
"grad_norm": 0.33237969736841255,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.738,
"step": 761
},
{
"epoch": 3.0944162436548224,
"grad_norm": 0.44137238279666374,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.761,
"step": 762
},
{
"epoch": 3.098477157360406,
"grad_norm": 0.3797869864963043,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.7252,
"step": 763
},
{
"epoch": 3.10253807106599,
"grad_norm": 0.2961740354584357,
"learning_rate": 1.508449870007656e-05,
"loss": 0.7288,
"step": 764
},
{
"epoch": 3.1065989847715736,
"grad_norm": 0.4446696963154877,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.7419,
"step": 765
},
{
"epoch": 3.110659898477157,
"grad_norm": 0.2723421851369137,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.7536,
"step": 766
},
{
"epoch": 3.114720812182741,
"grad_norm": 0.4168640413621046,
"learning_rate": 1.4919625713405e-05,
"loss": 0.7391,
"step": 767
},
{
"epoch": 3.118781725888325,
"grad_norm": 0.3304840241552262,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.742,
"step": 768
},
{
"epoch": 3.122842639593909,
"grad_norm": 0.4019941763245211,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.7248,
"step": 769
},
{
"epoch": 3.1269035532994924,
"grad_norm": 0.3087464883159267,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.7457,
"step": 770
},
{
"epoch": 3.130964467005076,
"grad_norm": 0.363648230874974,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.7346,
"step": 771
},
{
"epoch": 3.13502538071066,
"grad_norm": 0.27188450308844053,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.7568,
"step": 772
},
{
"epoch": 3.1390862944162436,
"grad_norm": 0.43845250311218126,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.7329,
"step": 773
},
{
"epoch": 3.1431472081218272,
"grad_norm": 0.2642384381691769,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.7649,
"step": 774
},
{
"epoch": 3.1472081218274113,
"grad_norm": 0.338293331024774,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.7195,
"step": 775
},
{
"epoch": 3.151269035532995,
"grad_norm": 0.27436565915905275,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.7492,
"step": 776
},
{
"epoch": 3.1553299492385785,
"grad_norm": 0.30259492454908654,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.7387,
"step": 777
},
{
"epoch": 3.1593908629441625,
"grad_norm": 0.2639400998864047,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.7583,
"step": 778
},
{
"epoch": 3.163451776649746,
"grad_norm": 0.26551641668032355,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.73,
"step": 779
},
{
"epoch": 3.16751269035533,
"grad_norm": 0.2944980787030092,
"learning_rate": 1.420958430225303e-05,
"loss": 0.758,
"step": 780
},
{
"epoch": 3.1715736040609137,
"grad_norm": 0.29781770495371135,
"learning_rate": 1.415527989393985e-05,
"loss": 0.7506,
"step": 781
},
{
"epoch": 3.1756345177664973,
"grad_norm": 0.34119082134701306,
"learning_rate": 1.410102255819891e-05,
"loss": 0.7481,
"step": 782
},
{
"epoch": 3.1796954314720813,
"grad_norm": 0.30851150885614853,
"learning_rate": 1.404681273201131e-05,
"loss": 0.7458,
"step": 783
},
{
"epoch": 3.183756345177665,
"grad_norm": 0.27361726931799646,
"learning_rate": 1.399265085197556e-05,
"loss": 0.7493,
"step": 784
},
{
"epoch": 3.187817258883249,
"grad_norm": 0.36306956336848056,
"learning_rate": 1.393853735430398e-05,
"loss": 0.7271,
"step": 785
},
{
"epoch": 3.1918781725888326,
"grad_norm": 0.3185922234214135,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.7277,
"step": 786
},
{
"epoch": 3.195939086294416,
"grad_norm": 0.28815719373360027,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.7519,
"step": 787
},
{
"epoch": 3.2,
"grad_norm": 0.30665509302604865,
"learning_rate": 1.377649151173163e-05,
"loss": 0.7478,
"step": 788
},
{
"epoch": 3.2040609137055838,
"grad_norm": 0.3239622910515329,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.7368,
"step": 789
},
{
"epoch": 3.2081218274111674,
"grad_norm": 0.2829991605182016,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.7311,
"step": 790
},
{
"epoch": 3.2121827411167514,
"grad_norm": 0.3925944810396778,
"learning_rate": 1.361489677627324e-05,
"loss": 0.7307,
"step": 791
},
{
"epoch": 3.216243654822335,
"grad_norm": 0.3117437800741885,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.7496,
"step": 792
},
{
"epoch": 3.2203045685279186,
"grad_norm": 0.31087275099758055,
"learning_rate": 1.350742335331241e-05,
"loss": 0.7609,
"step": 793
},
{
"epoch": 3.2243654822335026,
"grad_norm": 0.33879709967295524,
"learning_rate": 1.345376486102198e-05,
"loss": 0.7392,
"step": 794
},
{
"epoch": 3.228426395939086,
"grad_norm": 0.31237863299632385,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.7333,
"step": 795
},
{
"epoch": 3.2324873096446702,
"grad_norm": 0.3535882037053915,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.7455,
"step": 796
},
{
"epoch": 3.236548223350254,
"grad_norm": 0.27427926271935554,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.7412,
"step": 797
},
{
"epoch": 3.2406091370558374,
"grad_norm": 0.3151504447846241,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.7303,
"step": 798
},
{
"epoch": 3.2446700507614215,
"grad_norm": 0.27001938316373064,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.7475,
"step": 799
},
{
"epoch": 3.248730964467005,
"grad_norm": 0.2830417637242057,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.7481,
"step": 800
},
{
"epoch": 3.252791878172589,
"grad_norm": 0.3056894647613015,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.7522,
"step": 801
},
{
"epoch": 3.2568527918781727,
"grad_norm": 0.27733762682798163,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.7344,
"step": 802
},
{
"epoch": 3.2609137055837563,
"grad_norm": 0.32005797529676006,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.7362,
"step": 803
},
{
"epoch": 3.2649746192893403,
"grad_norm": 0.31805582361392576,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.7488,
"step": 804
},
{
"epoch": 3.269035532994924,
"grad_norm": 0.3143165155925378,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.738,
"step": 805
},
{
"epoch": 3.2730964467005075,
"grad_norm": 0.30874886747722174,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.7424,
"step": 806
},
{
"epoch": 3.2771573604060915,
"grad_norm": 0.3010319724195612,
"learning_rate": 1.276115892225764e-05,
"loss": 0.7333,
"step": 807
},
{
"epoch": 3.281218274111675,
"grad_norm": 0.26121271855233724,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.7286,
"step": 808
},
{
"epoch": 3.2852791878172587,
"grad_norm": 0.27455141861755006,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.745,
"step": 809
},
{
"epoch": 3.2893401015228427,
"grad_norm": 0.29415969513042417,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.77,
"step": 810
},
{
"epoch": 3.2934010152284263,
"grad_norm": 0.2911494860837187,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.7276,
"step": 811
},
{
"epoch": 3.29746192893401,
"grad_norm": 0.2599383228152168,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.7512,
"step": 812
},
{
"epoch": 3.301522842639594,
"grad_norm": 0.2791205334763667,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.7253,
"step": 813
},
{
"epoch": 3.3055837563451775,
"grad_norm": 0.27353854314388765,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.7418,
"step": 814
},
{
"epoch": 3.3096446700507616,
"grad_norm": 0.2510363756212501,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.7363,
"step": 815
},
{
"epoch": 3.313705583756345,
"grad_norm": 0.2728828695927957,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.7459,
"step": 816
},
{
"epoch": 3.3177664974619288,
"grad_norm": 0.2778220615086869,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.7182,
"step": 817
},
{
"epoch": 3.321827411167513,
"grad_norm": 0.2706334326092133,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.7524,
"step": 818
},
{
"epoch": 3.3258883248730964,
"grad_norm": 0.33292910716563195,
"learning_rate": 1.213055323844297e-05,
"loss": 0.7394,
"step": 819
},
{
"epoch": 3.3299492385786804,
