Enigma-8B-Entity / trainer_state.json
SeacowX's picture
Upload folder using huggingface_hub
a4e4a59 verified
Raw
History Blame Contribute Delete
83.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 477,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006289308176100629,
"grad_norm": 2.5318546295166016,
"learning_rate": 1.0416666666666667e-06,
"loss": 1.6018,
"step": 1
},
{
"epoch": 0.012578616352201259,
"grad_norm": 2.7459726333618164,
"learning_rate": 2.0833333333333334e-06,
"loss": 1.6895,
"step": 2
},
{
"epoch": 0.018867924528301886,
"grad_norm": 2.351409435272217,
"learning_rate": 3.125e-06,
"loss": 1.6238,
"step": 3
},
{
"epoch": 0.025157232704402517,
"grad_norm": 2.7000679969787598,
"learning_rate": 4.166666666666667e-06,
"loss": 1.5833,
"step": 4
},
{
"epoch": 0.031446540880503145,
"grad_norm": 2.837592601776123,
"learning_rate": 5.208333333333334e-06,
"loss": 1.5915,
"step": 5
},
{
"epoch": 0.03773584905660377,
"grad_norm": 2.69355845451355,
"learning_rate": 6.25e-06,
"loss": 1.6902,
"step": 6
},
{
"epoch": 0.0440251572327044,
"grad_norm": 2.639601707458496,
"learning_rate": 7.2916666666666674e-06,
"loss": 1.5898,
"step": 7
},
{
"epoch": 0.050314465408805034,
"grad_norm": 2.6636252403259277,
"learning_rate": 8.333333333333334e-06,
"loss": 1.5769,
"step": 8
},
{
"epoch": 0.05660377358490566,
"grad_norm": 3.113805055618286,
"learning_rate": 9.375000000000001e-06,
"loss": 1.6211,
"step": 9
},
{
"epoch": 0.06289308176100629,
"grad_norm": 2.6526496410369873,
"learning_rate": 1.0416666666666668e-05,
"loss": 1.5546,
"step": 10
},
{
"epoch": 0.06918238993710692,
"grad_norm": 2.6448655128479004,
"learning_rate": 1.1458333333333333e-05,
"loss": 1.4448,
"step": 11
},
{
"epoch": 0.07547169811320754,
"grad_norm": 3.240588665008545,
"learning_rate": 1.25e-05,
"loss": 1.6157,
"step": 12
},
{
"epoch": 0.08176100628930817,
"grad_norm": 2.678084135055542,
"learning_rate": 1.3541666666666666e-05,
"loss": 1.6053,
"step": 13
},
{
"epoch": 0.0880503144654088,
"grad_norm": 2.71440052986145,
"learning_rate": 1.4583333333333335e-05,
"loss": 1.6036,
"step": 14
},
{
"epoch": 0.09433962264150944,
"grad_norm": 2.2286577224731445,
"learning_rate": 1.5625e-05,
"loss": 1.4786,
"step": 15
},
{
"epoch": 0.10062893081761007,
"grad_norm": 2.3145461082458496,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.476,
"step": 16
},
{
"epoch": 0.1069182389937107,
"grad_norm": 2.355117082595825,
"learning_rate": 1.7708333333333335e-05,
"loss": 1.5789,
"step": 17
},
{
"epoch": 0.11320754716981132,
"grad_norm": 2.285625696182251,
"learning_rate": 1.8750000000000002e-05,
"loss": 1.5044,
"step": 18
},
{
"epoch": 0.11949685534591195,
"grad_norm": 2.5451812744140625,
"learning_rate": 1.9791666666666665e-05,
"loss": 1.4246,
"step": 19
},
{
"epoch": 0.12578616352201258,
"grad_norm": 2.028719425201416,
"learning_rate": 2.0833333333333336e-05,
"loss": 1.3892,
"step": 20
},
{
"epoch": 0.1320754716981132,
"grad_norm": 2.2321243286132812,
"learning_rate": 2.1875e-05,
"loss": 1.3488,
"step": 21
},
{
"epoch": 0.13836477987421383,
"grad_norm": 2.434558391571045,
"learning_rate": 2.2916666666666667e-05,
"loss": 1.223,