"grad_norm": 0.2714101935547049,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.7614,
"step": 820
},
{
"epoch": 3.334010152284264,
"grad_norm": 0.3322722662060862,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.7233,
"step": 821
},
{
"epoch": 3.3380710659898476,
"grad_norm": 0.2540056373156356,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.7463,
"step": 822
},
{
"epoch": 3.3421319796954316,
"grad_norm": 0.28821523737476706,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.743,
"step": 823
},
{
"epoch": 3.3461928934010152,
"grad_norm": 0.2825372593715907,
"learning_rate": 1.187045262895488e-05,
"loss": 0.7439,
"step": 824
},
{
"epoch": 3.350253807106599,
"grad_norm": 0.26555647709243396,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.7476,
"step": 825
},
{
"epoch": 3.354314720812183,
"grad_norm": 0.24979621920316314,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.7732,
"step": 826
},
{
"epoch": 3.3583756345177664,
"grad_norm": 0.2496024750697965,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.7254,
"step": 827
},
{
"epoch": 3.36243654822335,
"grad_norm": 0.25182054657540387,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.7543,
"step": 828
},
{
"epoch": 3.366497461928934,
"grad_norm": 0.26570562705892076,
"learning_rate": 1.161198884952377e-05,
"loss": 0.7695,
"step": 829
},
{
"epoch": 3.3705583756345177,
"grad_norm": 0.23489487085423405,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.762,
"step": 830
},
{
"epoch": 3.3746192893401017,
"grad_norm": 0.2501080611325196,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.7365,
"step": 831
},
{
"epoch": 3.3786802030456853,
"grad_norm": 0.2556820948346217,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.7829,
"step": 832
},
{
"epoch": 3.382741116751269,
"grad_norm": 0.23456955137291757,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.7516,
"step": 833
},
{
"epoch": 3.386802030456853,
"grad_norm": 0.2483349006205051,
"learning_rate": 1.135521394010506e-05,
"loss": 0.7561,
"step": 834
},
{
"epoch": 3.3908629441624365,
"grad_norm": 0.2621579025203293,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.7437,
"step": 835
},
{
"epoch": 3.3949238578680205,
"grad_norm": 0.234798906520539,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.7451,
"step": 836
},
{
"epoch": 3.398984771573604,
"grad_norm": 0.2420752405876616,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.731,
"step": 837
},
{
"epoch": 3.4030456852791877,
"grad_norm": 0.24001912636883815,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.7305,
"step": 838
},
{
"epoch": 3.4071065989847718,
"grad_norm": 0.251533688715336,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.7391,
"step": 839
},
{
"epoch": 3.4111675126903553,
"grad_norm": 0.25744572484429745,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.7446,
"step": 840
},
{
"epoch": 3.415228426395939,
"grad_norm": 0.231718172223051,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.75,
"step": 841
},
{
"epoch": 3.419289340101523,
"grad_norm": 0.2576424502903035,
"learning_rate": 1.0948015846744e-05,
"loss": 0.7274,
"step": 842
},
{
"epoch": 3.4233502538071066,
"grad_norm": 0.2819264357076971,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.7612,
"step": 843
},
{
"epoch": 3.42741116751269,
"grad_norm": 0.23170687077949342,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.745,
"step": 844
},
{
"epoch": 3.431472081218274,
"grad_norm": 0.2889169574975605,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.7404,
"step": 845
},
{
"epoch": 3.4355329949238578,
"grad_norm": 0.24752001491585654,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.7402,
"step": 846
},
{
"epoch": 3.439593908629442,
"grad_norm": 0.31282878984182255,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.7414,
"step": 847
},
{
"epoch": 3.4436548223350254,
"grad_norm": 0.27827445872160866,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.7464,
"step": 848
},
{
"epoch": 3.447715736040609,
"grad_norm": 0.25959820978042525,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.7596,
"step": 849
},
{
"epoch": 3.451776649746193,
"grad_norm": 0.28009757453038014,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.7331,
"step": 850
},
{
"epoch": 3.4558375634517766,
"grad_norm": 0.26441163416028046,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.7417,
"step": 851
},
{
"epoch": 3.4598984771573607,
"grad_norm": 0.22478446900055882,
"learning_rate": 1.044560362975474e-05,
"loss": 0.7281,
"step": 852
},
{
"epoch": 3.4639593908629442,
"grad_norm": 0.26380059974827835,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.7268,
"step": 853
},
{
"epoch": 3.468020304568528,
"grad_norm": 0.2236590626384194,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.7242,
"step": 854
},
{
"epoch": 3.472081218274112,
"grad_norm": 0.225105392287092,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.707,
"step": 855
},
{
"epoch": 3.4761421319796955,
"grad_norm": 0.25097569472952364,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.7306,
"step": 856
},
{
"epoch": 3.480203045685279,
"grad_norm": 0.2403426115585771,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.749,
"step": 857
},
{
"epoch": 3.484263959390863,
"grad_norm": 0.2446777617970813,
"learning_rate": 1.014782400334433e-05,
"loss": 0.7279,
"step": 858
},
{
"epoch": 3.4883248730964467,
"grad_norm": 0.2556873660668158,
"learning_rate": 1.009846945771296e-05,
"loss": 0.7414,
"step": 859
},
{
"epoch": 3.4923857868020303,
"grad_norm": 0.2609147461889837,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.758,
"step": 860
},
{
"epoch": 3.4964467005076143,
"grad_norm": 0.2609577965875469,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.757,
"step": 861
},
{
"epoch": 3.500507614213198,
"grad_norm": 0.25279183390096327,
"learning_rate": 9.950885880977891e-06,
"loss": 0.7288,
"step": 862
},
{
"epoch": 3.5045685279187815,
"grad_norm": 0.27486794928859987,
"learning_rate": 9.901852696139382e-06,
"loss": 0.746,
"step": 863
},
{
"epoch": 3.5086294416243655,
"grad_norm": 0.25677685542325684,
"learning_rate": 9.852900840391027e-06,
"loss": 0.7539,
"step": 864
},
{
"epoch": 3.512690355329949,
"grad_norm": 1.521482654503753,
"learning_rate": 9.804030707984313e-06,
"loss": 0.7517,
"step": 865
},
{
"epoch": 3.516751269035533,
"grad_norm": 0.2575599981638263,
"learning_rate": 9.755242692512599e-06,
"loss": 0.7333,
"step": 866
},
{
"epoch": 3.5208121827411167,
"grad_norm": 0.2417441375571431,
"learning_rate": 9.70653718690782e-06,
"loss": 0.7534,
"step": 867
},
{
"epoch": 3.5248730964467008,
"grad_norm": 0.24620512184778567,
"learning_rate": 9.657914583437454e-06,
"loss": 0.7414,
"step": 868
},
{
"epoch": 3.5289340101522844,
"grad_norm": 0.24115043325920094,
"learning_rate": 9.609375273701246e-06,
"loss": 0.7362,
"step": 869
},
{
"epoch": 3.532994923857868,
"grad_norm": 0.2752632961058416,
"learning_rate": 9.560919648628133e-06,
"loss": 0.7655,
"step": 870
},
{
"epoch": 3.537055837563452,
"grad_norm": 0.23278488625594862,
"learning_rate": 9.512548098473047e-06,
"loss": 0.739,
"step": 871
},
{
"epoch": 3.5411167512690356,
"grad_norm": 0.25951889371596865,
"learning_rate": 9.464261012813825e-06,
"loss": 0.7352,
"step": 872
},
{
"epoch": 3.545177664974619,
"grad_norm": 0.2290229790246288,
"learning_rate": 9.416058780547987e-06,
"loss": 0.731,
"step": 873
},
{
"epoch": 3.549238578680203,
"grad_norm": 0.24659024883115266,
"learning_rate": 9.367941789889714e-06,
"loss": 0.7694,
"step": 874
},
{
"epoch": 3.553299492385787,
"grad_norm": 0.25787404882197246,
"learning_rate": 9.319910428366607e-06,
"loss": 0.7276,
"step": 875
},
{
"epoch": 3.5573604060913704,
"grad_norm": 0.23615790052667904,
"learning_rate": 9.271965082816667e-06,
"loss": 0.732,
"step": 876
},
{
"epoch": 3.5614213197969544,
"grad_norm": 0.25379973725955685,
"learning_rate": 9.224106139385111e-06,
"loss": 0.7439,
"step": 877
},
{
"epoch": 3.565482233502538,
"grad_norm": 0.24250996181672738,
"learning_rate": 9.176333983521291e-06,
"loss": 0.7336,
"step": 878
},
{
"epoch": 3.5695431472081216,
"grad_norm": 0.27360033387740423,
"learning_rate": 9.12864899997558e-06,
"loss": 0.7243,
"step": 879
},
{
"epoch": 3.5736040609137056,
"grad_norm": 0.2508165228700863,
"learning_rate": 9.08105157279628e-06,
"loss": 0.7357,
"step": 880
},
{
"epoch": 3.5776649746192892,
"grad_norm": 0.25227549130403126,
"learning_rate": 9.03354208532653e-06,
"loss": 0.753,
"step": 881
},
{
"epoch": 3.581725888324873,
"grad_norm": 0.24155626292668905,
"learning_rate": 8.986120920201205e-06,
"loss": 0.7538,
"step": 882
},
{
"epoch": 3.585786802030457,
"grad_norm": 0.27107708304424194,
"learning_rate": 8.938788459343852e-06,
"loss": 0.7375,
"step": 883
},
{
"epoch": 3.5898477157360404,
"grad_norm": 0.2511326734171602,
"learning_rate": 8.8915450839636e-06,
"loss": 0.7288,
"step": 884
},
{
"epoch": 3.5939086294416245,
"grad_norm": 0.23109212835916476,
"learning_rate": 8.844391174552116e-06,
"loss": 0.7616,
"step": 885
},
{
"epoch": 3.597969543147208,
"grad_norm": 0.26245808772545715,
"learning_rate": 8.797327110880479e-06,
"loss": 0.7554,
"step": 886
},
{
"epoch": 3.602030456852792,
"grad_norm": 0.26545054867805085,
"learning_rate": 8.750353271996206e-06,
"loss": 0.7429,
"step": 887
},
{
"epoch": 3.6060913705583757,
"grad_norm": 0.22464792750044918,
"learning_rate": 8.703470036220132e-06,
"loss": 0.7491,
"step": 888
},
{
"epoch": 3.6101522842639593,
"grad_norm": 0.2307376202774878,
"learning_rate": 8.656677781143394e-06,
"loss": 0.7523,
"step": 889
},
{
"epoch": 3.6142131979695433,
"grad_norm": 0.23944127080137503,
"learning_rate": 8.609976883624377e-06,
"loss": 0.7417,
"step": 890
},
{
"epoch": 3.618274111675127,
"grad_norm": 0.22982245136354854,
"learning_rate": 8.563367719785698e-06,
"loss": 0.7139,
"step": 891
},
{
"epoch": 3.6223350253807105,
"grad_norm": 0.22948869782898942,
"learning_rate": 8.516850665011138e-06,
"loss": 0.7405,
"step": 892
},
{
"epoch": 3.6263959390862945,
"grad_norm": 0.22700879030682156,
"learning_rate": 8.47042609394269e-06,
"loss": 0.7395,
"step": 893
},
{
"epoch": 3.630456852791878,
"grad_norm": 0.23403438807322638,
"learning_rate": 8.424094380477432e-06,
"loss": 0.7403,
"step": 894
},
{
"epoch": 3.6345177664974617,
"grad_norm": 0.2599099157052386,
"learning_rate": 8.37785589776465e-06,
"loss": 0.7594,
"step": 895
},
{
"epoch": 3.6385786802030458,
"grad_norm": 0.24402970348208128,
"learning_rate": 8.331711018202694e-06,
"loss": 0.7398,
"step": 896
},
{
"epoch": 3.6426395939086293,
"grad_norm": 0.2499812856959354,
"learning_rate": 8.285660113436104e-06,
"loss": 0.75,
"step": 897
},
{
"epoch": 3.646700507614213,
"grad_norm": 0.26496932894568204,
"learning_rate": 8.239703554352527e-06,
"loss": 0.7405,
"step": 898
},
{
"epoch": 3.650761421319797,
"grad_norm": 0.24588632730665527,
"learning_rate": 8.193841711079775e-06,
"loss": 0.749,
"step": 899
},
{
"epoch": 3.6548223350253806,
"grad_norm": 0.22798888412961407,
"learning_rate": 8.148074952982828e-06,
"loss": 0.7565,
"step": 900
},
{
"epoch": 3.6588832487309646,
"grad_norm": 0.26382430874419904,
"learning_rate": 8.102403648660859e-06,
"loss": 0.7253,
"step": 901
},
{
"epoch": 3.662944162436548,
"grad_norm": 0.2230223665177076,
"learning_rate": 8.056828165944282e-06,
"loss": 0.7324,
"step": 902
},
{
"epoch": 3.6670050761421322,
"grad_norm": 0.23200546768626257,
"learning_rate": 8.011348871891762e-06,
"loss": 0.7272,
"step": 903
},
{
"epoch": 3.671065989847716,
"grad_norm": 0.22992878540445752,
"learning_rate": 7.965966132787287e-06,
"loss": 0.7445,
"step": 904
},
{
"epoch": 3.6751269035532994,
"grad_norm": 0.23633450124422825,
"learning_rate": 7.920680314137189e-06,
"loss": 0.7217,
"step": 905
},
{
"epoch": 3.6791878172588834,
"grad_norm": 0.23273894533582937,
"learning_rate": 7.875491780667246e-06,
"loss": 0.7738,
"step": 906
},
{
"epoch": 3.683248730964467,
"grad_norm": 0.2322736235018593,
"learning_rate": 7.830400896319667e-06,
"loss": 0.7447,
"step": 907
},
{
"epoch": 3.6873096446700506,
"grad_norm": 0.26055207046750123,
"learning_rate": 7.785408024250259e-06,
"loss": 0.7399,
"step": 908
},
{
"epoch": 3.6913705583756347,
"grad_norm": 0.20152296096342115,
"learning_rate": 7.74051352682542e-06,
"loss": 0.7447,
"step": 909
},
{
"epoch": 3.6954314720812182,
"grad_norm": 0.24735431056615678,
"learning_rate": 7.695717765619257e-06,
"loss": 0.7419,
"step": 910
},
{
"epoch": 3.699492385786802,
"grad_norm": 0.25129461689703236,
"learning_rate": 7.651021101410673e-06,
"loss": 0.7503,
"step": 911
},
{
"epoch": 3.703553299492386,
"grad_norm": 0.2376847706323741,
"learning_rate": 7.606423894180464e-06,
"loss": 0.7393,
"step": 912
},
{
"epoch": 3.7076142131979695,
"grad_norm": 0.24136592348794328,
"learning_rate": 7.56192650310839e-06,
"loss": 0.7442,
"step": 913
},
{
"epoch": 3.711675126903553,
"grad_norm": 0.22503619709470735,
"learning_rate": 7.517529286570349e-06,
"loss": 0.7411,
"step": 914
},
{
"epoch": 3.715736040609137,
"grad_norm": 0.23227373547827485,
"learning_rate": 7.473232602135387e-06,
"loss": 0.7389,
"step": 915
},
{
"epoch": 3.7197969543147207,
"grad_norm": 0.2316689555657671,
"learning_rate": 7.429036806562935e-06,
"loss": 0.7553,
"step": 916
},
{
"epoch": 3.7238578680203047,
"grad_norm": 0.21575118586596123,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.7314,
"step": 917
},
{
"epoch": 3.7279187817258883,
"grad_norm": 0.25045013851788955,
"learning_rate": 7.340949304977567e-06,
"loss": 0.7544,
"step": 918
},
{
"epoch": 3.7319796954314723,
"grad_norm": 0.21393928325341693,
"learning_rate": 7.297058308409282e-06,
"loss": 0.7329,
"step": 919
},
{
"epoch": 3.736040609137056,
"grad_norm": 0.2372466928285262,
"learning_rate": 7.25326961958704e-06,
"loss": 0.7481,
"step": 920
},
{
"epoch": 3.7401015228426395,
"grad_norm": 0.2521778142033679,
"learning_rate": 7.209583591178921e-06,
"loss": 0.7484,
"step": 921
},
{
"epoch": 3.7441624365482236,
"grad_norm": 0.24188938258053216,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.7544,
"step": 922
},
{
"epoch": 3.748223350253807,
"grad_norm": 0.23184664300726707,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.7388,
"step": 923
},
{
"epoch": 3.7522842639593907,
"grad_norm": 0.2510462743381437,
"learning_rate": 7.079144982700909e-06,
"loss": 0.7282,
"step": 924
},
{
"epoch": 3.7563451776649748,
"grad_norm": 0.2980168710199434,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.7561,
"step": 925
},
{
"epoch": 3.7604060913705584,
"grad_norm": 0.24450263780934453,
"learning_rate": 6.99270564069757e-06,
"loss": 0.7294,
"step": 926
},
{
"epoch": 3.764467005076142,
"grad_norm": 0.24230554082720565,
"learning_rate": 6.949642934304375e-06,
"loss": 0.7252,
"step": 927
},
{
"epoch": 3.768527918781726,
"grad_norm": 0.22714615069749453,
"learning_rate": 6.906685333692871e-06,
"loss": 0.7576,
"step": 928
},
{
"epoch": 3.7725888324873096,
"grad_norm": 0.25676800727043986,
"learning_rate": 6.86383318483769e-06,
"loss": 0.7315,
"step": 929
},
{
"epoch": 3.776649746192893,
"grad_norm": 0.23243017026901552,
"learning_rate": 6.821086832864139e-06,
"loss": 0.7501,
"step": 930
},
{
"epoch": 3.780710659898477,
"grad_norm": 0.23387701007419862,
"learning_rate": 6.77844662204546e-06,
"loss": 0.7689,
"step": 931
},
{
"epoch": 3.784771573604061,
"grad_norm": 0.24240846133275248,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.7036,
"step": 932
},
{
"epoch": 3.788832487309645,
"grad_norm": 0.23625124060915847,
"learning_rate": 6.693485996688695e-06,
"loss": 0.7391,
"step": 933
},
{
"epoch": 3.7928934010152284,
"grad_norm": 0.2447453411043934,
"learning_rate": 6.651166266411801e-06,
"loss": 0.7406,
"step": 934
},
{
"epoch": 3.796954314720812,
"grad_norm": 0.25843158856773174,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.743,
"step": 935
},
{
"epoch": 3.801015228426396,
"grad_norm": 0.26033849403138104,
"learning_rate": 6.566849674844711e-06,
"loss": 0.7265,
"step": 936
},
{
"epoch": 3.8050761421319796,
"grad_norm": 0.2303788156060688,
"learning_rate": 6.524853492628747e-06,
"loss": 0.7392,
"step": 937
},
{
"epoch": 3.8091370558375637,
"grad_norm": 0.2250808956355806,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.7596,