"step": 22
},
{
"epoch": 0.14465408805031446,
"grad_norm": 2.4220376014709473,
"learning_rate": 2.3958333333333334e-05,
"loss": 1.2317,
"step": 23
},
{
"epoch": 0.1509433962264151,
"grad_norm": 2.2280728816986084,
"learning_rate": 2.5e-05,
"loss": 1.1992,
"step": 24
},
{
"epoch": 0.15723270440251572,
"grad_norm": 2.202923059463501,
"learning_rate": 2.604166666666667e-05,
"loss": 1.1245,
"step": 25
},
{
"epoch": 0.16352201257861634,
"grad_norm": 1.8481968641281128,
"learning_rate": 2.7083333333333332e-05,
"loss": 1.2006,
"step": 26
},
{
"epoch": 0.16981132075471697,
"grad_norm": 1.5634678602218628,
"learning_rate": 2.8125000000000003e-05,
"loss": 1.0542,
"step": 27
},
{
"epoch": 0.1761006289308176,
"grad_norm": 1.4136722087860107,
"learning_rate": 2.916666666666667e-05,
"loss": 1.033,
"step": 28
},
{
"epoch": 0.18238993710691823,
"grad_norm": 1.625587821006775,
"learning_rate": 3.0208333333333334e-05,
"loss": 0.9276,
"step": 29
},
{
"epoch": 0.18867924528301888,
"grad_norm": 1.856663703918457,
"learning_rate": 3.125e-05,
"loss": 0.9224,
"step": 30
},
{
"epoch": 0.1949685534591195,
"grad_norm": 1.6114718914031982,
"learning_rate": 3.229166666666667e-05,
"loss": 0.9932,
"step": 31
},
{
"epoch": 0.20125786163522014,
"grad_norm": 1.60017728805542,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.9436,
"step": 32
},
{
"epoch": 0.20754716981132076,
"grad_norm": 1.2032058238983154,
"learning_rate": 3.4375e-05,
"loss": 0.83,
"step": 33
},
{
"epoch": 0.2138364779874214,
"grad_norm": 1.226300835609436,
"learning_rate": 3.541666666666667e-05,
"loss": 0.8821,
"step": 34
},
{
"epoch": 0.22012578616352202,
"grad_norm": 1.2800050973892212,
"learning_rate": 3.6458333333333336e-05,
"loss": 0.8475,
"step": 35
},
{
"epoch": 0.22641509433962265,
"grad_norm": 0.9750305414199829,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.8542,
"step": 36
},
{
"epoch": 0.23270440251572327,
"grad_norm": 0.8264267444610596,
"learning_rate": 3.854166666666667e-05,
"loss": 0.8693,
"step": 37
},
{
"epoch": 0.2389937106918239,
"grad_norm": 1.2760239839553833,
"learning_rate": 3.958333333333333e-05,
"loss": 0.8813,
"step": 38
},
{
"epoch": 0.24528301886792453,
"grad_norm": 1.0647022724151611,
"learning_rate": 4.0625000000000005e-05,
"loss": 0.8369,
"step": 39
},
{
"epoch": 0.25157232704402516,
"grad_norm": 0.7672545313835144,
"learning_rate": 4.166666666666667e-05,
"loss": 0.7216,
"step": 40
},
{
"epoch": 0.2578616352201258,
"grad_norm": 0.712559700012207,
"learning_rate": 4.270833333333333e-05,
"loss": 0.7603,
"step": 41
},
{
"epoch": 0.2641509433962264,
"grad_norm": 0.6863624453544617,
"learning_rate": 4.375e-05,
"loss": 0.7656,
"step": 42
},
{
"epoch": 0.27044025157232704,
"grad_norm": 0.9794521331787109,
"learning_rate": 4.4791666666666673e-05,
"loss": 0.8198,
"step": 43
},
{
"epoch": 0.27672955974842767,
"grad_norm": 0.8873756527900696,
"learning_rate": 4.5833333333333334e-05,
"loss": 0.8332,
"step": 44
},
{
"epoch": 0.2830188679245283,
"grad_norm": 0.7220881581306458,
"learning_rate": 4.6875e-05,
"loss": 0.8236,
"step": 45
},
{
"epoch": 0.2893081761006289,
"grad_norm": 0.9265381097793579,