"step": 938
},
{
"epoch": 3.8131979695431473,
"grad_norm": 0.22212031108153166,
"learning_rate": 6.441187046486648e-06,
"loss": 0.7201,
"step": 939
},
{
"epoch": 3.817258883248731,
"grad_norm": 0.265299499761227,
"learning_rate": 6.399517456398567e-06,
"loss": 0.7549,
"step": 940
},
{
"epoch": 3.821319796954315,
"grad_norm": 0.2078882009691513,
"learning_rate": 6.357957402727164e-06,
"loss": 0.7249,
"step": 941
},
{
"epoch": 3.8253807106598985,
"grad_norm": 0.21527356033807118,
"learning_rate": 6.316507220191395e-06,
"loss": 0.7403,
"step": 942
},
{
"epoch": 3.829441624365482,
"grad_norm": 0.24764348029490085,
"learning_rate": 6.275167242625331e-06,
"loss": 0.7422,
"step": 943
},
{
"epoch": 3.833502538071066,
"grad_norm": 0.22004098644932552,
"learning_rate": 6.233937802975471e-06,
"loss": 0.7391,
"step": 944
},
{
"epoch": 3.8375634517766497,
"grad_norm": 0.22749877075828798,
"learning_rate": 6.192819233298046e-06,
"loss": 0.7559,
"step": 945
},
{
"epoch": 3.8416243654822333,
"grad_norm": 0.22858436995937756,
"learning_rate": 6.151811864756383e-06,
"loss": 0.7231,
"step": 946
},
{
"epoch": 3.8456852791878173,
"grad_norm": 0.20151570292625065,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.743,
"step": 947
},
{
"epoch": 3.849746192893401,
"grad_norm": 0.2068047205090228,
"learning_rate": 6.070132051252868e-06,
"loss": 0.7314,
"step": 948
},
{
"epoch": 3.8538071065989845,
"grad_norm": 0.2910426669889258,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.7366,
"step": 949
},
{
"epoch": 3.8578680203045685,
"grad_norm": 0.21071386184491267,
"learning_rate": 5.988900993811575e-06,
"loss": 0.7436,
"step": 950
},
{
"epoch": 3.861928934010152,
"grad_norm": 0.22849085311042258,
"learning_rate": 5.948454566959363e-06,
"loss": 0.7555,
"step": 951
},
{
"epoch": 3.865989847715736,
"grad_norm": 0.2255294508140575,
"learning_rate": 5.908121309322328e-06,
"loss": 0.7508,
"step": 952
},
{
"epoch": 3.8700507614213198,
"grad_norm": 0.22562347706919952,
"learning_rate": 5.867901545738976e-06,
"loss": 0.7519,
"step": 953
},
{
"epoch": 3.874111675126904,
"grad_norm": 0.20458850987797497,
"learning_rate": 5.827795600133774e-06,
"loss": 0.7458,
"step": 954
},
{
"epoch": 3.8781725888324874,
"grad_norm": 0.22629134938810985,
"learning_rate": 5.787803795514466e-06,
"loss": 0.7218,
"step": 955
},
{
"epoch": 3.882233502538071,
"grad_norm": 0.23317151394054736,
"learning_rate": 5.747926453969576e-06,
"loss": 0.7422,
"step": 956
},
{
"epoch": 3.886294416243655,
"grad_norm": 0.22632132880013398,
"learning_rate": 5.708163896665708e-06,
"loss": 0.7186,
"step": 957
},
{
"epoch": 3.8903553299492386,
"grad_norm": 0.22208728288117852,
"learning_rate": 5.668516443845047e-06,
"loss": 0.7404,
"step": 958
},
{
"epoch": 3.894416243654822,
"grad_norm": 0.2160528938183103,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.7405,
"step": 959
},
{
"epoch": 3.8984771573604062,
"grad_norm": 0.21110092988902335,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.7039,
"step": 960
},
{
"epoch": 3.90253807106599,
"grad_norm": 0.21201570315614024,
"learning_rate": 5.550267900783019e-06,
"loss": 0.7549,
"step": 961
},
{
"epoch": 3.9065989847715734,
"grad_norm": 0.21311262784525606,
"learning_rate": 5.511084049737623e-06,
"loss": 0.7173,
"step": 962
},
{
"epoch": 3.9106598984771574,
"grad_norm": 0.21411990251735186,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.7435,
"step": 963
},
{
"epoch": 3.914720812182741,
"grad_norm": 0.21849107730150769,
"learning_rate": 5.433066737499948e-06,
"loss": 0.7541,
"step": 964
},
{
"epoch": 3.9187817258883246,
"grad_norm": 0.20294153131438714,
"learning_rate": 5.394233904648376e-06,
"loss": 0.7462,
"step": 965
},
{
"epoch": 3.9228426395939087,
"grad_norm": 0.21463427345226635,
"learning_rate": 5.355518704628997e-06,
"loss": 0.7239,
"step": 966
},
{
"epoch": 3.9269035532994923,
"grad_norm": 0.23671657537172908,
"learning_rate": 5.316921449248731e-06,
"loss": 0.7487,
"step": 967
},
{
"epoch": 3.9309644670050763,
"grad_norm": 0.23363893632078345,
"learning_rate": 5.278442449364538e-06,
"loss": 0.7354,
"step": 968
},
{
"epoch": 3.93502538071066,
"grad_norm": 0.19520580298691356,
"learning_rate": 5.240082014881016e-06,
"loss": 0.7448,
"step": 969
},
{
"epoch": 3.939086294416244,
"grad_norm": 0.19230981141514814,
"learning_rate": 5.201840454747822e-06,
"loss": 0.748,
"step": 970
},
{
"epoch": 3.9431472081218275,
"grad_norm": 0.20023436700406883,
"learning_rate": 5.163718076957223e-06,
"loss": 0.7116,
"step": 971
},
{
"epoch": 3.947208121827411,
"grad_norm": 0.21296565270842732,
"learning_rate": 5.125715188541609e-06,
"loss": 0.7229,
"step": 972
},
{
"epoch": 3.951269035532995,
"grad_norm": 0.22662415287060791,
"learning_rate": 5.087832095571021e-06,
"loss": 0.7368,
"step": 973
},
{
"epoch": 3.9553299492385787,
"grad_norm": 0.22248967061653518,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.7652,
"step": 974
},
{
"epoch": 3.9593908629441623,
"grad_norm": 0.2099808059460677,
"learning_rate": 5.01242651541854e-06,
"loss": 0.7357,
"step": 975
},
{
"epoch": 3.9634517766497463,
"grad_norm": 0.20610965580442173,
"learning_rate": 4.974904635542815e-06,
"loss": 0.7547,
"step": 976
},
{
"epoch": 3.96751269035533,
"grad_norm": 0.21603019923085653,
"learning_rate": 4.937503765719582e-06,
"loss": 0.7217,
"step": 977
},
{
"epoch": 3.9715736040609135,
"grad_norm": 0.23432127437655537,
"learning_rate": 4.900224207170299e-06,
"loss": 0.7483,
"step": 978
},
{
"epoch": 3.9756345177664976,
"grad_norm": 0.2579018922203181,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.7227,
"step": 979
},
{
"epoch": 3.979695431472081,
"grad_norm": 0.2424333544891702,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.7498,
"step": 980
},
{
"epoch": 3.9837563451776647,
"grad_norm": 0.21306614085277784,
"learning_rate": 4.789116396710924e-06,
"loss": 0.7365,
"step": 981
},
{
"epoch": 3.987817258883249,
"grad_norm": 0.21639303553147837,
"learning_rate": 4.752325075895368e-06,
"loss": 0.75,
"step": 982
},
{
"epoch": 3.9918781725888324,
"grad_norm": 0.2112003757142461,
"learning_rate": 4.715656557757473e-06,
"loss": 0.7479,
"step": 983
},
{
"epoch": 3.9959390862944164,
"grad_norm": 0.20253527571826366,
"learning_rate": 4.679111137620442e-06,
"loss": 0.731,
"step": 984
},
{
"epoch": 4.0,
"grad_norm": 0.41521391532492863,
"learning_rate": 4.6426891098160585e-06,
"loss": 1.2917,
"step": 985
},
{
"epoch": 4.004060913705584,
"grad_norm": 0.2850560941440662,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.7141,
"step": 986
},
{
"epoch": 4.008121827411167,
"grad_norm": 0.2888542907877441,
"learning_rate": 4.570216403561141e-06,
"loss": 0.7148,
"step": 987
},
{
"epoch": 4.012182741116751,
"grad_norm": 0.27491747442561987,
"learning_rate": 4.534166308795815e-06,
"loss": 0.7344,
"step": 988
},
{
"epoch": 4.016243654822335,
"grad_norm": 0.28179527762540824,
"learning_rate": 4.498240773728859e-06,
"loss": 0.7081,
"step": 989
},
{
"epoch": 4.020304568527918,
"grad_norm": 0.2518788378654235,
"learning_rate": 4.462440087699609e-06,
"loss": 0.7259,
"step": 990
},
{
"epoch": 4.024365482233502,
"grad_norm": 0.2942921376143902,
"learning_rate": 4.426764539041861e-06,
"loss": 0.7102,
"step": 991
},
{
"epoch": 4.0284263959390865,
"grad_norm": 0.2857739782152199,
"learning_rate": 4.391214415081582e-06,
"loss": 0.7199,
"step": 992
},
{
"epoch": 4.03248730964467,
"grad_norm": 0.24117619047634056,
"learning_rate": 4.355790002134579e-06,
"loss": 0.704,
"step": 993
},
{
"epoch": 4.036548223350254,
"grad_norm": 0.24701796794209915,
"learning_rate": 4.320491585504207e-06,
"loss": 0.7082,
"step": 994
},
{
"epoch": 4.040609137055838,
"grad_norm": 0.2747447351801685,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.7164,
"step": 995
},
{
"epoch": 4.044670050761422,
"grad_norm": 0.2557174468414056,
"learning_rate": 4.250273877330691e-06,
"loss": 0.7192,
"step": 996
},
{
"epoch": 4.048730964467005,
"grad_norm": 0.26971791828027597,
"learning_rate": 4.215355151311313e-06,
"loss": 0.7046,
"step": 997
},
{
"epoch": 4.052791878172589,
"grad_norm": 0.24428970205258202,
"learning_rate": 4.180563552651542e-06,
"loss": 0.6962,
"step": 998
},
{
"epoch": 4.056852791878173,
"grad_norm": 0.2666870737798325,
"learning_rate": 4.145899361558147e-06,
"loss": 0.7145,
"step": 999
},
{
"epoch": 4.060913705583756,
"grad_norm": 0.28527503856519576,
"learning_rate": 4.111362857211738e-06,
"loss": 0.7295,
"step": 1000
},
{
"epoch": 4.06497461928934,
"grad_norm": 0.24348351825634046,
"learning_rate": 4.076954317764592e-06,
"loss": 0.7155,
"step": 1001
},
{
"epoch": 4.069035532994924,
"grad_norm": 0.21531748782793372,
"learning_rate": 4.042674020338335e-06,
"loss": 0.7031,
"step": 1002
},
{
"epoch": 4.073096446700507,
"grad_norm": 0.25618200384083717,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.7012,
"step": 1003
},
{
"epoch": 4.077157360406091,
"grad_norm": 0.2817513510872296,
"learning_rate": 3.974499254868674e-06,
"loss": 0.7155,
"step": 1004
},
{
"epoch": 4.081218274111675,
"grad_norm": 0.22361220658714553,
"learning_rate": 3.940605335895451e-06,
"loss": 0.7107,
"step": 1005
},
{
"epoch": 4.0852791878172585,
"grad_norm": 0.21638444670819842,
"learning_rate": 3.90684075707908e-06,
"loss": 0.7024,
"step": 1006
},
{
"epoch": 4.0893401015228426,
"grad_norm": 0.25076449405014356,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.727,
"step": 1007
},
{
"epoch": 4.093401015228427,
"grad_norm": 0.22795220989908724,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.701,
"step": 1008
},
{
"epoch": 4.09746192893401,
"grad_norm": 0.21653887558824297,
"learning_rate": 3.806325775702304e-06,
"loss": 0.709,
"step": 1009
},
{
"epoch": 4.101522842639594,
"grad_norm": 0.23363907056199035,
"learning_rate": 3.773081266416434e-06,
"loss": 0.7121,
"step": 1010
},
{
"epoch": 4.105583756345178,
"grad_norm": 0.2484094886279314,
"learning_rate": 3.739967446503245e-06,
"loss": 0.6961,
"step": 1011
},
{
"epoch": 4.109644670050762,
"grad_norm": 0.2129494704600493,
"learning_rate": 3.706984582656894e-06,
"loss": 0.7008,
"step": 1012
},
{
"epoch": 4.113705583756345,
"grad_norm": 0.21217402119089102,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.7095,
"step": 1013
},
{
"epoch": 4.117766497461929,
"grad_norm": 0.2447940648563814,
"learning_rate": 3.641412784665648e-06,
"loss": 0.6999,
"step": 1014
},
{
"epoch": 4.121827411167513,
"grad_norm": 0.25104323364533415,
"learning_rate": 3.608824378627005e-06,
"loss": 0.7023,
"step": 1015
},
{
"epoch": 4.125888324873096,
"grad_norm": 0.22674342612180473,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.7353,
"step": 1016
},
{
"epoch": 4.12994923857868,
"grad_norm": 0.21052184913367925,
"learning_rate": 3.544043864774269e-06,
"loss": 0.7086,
"step": 1017
},
{
"epoch": 4.134010152284264,
"grad_norm": 0.2143451385751272,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.7201,
"step": 1018
},
{
"epoch": 4.138071065989847,
"grad_norm": 0.22573926325317895,
"learning_rate": 3.479793485887819e-06,
"loss": 0.726,
"step": 1019
},
{
"epoch": 4.1421319796954315,
"grad_norm": 0.20850431478139123,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.7069,
"step": 1020
},
{
"epoch": 4.1461928934010155,
"grad_norm": 0.23320933559391355,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.7201,
"step": 1021
},
{
"epoch": 4.150253807106599,
"grad_norm": 0.24789716827428848,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.7112,
"step": 1022
},
{
"epoch": 4.154314720812183,
"grad_norm": 0.21755084398521404,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.7043,
"step": 1023
},
{
"epoch": 4.158375634517767,
"grad_norm": 0.19663474021805777,
"learning_rate": 3.321500420335e-06,
"loss": 0.7349,
"step": 1024
},
{
"epoch": 4.16243654822335,
"grad_norm": 0.22931937425262247,
"learning_rate": 3.290243771741275e-06,
"loss": 0.709,
"step": 1025
},
{
"epoch": 4.166497461928934,
"grad_norm": 0.21500253239701286,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.7039,
"step": 1026
},
{
"epoch": 4.170558375634518,
"grad_norm": 0.23206521450620168,
"learning_rate": 3.228134459445149e-06,
"loss": 0.7176,
"step": 1027
},
{
"epoch": 4.174619289340102,
"grad_norm": 0.19993807293873533,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.7098,
"step": 1028
},
{
"epoch": 4.178680203045685,
"grad_norm": 0.20085679030024695,
"learning_rate": 3.166565459258513e-06,
"loss": 0.7057,
"step": 1029
},
{
"epoch": 4.182741116751269,
"grad_norm": 0.2021915048534329,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.7084,
"step": 1030
},
{
"epoch": 4.186802030456853,
"grad_norm": 0.19770791555426856,
"learning_rate": 3.105538754650419e-06,
"loss": 0.6914,
"step": 1031
},
{
"epoch": 4.190862944162436,
"grad_norm": 0.22757057414678655,
"learning_rate": 3.07522937824712e-06,
"loss": 0.7142,
"step": 1032
},
{
"epoch": 4.19492385786802,
"grad_norm": 0.22239309081799277,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.7113,
"step": 1033
},
{
"epoch": 4.198984771573604,
"grad_norm": 0.21856109868425377,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.6911,
"step": 1034
},
{
"epoch": 4.2030456852791875,
"grad_norm": 0.22064476833728447,
"learning_rate": 2.985120078631465e-06,
"loss": 0.6804,
"step": 1035
},
{
"epoch": 4.207106598984772,
"grad_norm": 0.188327941405046,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.6916,
"step": 1036
},
{
"epoch": 4.211167512690356,
"grad_norm": 0.21221941806893596,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.7101,
"step": 1037
},
{
"epoch": 4.215228426395939,
"grad_norm": 0.20174563385899463,
"learning_rate": 2.896244091927678e-06,
"loss": 0.704,
"step": 1038
},
{
"epoch": 4.219289340101523,
"grad_norm": 0.20613064430934083,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.7144,
"step": 1039
},
{
"epoch": 4.223350253807107,
"grad_norm": 0.20587529730483234,
"learning_rate": 2.837681792951994e-06,
"loss": 0.7253,
"step": 1040
},
{
"epoch": 4.22741116751269,
"grad_norm": 0.20815904060410292,
"learning_rate": 2.808607860255981e-06,
"loss": 0.7254,
"step": 1041
},
{
"epoch": 4.231472081218274,
"grad_norm": 0.22856615650465031,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.7024,
"step": 1042
},
{
"epoch": 4.235532994923858,
"grad_norm": 0.22051415732907617,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.7106,
"step": 1043
},
{
"epoch": 4.239593908629441,
"grad_norm": 0.2034090237994826,
"learning_rate": 2.722217735873718e-06,
"loss": 0.6982,
"step": 1044
},
{
"epoch": 4.243654822335025,
"grad_norm": 0.2005484451504613,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.6862,
"step": 1045
},
{
"epoch": 4.247715736040609,
"grad_norm": 0.2299513609659669,
"learning_rate": 2.665319697490205e-06,
"loss": 0.7129,
"step": 1046
},
{
"epoch": 4.251776649746193,
"grad_norm": 0.20473306780276238,
"learning_rate": 2.637079980714945e-06,
"loss": 0.7008,
"step": 1047
},
{
"epoch": 4.255837563451776,
"grad_norm": 0.1933200458488444,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.7165,
"step": 1048
},
{
"epoch": 4.2598984771573605,
"grad_norm": 0.21047418943172733,
"learning_rate": 2.581020289189571e-06,
"loss": 0.6824,
"step": 1049
},
{
"epoch": 4.2639593908629445,
"grad_norm": 0.20577843477703464,
"learning_rate": 2.553200765936501e-06,
"loss": 0.685,
"step": 1050
},
{
"epoch": 4.268020304568528,
"grad_norm": 0.21802997559316606,
"learning_rate": 2.525521756805962e-06,
"loss": 0.6859,
"step": 1051
},
{
"epoch": 4.272081218274112,
"grad_norm": 0.191364305721244,
"learning_rate": 2.497983484720885e-06,
"loss": 0.7019,
"step": 1052
},
{
"epoch": 4.276142131979696,
"grad_norm": 0.20644661986147728,
"learning_rate": 2.470586171470728e-06,
"loss": 0.7325,
"step": 1053
},
{
"epoch": 4.280203045685279,
"grad_norm": 0.25262109675549166,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.7011,
"step": 1054
},
{
"epoch": 4.284263959390863,
"grad_norm": 0.21359875850454293,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.7075,
"step": 1055
},
{
"epoch": 4.288324873096447,
"grad_norm": 0.21388043789542105,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.7074,
"step": 1056
},
{
"epoch": 4.29238578680203,