"learning_rate": 4.791666666666667e-05,
"loss": 0.7359,
"step": 46
},
{
"epoch": 0.29559748427672955,
"grad_norm": 0.7147327661514282,
"learning_rate": 4.8958333333333335e-05,
"loss": 0.7821,
"step": 47
},
{
"epoch": 0.3018867924528302,
"grad_norm": 0.672863781452179,
"learning_rate": 5e-05,
"loss": 0.6891,
"step": 48
},
{
"epoch": 0.3081761006289308,
"grad_norm": 0.9889981150627136,
"learning_rate": 4.9999329662935534e-05,
"loss": 0.8453,
"step": 49
},
{
"epoch": 0.31446540880503143,
"grad_norm": 0.9827463030815125,
"learning_rate": 4.999731868769027e-05,
"loss": 0.7842,
"step": 50
},
{
"epoch": 0.32075471698113206,
"grad_norm": 0.7701772451400757,
"learning_rate": 4.99939671821067e-05,
"loss": 0.7875,
"step": 51
},
{
"epoch": 0.3270440251572327,
"grad_norm": 0.6708323359489441,
"learning_rate": 4.998927532591592e-05,
"loss": 0.6967,
"step": 52
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.678805947303772,
"learning_rate": 4.9983243370727914e-05,
"loss": 0.708,
"step": 53
},
{
"epoch": 0.33962264150943394,
"grad_norm": 1.3635503053665161,
"learning_rate": 4.9975871640018154e-05,
"loss": 0.8041,
"step": 54
},
{
"epoch": 0.34591194968553457,
"grad_norm": 1.0370304584503174,
"learning_rate": 4.996716052911017e-05,
"loss": 0.6872,
"step": 55
},
{
"epoch": 0.3522012578616352,
"grad_norm": 0.7008804678916931,
"learning_rate": 4.99571105051544e-05,
"loss": 0.7349,
"step": 56
},
{
"epoch": 0.3584905660377358,
"grad_norm": 0.8846171498298645,
"learning_rate": 4.994572210710315e-05,
"loss": 0.678,
"step": 57
},
{
"epoch": 0.36477987421383645,
"grad_norm": 1.070142149925232,
"learning_rate": 4.993299594568163e-05,
"loss": 0.6986,
"step": 58
},
{
"epoch": 0.3710691823899371,
"grad_norm": 0.6238954067230225,
"learning_rate": 4.9918932703355256e-05,
"loss": 0.7054,
"step": 59
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.7648403644561768,
"learning_rate": 4.990353313429303e-05,
"loss": 0.7182,
"step": 60
},
{
"epoch": 0.3836477987421384,
"grad_norm": 1.210137128829956,
"learning_rate": 4.988679806432712e-05,
"loss": 0.7162,
"step": 61
},
{
"epoch": 0.389937106918239,
"grad_norm": 0.6953855752944946,
"learning_rate": 4.9868728390908526e-05,
"loss": 0.6737,
"step": 62
},
{
"epoch": 0.39622641509433965,
"grad_norm": 0.6835370659828186,
"learning_rate": 4.9849325083059e-05,
"loss": 0.7371,
"step": 63
},
{
"epoch": 0.4025157232704403,
"grad_norm": 0.7523362636566162,
"learning_rate": 4.982858918131906e-05,
"loss": 0.7202,
"step": 64
},
{
"epoch": 0.4088050314465409,
"grad_norm": 0.9010921120643616,
"learning_rate": 4.980652179769218e-05,
"loss": 0.7486,
"step": 65
},
{
"epoch": 0.41509433962264153,
"grad_norm": 0.6029642820358276,
"learning_rate": 4.978312411558518e-05,
"loss": 0.6713,
"step": 66
},
{
"epoch": 0.42138364779874216,
"grad_norm": 0.8067546486854553,
"learning_rate": 4.9758397389744734e-05,
"loss": 0.7145,
"step": 67
},
{
"epoch": 0.4276729559748428,
"grad_norm": 0.6919695138931274,
"learning_rate": 4.973234294619011e-05,
"loss": 0.6896,
"step": 68
},
{
"epoch": 0.4339622641509434,
"grad_norm": 0.7718349099159241,
"learning_rate": 4.9704962182142044e-05,
"loss": 0.6577,