"grad_norm": 0.21352931611341033,
"learning_rate": 2.362410902836978e-06,
"loss": 0.728,
"step": 1057
},
{
"epoch": 4.296446700507614,
"grad_norm": 0.22847629191396904,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.7241,
"step": 1058
},
{
"epoch": 4.300507614213198,
"grad_norm": 0.23089562626153767,
"learning_rate": 2.309174704446411e-06,
"loss": 0.7199,
"step": 1059
},
{
"epoch": 4.304568527918782,
"grad_norm": 0.19824381991111542,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.6988,
"step": 1060
},
{
"epoch": 4.308629441624365,
"grad_norm": 0.18918781890298583,
"learning_rate": 2.256508422807855e-06,
"loss": 0.7021,
"step": 1061
},
{
"epoch": 4.312690355329949,
"grad_norm": 0.19534730528331284,
"learning_rate": 2.230389531696946e-06,
"loss": 0.7243,
"step": 1062
},
{
"epoch": 4.316751269035533,
"grad_norm": 0.20453014169090913,
"learning_rate": 2.204413754585857e-06,
"loss": 0.7016,
"step": 1063
},
{
"epoch": 4.320812182741117,
"grad_norm": 0.22813799544126379,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.7138,
"step": 1064
},
{
"epoch": 4.324873096446701,
"grad_norm": 0.21591015757115095,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.7185,
"step": 1065
},
{
"epoch": 4.328934010152285,
"grad_norm": 0.19051583174113043,
"learning_rate": 2.127347193531757e-06,
"loss": 0.6865,
"step": 1066
},
{
"epoch": 4.332994923857868,
"grad_norm": 0.197738767215128,
"learning_rate": 2.101945952921942e-06,
"loss": 0.7133,
"step": 1067
},
{
"epoch": 4.337055837563452,
"grad_norm": 0.19306980403817162,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.6853,
"step": 1068
},
{
"epoch": 4.341116751269036,
"grad_norm": 0.20701412900943456,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.7125,
"step": 1069
},
{
"epoch": 4.345177664974619,
"grad_norm": 0.2222600851334625,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.7138,
"step": 1070
},
{
"epoch": 4.349238578680203,
"grad_norm": 0.2038571945945684,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.6964,
"step": 1071
},
{
"epoch": 4.353299492385787,
"grad_norm": 0.20649965594326167,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.723,
"step": 1072
},
{
"epoch": 4.35736040609137,
"grad_norm": 0.18591726370957262,
"learning_rate": 1.95257270594031e-06,
"loss": 0.7062,
"step": 1073
},
{
"epoch": 4.361421319796954,
"grad_norm": 0.1932532745987076,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.6838,
"step": 1074
},
{
"epoch": 4.365482233502538,
"grad_norm": 0.19532573667941017,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.6918,
"step": 1075
},
{
"epoch": 4.369543147208121,
"grad_norm": 0.19948698389943428,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.7069,
"step": 1076
},
{
"epoch": 4.3736040609137055,
"grad_norm": 0.1922711433139182,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.693,
"step": 1077
},
{
"epoch": 4.3776649746192895,
"grad_norm": 0.19673311585872652,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.7116,
"step": 1078
},
{
"epoch": 4.381725888324873,
"grad_norm": 0.1939681776940388,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.7083,
"step": 1079
},
{
"epoch": 4.385786802030457,
"grad_norm": 0.1969351398842089,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.7206,
"step": 1080
},
{
"epoch": 4.389847715736041,
"grad_norm": 0.18849745282487337,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.6898,
"step": 1081
},
{
"epoch": 4.393908629441625,
"grad_norm": 0.19388382673699417,
"learning_rate": 1.738336950434061e-06,
"loss": 0.6968,
"step": 1082
},
{
"epoch": 4.397969543147208,
"grad_norm": 0.1794724896458433,
"learning_rate": 1.715265841375957e-06,
"loss": 0.6816,
"step": 1083
},
{
"epoch": 4.402030456852792,
"grad_norm": 0.18856207860651322,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.7153,
"step": 1084
},
{
"epoch": 4.406091370558376,
"grad_norm": 0.19229573269486613,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.7178,
"step": 1085
},
{
"epoch": 4.410152284263959,
"grad_norm": 0.20801158806456688,
"learning_rate": 1.646936827927441e-06,
"loss": 0.699,
"step": 1086
},
{
"epoch": 4.414213197969543,
"grad_norm": 0.19186933974991496,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.7039,
"step": 1087
},
{
"epoch": 4.418274111675127,
"grad_norm": 0.2003696855683119,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.7076,
"step": 1088
},
{
"epoch": 4.42233502538071,
"grad_norm": 0.1971319476563732,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.7219,
"step": 1089
},
{
"epoch": 4.426395939086294,
"grad_norm": 0.18854943099636698,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.7067,
"step": 1090
},
{
"epoch": 4.430456852791878,
"grad_norm": 0.18925292613909386,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.7102,
"step": 1091
},
{
"epoch": 4.4345177664974615,
"grad_norm": 0.1901411296550555,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.7283,
"step": 1092
},
{
"epoch": 4.438578680203046,
"grad_norm": 0.18594741297652848,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.7113,
"step": 1093
},
{
"epoch": 4.44263959390863,
"grad_norm": 0.20714302815473834,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.7056,
"step": 1094
},
{
"epoch": 4.446700507614214,
"grad_norm": 0.19738463771182943,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.72,
"step": 1095
},
{
"epoch": 4.450761421319797,
"grad_norm": 0.20932127089336555,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.722,
"step": 1096
},
{
"epoch": 4.454822335025381,
"grad_norm": 0.1905487757196375,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.7152,
"step": 1097
},
{
"epoch": 4.458883248730965,
"grad_norm": 0.1909907811050504,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.7109,
"step": 1098
},
{
"epoch": 4.462944162436548,
"grad_norm": 0.20086472481900297,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.7197,
"step": 1099
},
{
"epoch": 4.467005076142132,
"grad_norm": 0.1937168329569741,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.7232,
"step": 1100
},
{
"epoch": 4.471065989847716,
"grad_norm": 0.19471697744572478,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.721,
"step": 1101
},
{
"epoch": 4.475126903553299,
"grad_norm": 0.19065070910110418,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.681,
"step": 1102
},
{
"epoch": 4.479187817258883,
"grad_norm": 0.20268952854279956,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.7223,
"step": 1103
},
{
"epoch": 4.483248730964467,
"grad_norm": 0.1818363055816385,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.7103,
"step": 1104
},
{
"epoch": 4.4873096446700504,
"grad_norm": 0.21217213420997766,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.7177,
"step": 1105
},
{
"epoch": 4.4913705583756345,
"grad_norm": 0.21028157926866825,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.7216,
"step": 1106
},
{
"epoch": 4.4954314720812185,
"grad_norm": 0.2052122571680612,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.7238,
"step": 1107
},
{
"epoch": 4.499492385786802,
"grad_norm": 0.1852813097783252,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.7225,
"step": 1108
},
{
"epoch": 4.503553299492386,
"grad_norm": 0.1864694441265238,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.6991,
"step": 1109
},
{
"epoch": 4.50761421319797,
"grad_norm": 0.19718432078748988,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.7029,
"step": 1110
},
{
"epoch": 4.511675126903553,
"grad_norm": 0.193249898839053,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.7244,
"step": 1111
},
{
"epoch": 4.515736040609137,
"grad_norm": 0.19981859658046794,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.7002,
"step": 1112
},
{
"epoch": 4.519796954314721,
"grad_norm": 0.18900390090160304,
"learning_rate": 1.092402023876933e-06,
"loss": 0.7026,
"step": 1113
},
{
"epoch": 4.523857868020304,
"grad_norm": 0.18322687252270242,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.6955,
"step": 1114
},
{
"epoch": 4.527918781725888,
"grad_norm": 0.17903451553474023,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.7039,
"step": 1115
},
{
"epoch": 4.531979695431472,
"grad_norm": 0.18290996700820106,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.7013,
"step": 1116
},
{
"epoch": 4.536040609137056,
"grad_norm": 0.18118485292689418,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.7115,
"step": 1117
},
{
"epoch": 4.540101522842639,
"grad_norm": 0.17815244453066895,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.7073,
"step": 1118
},
{
"epoch": 4.544162436548223,
"grad_norm": 0.18243752025517682,
"learning_rate": 9.841383857091947e-07,
"loss": 0.7138,
"step": 1119
},
{
"epoch": 4.548223350253807,
"grad_norm": 0.19375951855218576,
"learning_rate": 9.666296380321616e-07,
"loss": 0.6995,
"step": 1120
},
{
"epoch": 4.552284263959391,
"grad_norm": 0.18311312272830696,
"learning_rate": 9.492741825034124e-07,
"loss": 0.7189,
"step": 1121
},
{
"epoch": 4.556345177664975,
"grad_norm": 0.21522556759336556,
"learning_rate": 9.320721589013892e-07,
"loss": 0.7016,
"step": 1122
},
{
"epoch": 4.560406091370559,
"grad_norm": 0.18525990329888184,
"learning_rate": 9.150237057688339e-07,
"loss": 0.7236,
"step": 1123
},
{
"epoch": 4.564467005076142,
"grad_norm": 0.19100985392712147,
"learning_rate": 8.981289604116328e-07,
"loss": 0.7319,
"step": 1124
},
{
"epoch": 4.568527918781726,
"grad_norm": 0.18582117579527035,
"learning_rate": 8.813880588977542e-07,
"loss": 0.6971,
"step": 1125
},
{
"epoch": 4.57258883248731,
"grad_norm": 0.17973387718573708,
"learning_rate": 8.648011360561126e-07,
"loss": 0.7077,
"step": 1126
},
{
"epoch": 4.576649746192894,
"grad_norm": 0.1815863256342077,
"learning_rate": 8.483683254755037e-07,
"loss": 0.7228,
"step": 1127
},
{
"epoch": 4.580710659898477,
"grad_norm": 0.18242793546140046,
"learning_rate": 8.320897595035227e-07,
"loss": 0.7065,
"step": 1128
},
{
"epoch": 4.584771573604061,
"grad_norm": 0.17631872571388305,
"learning_rate": 8.159655692455093e-07,
"loss": 0.6661,
"step": 1129
},
{
"epoch": 4.588832487309645,
"grad_norm": 0.18136267235846454,
"learning_rate": 7.999958845634648e-07,
"loss": 0.715,
"step": 1130
},
{
"epoch": 4.592893401015228,
"grad_norm": 0.19885035055577172,
"learning_rate": 7.841808340750478e-07,
"loss": 0.7122,
"step": 1131
},
{
"epoch": 4.596954314720812,
"grad_norm": 0.18868060008115312,
"learning_rate": 7.685205451524869e-07,
"loss": 0.7082,
"step": 1132
},
{
"epoch": 4.601015228426396,
"grad_norm": 0.19688723956130258,
"learning_rate": 7.530151439216027e-07,
"loss": 0.6919,
"step": 1133
},
{
"epoch": 4.6050761421319795,
"grad_norm": 0.1892381526419787,
"learning_rate": 7.376647552607675e-07,
"loss": 0.734,
"step": 1134
},
{
"epoch": 4.6091370558375635,
"grad_norm": 0.1857183939107732,
"learning_rate": 7.224695027998963e-07,
"loss": 0.6844,
"step": 1135
},
{
"epoch": 4.6131979695431475,
"grad_norm": 0.20198762895197114,
"learning_rate": 7.07429508919466e-07,
"loss": 0.7068,
"step": 1136
},
{
"epoch": 4.617258883248731,
"grad_norm": 0.21442069467122712,
"learning_rate": 6.925448947495206e-07,
"loss": 0.7041,
"step": 1137
},
{
"epoch": 4.621319796954315,
"grad_norm": 0.17271781215280485,
"learning_rate": 6.778157801686936e-07,
"loss": 0.6907,
"step": 1138
},
{
"epoch": 4.625380710659899,
"grad_norm": 0.19940047554083049,
"learning_rate": 6.632422838032515e-07,
"loss": 0.6926,
"step": 1139
},
{
"epoch": 4.629441624365482,
"grad_norm": 0.18201464122924066,
"learning_rate": 6.488245230261281e-07,
"loss": 0.7152,
"step": 1140
},
{
"epoch": 4.633502538071066,
"grad_norm": 0.19106350068613595,
"learning_rate": 6.345626139559868e-07,
"loss": 0.7128,
"step": 1141
},
{
"epoch": 4.63756345177665,
"grad_norm": 0.19575171775220232,
"learning_rate": 6.204566714562866e-07,
"loss": 0.7098,
"step": 1142
},
{
"epoch": 4.641624365482233,
"grad_norm": 0.1860021445236794,
"learning_rate": 6.06506809134344e-07,
"loss": 0.7195,
"step": 1143
},
{
"epoch": 4.645685279187817,
"grad_norm": 0.20165562305028725,
"learning_rate": 5.927131393404373e-07,
"loss": 0.7025,
"step": 1144
},
{
"epoch": 4.649746192893401,
"grad_norm": 0.19111839749815457,
"learning_rate": 5.790757731668817e-07,
"loss": 0.7139,
"step": 1145
},
{
"epoch": 4.653807106598984,
"grad_norm": 0.1749719770564701,
"learning_rate": 5.655948204471507e-07,
"loss": 0.6775,
"step": 1146
},
{
"epoch": 4.657868020304568,
"grad_norm": 0.19815546728124422,
"learning_rate": 5.522703897549875e-07,
"loss": 0.7394,
"step": 1147
},
{
"epoch": 4.661928934010152,
"grad_norm": 0.18961130066249993,
"learning_rate": 5.391025884035239e-07,
"loss": 0.7168,
"step": 1148
},
{
"epoch": 4.6659898477157356,
"grad_norm": 0.1893287221057379,
"learning_rate": 5.260915224444207e-07,
"loss": 0.7202,
"step": 1149
},
{
"epoch": 4.67005076142132,
"grad_norm": 0.19438061826481337,
"learning_rate": 5.132372966670129e-07,
"loss": 0.7209,
"step": 1150
},
{
"epoch": 4.674111675126904,
"grad_norm": 0.21047134904569653,
"learning_rate": 5.005400145974704e-07,
"loss": 0.7055,
"step": 1151
},
{
"epoch": 4.678172588832488,
"grad_norm": 0.2180569002551968,
"learning_rate": 4.879997784979562e-07,
"loss": 0.7063,
"step": 1152
},
{
"epoch": 4.682233502538071,
"grad_norm": 0.19582195435858835,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.711,
"step": 1153
},
{
"epoch": 4.686294416243655,
"grad_norm": 0.1892357703752406,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.7193,
"step": 1154
},
{
"epoch": 4.690355329949239,
"grad_norm": 0.18358188080866097,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.7025,
"step": 1155
},
{
"epoch": 4.694416243654822,
"grad_norm": 0.19976708201672577,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.7417,
"step": 1156
},
{
"epoch": 4.698477157360406,
"grad_norm": 0.1844356894032601,
"learning_rate": 4.27657778143431e-07,
"loss": 0.6891,
"step": 1157
},
{
"epoch": 4.70253807106599,
"grad_norm": 0.19397329457267418,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.6858,
"step": 1158
},
{
"epoch": 4.706598984771573,
"grad_norm": 0.20027813257836688,
"learning_rate": 4.046237371675177e-07,
"loss": 0.7183,
"step": 1159
},
{
"epoch": 4.710659898477157,
"grad_norm": 0.1835685309560609,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.7043,
"step": 1160
},
{
"epoch": 4.714720812182741,
"grad_norm": 0.17921532922278718,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.7116,
"step": 1161
},
{
"epoch": 4.718781725888325,
"grad_norm": 0.1762036906039039,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.6769,
"step": 1162
},
{
"epoch": 4.7228426395939085,
"grad_norm": 0.18438117973380222,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.7254,
"step": 1163
},
{
"epoch": 4.7269035532994925,
"grad_norm": 0.17140330163611783,
"learning_rate": 3.498020255137813e-07,
"loss": 0.7155,
"step": 1164
},
{
"epoch": 4.730964467005077,
"grad_norm": 0.18780148772107533,
"learning_rate": 3.393121163057811e-07,
"loss": 0.7045,
"step": 1165
},
{
"epoch": 4.73502538071066,
"grad_norm": 0.17916428013506577,
"learning_rate": 3.289805515751399e-07,
"loss": 0.7051,
"step": 1166
},
{
"epoch": 4.739086294416244,
"grad_norm": 0.1900846470364545,
"learning_rate": 3.188074145308573e-07,
"loss": 0.7081,
"step": 1167
},
{
"epoch": 4.743147208121828,
"grad_norm": 0.18398667783682163,
"learning_rate": 3.087927871059804e-07,
"loss": 0.7171,
"step": 1168
},
{
"epoch": 4.747208121827411,
"grad_norm": 0.1786797493285871,
"learning_rate": 2.989367499569418e-07,
"loss": 0.7235,
"step": 1169
},
{
"epoch": 4.751269035532995,
"grad_norm": 0.191264731060291,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.7131,
"step": 1170
},
{
"epoch": 4.755329949238579,
"grad_norm": 0.1860291435316892,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.6963,
"step": 1171
},
{
"epoch": 4.759390862944162,
"grad_norm": 0.19895794302931868,
"learning_rate": 2.703209675663887e-07,
"loss": 0.6977,
"step": 1172
},
{
"epoch": 4.763451776649746,
"grad_norm": 0.18506146006581123,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.6977,
"step": 1173
},
{
"epoch": 4.76751269035533,
"grad_norm": 0.17779966728452806,
"learning_rate": 2.520381518804471e-07,
"loss": 0.7038,
"step": 1174
},
{
"epoch": 4.771573604060913,
"grad_norm": 0.18181357309939802,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.7056,
"step": 1175
},
{
"epoch": 4.775634517766497,
"grad_norm": 0.18617462367358958,
"learning_rate": 2.343915243930317e-07,
"loss": 0.7099,
"step": 1176
},
{
"epoch": 4.779695431472081,
"grad_norm": 0.17800999134377843,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.7138,
"step": 1177
},
{
"epoch": 4.783756345177665,
"grad_norm": 0.17891939462393,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.7304,
"step": 1178
},
{
"epoch": 4.787817258883249,
"grad_norm": 0.1794217627811603,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.7185,
"step": 1179
},
{
"epoch": 4.791878172588833,
"grad_norm": 0.17562033309283775,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.7082,
"step": 1180
},
{
"epoch": 4.795939086294416,
"grad_norm": 0.17587695987018678,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.7243,
"step": 1181
},
{
"epoch": 4.8,
"grad_norm": 0.17228679354461118,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.707,
"step": 1182
},
{
"epoch": 4.804060913705584,
"grad_norm": 0.17940519247993464,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.7088,
"step": 1183
},
{
"epoch": 4.808121827411168,
"grad_norm": 0.1812927905758579,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.699,
"step": 1184
},
{
"epoch": 4.812182741116751,
"grad_norm": 0.1767516259431487,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.7138,
"step": 1185
},
{
"epoch": 4.816243654822335,
"grad_norm": 0.179376907713332,
"learning_rate": 1.557203888516745e-07,
"loss": 0.6996,
"step": 1186
},
{
"epoch": 4.820304568527919,
"grad_norm": 0.18081525033039156,
"learning_rate": 1.487313098987131e-07,
"loss": 0.7293,
"step": 1187
},
{
"epoch": 4.824365482233502,
"grad_norm": 0.18059810664281045,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.7186,
"step": 1188
},
{
"epoch": 4.828426395939086,
"grad_norm": 0.1738613382670985,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.7082,
"step": 1189
},
{
"epoch": 4.83248730964467,
"grad_norm": 1.1810657649736291,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.739,
"step": 1190
},
{
"epoch": 4.8365482233502535,
"grad_norm": 0.1794829688930156,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.7168,
"step": 1191
},
{
"epoch": 4.8406091370558375,
"grad_norm": 0.1738768213722039,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.7036,
"step": 1192
},
{
"epoch": 4.8446700507614215,
"grad_norm": 0.17883833126382923,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.6865,
"step": 1193
},
{
"epoch": 4.848730964467005,
"grad_norm": 0.18101989369032817,
"learning_rate": 1.042873700093061e-07,
"loss": 0.7242,
"step": 1194
},
{
"epoch": 4.852791878172589,
"grad_norm": 0.18415529671136324,
"learning_rate": 9.857879404556291e-08,
"loss": 0.7092,
"step": 1195
},
{
"epoch": 4.856852791878173,
"grad_norm": 0.18049812977584118,
"learning_rate": 9.303050139225722e-08,
"loss": 0.6998,
"step": 1196
},
{
"epoch": 4.860913705583757,
"grad_norm": 0.1785866092049721,
"learning_rate": 8.76425367345779e-08,
"loss": 0.6834,
"step": 1197
},
{
"epoch": 4.86497461928934,
"grad_norm": 0.1785392029303703,
"learning_rate": 8.241494346644897e-08,
"loss": 0.7068,
"step": 1198
},
{
"epoch": 4.869035532994924,
"grad_norm": 0.17428309998770658,
"learning_rate": 7.734776369019204e-08,
"loss": 0.6972,
"step": 1199
},
{
"epoch": 4.873096446700508,
"grad_norm": 0.17835898852518464,
"learning_rate": 7.244103821617332e-08,
"loss": 0.7108,
"step": 1200
},
{
"epoch": 4.877157360406091,
"grad_norm": 0.1805023163489205,
"learning_rate": 6.769480656248606e-08,
"loss": 0.692,
"step": 1201
},
{
"epoch": 4.881218274111675,
"grad_norm": 0.17977621695878188,
"learning_rate": 6.310910695462635e-08,
"loss": 0.6969,
"step": 1202
},
{
"epoch": 4.885279187817259,
"grad_norm": 0.17239587879760473,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.6992,
"step": 1203
},
{
"epoch": 4.889340101522842,
"grad_norm": 0.1847141639843027,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.707,
"step": 1204
},
{
"epoch": 4.893401015228426,
"grad_norm": 0.17306432330417978,
"learning_rate": 5.031556326567488e-08,
"loss": 0.7072,
"step": 1205
},
{
"epoch": 4.8974619289340104,
"grad_norm": 0.17246899597817691,
"learning_rate": 4.637234823364312e-08,
"loss": 0.7248,
"step": 1206
},
{
"epoch": 4.901522842639594,
"grad_norm": 0.18157931356114895,
"learning_rate": 4.258983697558838e-08,
"loss": 0.7172,
"step": 1207
},
{
"epoch": 4.905583756345178,
"grad_norm": 0.17572327512980632,
"learning_rate": 3.896805995533548e-08,
"loss": 0.7028,
"step": 1208
},
{
"epoch": 4.909644670050762,
"grad_norm": 0.1853567378749109,
"learning_rate": 3.550704634218028e-08,
"loss": 0.7123,
"step": 1209
},
{
"epoch": 4.913705583756345,
"grad_norm": 0.1788108246321232,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.7135,
"step": 1210
},
{
"epoch": 4.917766497461929,
"grad_norm": 0.17877208706013126,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.7134,
"step": 1211
},
{
"epoch": 4.921827411167513,
"grad_norm": 0.17178749878386051,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.7054,
"step": 1212
},
{
"epoch": 4.925888324873096,
"grad_norm": 0.18026445902409957,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.6817,
"step": 1213
},
{
"epoch": 4.92994923857868,
"grad_norm": 0.18092496265885005,
"learning_rate": 2.061435966221881e-08,
"loss": 0.7179,
"step": 1214
},
{
"epoch": 4.934010152284264,
"grad_norm": 0.18330859759394563,
"learning_rate": 1.811846652477245e-08,
"loss": 0.7068,
"step": 1215
},
{
"epoch": 4.938071065989847,
"grad_norm": 0.17551458763981342,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.6972,
"step": 1216
},
{
"epoch": 4.942131979695431,
"grad_norm": 0.17988455474433215,
"learning_rate": 1.360949303591097e-08,
"loss": 0.7066,
"step": 1217
},
{
"epoch": 4.946192893401015,
"grad_norm": 0.17438539082294452,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.7204,
"step": 1218
},
{
"epoch": 4.950253807106599,
"grad_norm": 0.1777485669648779,
"learning_rate": 9.744388816668172e-09,
"loss": 0.712,
"step": 1219
},
{
"epoch": 4.9543147208121825,
"grad_norm": 0.1806481462929512,
"learning_rate": 8.05332740472009e-09,
"loss": 0.7203,
"step": 1220
},
{
"epoch": 4.9583756345177665,
"grad_norm": 0.17547822411966987,
"learning_rate": 6.523278382872811e-09,
"loss": 0.7232,
"step": 1221
},
{
"epoch": 4.962436548223351,
"grad_norm": 0.178263122194253,
"learning_rate": 5.15425407393133e-09,
"loss": 0.7068,
"step": 1222
},
{
"epoch": 4.966497461928934,
"grad_norm": 0.17934887129542185,
"learning_rate": 3.94626550383137e-09,
"loss": 0.7235,
"step": 1223
},
{
"epoch": 4.970558375634518,
"grad_norm": 0.1775201512943909,
"learning_rate": 2.899322401546112e-09,
"loss": 0.7,
"step": 1224
},
{
"epoch": 4.974619289340102,
"grad_norm": 0.18437794081475634,
"learning_rate": 2.013433199010706e-09,
"loss": 0.7029,
"step": 1225
},
{
"epoch": 4.978680203045685,
"grad_norm": 0.1788793658490742,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.6835,
"step": 1226
},
{
"epoch": 4.982741116751269,
"grad_norm": 0.1782908175964806,
"learning_rate": 7.248437353468695e-10,
"loss": 0.7247,
"step": 1227
},
{
"epoch": 4.986802030456853,
"grad_norm": 0.17960842774184113,
"learning_rate": 3.221538523412449e-10,
"loss": 0.7062,
"step": 1228
},
{
"epoch": 4.990862944162437,
"grad_norm": 0.18477424162434355,
"learning_rate": 8.053862524670663e-11,
"loss": 0.6891,
"step": 1229
},
{
"epoch": 4.99492385786802,
"grad_norm": 0.17939234139703134,
"learning_rate": 0.0,
"loss": 0.6987,
"step": 1230
},
{
"epoch": 4.99492385786802,
"step": 1230,
"total_flos": 6.032180188216295e+18,
"train_loss": 0.8038945751461556,
"train_runtime": 33001.4558,
"train_samples_per_second": 4.774,
"train_steps_per_second": 0.037
}
],
"logging_steps": 1.0,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"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": 6.032180188216295e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}