"step": 69
},
{
"epoch": 0.44025157232704404,
"grad_norm": 1.3733798265457153,
"learning_rate": 4.967625656594782e-05,
"loss": 0.8091,
"step": 70
},
{
"epoch": 0.44654088050314467,
"grad_norm": 0.8210477232933044,
"learning_rate": 4.9646227637002515e-05,
"loss": 0.6669,
"step": 71
},
{
"epoch": 0.4528301886792453,
"grad_norm": 0.6245367527008057,
"learning_rate": 4.961487700566646e-05,
"loss": 0.669,
"step": 72
},
{
"epoch": 0.4591194968553459,
"grad_norm": 0.8577013611793518,
"learning_rate": 4.958220635317886e-05,
"loss": 0.6655,
"step": 73
},
{
"epoch": 0.46540880503144655,
"grad_norm": 1.2046078443527222,
"learning_rate": 4.9548217431567665e-05,
"loss": 0.8514,
"step": 74
},
{
"epoch": 0.4716981132075472,
"grad_norm": 0.9333224296569824,
"learning_rate": 4.95129120635556e-05,
"loss": 0.6504,
"step": 75
},
{
"epoch": 0.4779874213836478,
"grad_norm": 0.8769272565841675,
"learning_rate": 4.9476292142462374e-05,
"loss": 0.6981,
"step": 76
},
{
"epoch": 0.48427672955974843,
"grad_norm": 0.764488935470581,
"learning_rate": 4.943835963210324e-05,
"loss": 0.6861,
"step": 77
},
{
"epoch": 0.49056603773584906,
"grad_norm": 0.7639757990837097,
"learning_rate": 4.939911656668361e-05,
"loss": 0.6538,
"step": 78
},
{
"epoch": 0.4968553459119497,
"grad_norm": 0.8213962912559509,
"learning_rate": 4.9358565050689985e-05,
"loss": 0.7458,
"step": 79
},
{
"epoch": 0.5031446540880503,
"grad_norm": 0.8631553053855896,
"learning_rate": 4.93167072587771e-05,
"loss": 0.6948,
"step": 80
},
{
"epoch": 0.5094339622641509,
"grad_norm": 0.6010993123054504,
"learning_rate": 4.92735454356513e-05,
"loss": 0.5975,
"step": 81
},
{
"epoch": 0.5157232704402516,
"grad_norm": 0.7331316471099854,
"learning_rate": 4.922908189595018e-05,
"loss": 0.8162,
"step": 82
},
{
"epoch": 0.5220125786163522,
"grad_norm": 0.835134744644165,
"learning_rate": 4.9183319024118415e-05,
"loss": 0.6929,
"step": 83
},
{
"epoch": 0.5283018867924528,
"grad_norm": 0.8098347783088684,
"learning_rate": 4.9136259274279955e-05,
"loss": 0.6894,
"step": 84
},
{
"epoch": 0.5345911949685535,
"grad_norm": 0.8206546902656555,
"learning_rate": 4.908790517010636e-05,
"loss": 0.6453,
"step": 85
},
{
"epoch": 0.5408805031446541,
"grad_norm": 1.0443239212036133,
"learning_rate": 4.903825930468149e-05,
"loss": 0.726,
"step": 86
},
{
"epoch": 0.5471698113207547,
"grad_norm": 0.7683538198471069,
"learning_rate": 4.898732434036244e-05,
"loss": 0.6919,
"step": 87
},
{
"epoch": 0.5534591194968553,
"grad_norm": 0.7902626991271973,
"learning_rate": 4.893510300863676e-05,
"loss": 0.6655,
"step": 88
},
{
"epoch": 0.559748427672956,
"grad_norm": 0.7945076823234558,
"learning_rate": 4.8881598109976004e-05,
"loss": 0.6355,
"step": 89
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.9146498441696167,
"learning_rate": 4.8826812513685487e-05,
"loss": 0.6841,
"step": 90
},
{
"epoch": 0.5723270440251572,
"grad_norm": 0.6513539552688599,
"learning_rate": 4.877074915775049e-05,
"loss": 0.6766,
"step": 91
},
{
"epoch": 0.5786163522012578,
"grad_norm": 0.7359333038330078,
"learning_rate": 4.8713411048678635e-05,
"loss": 0.6613,
"step": 92
},
{
"epoch": 0.5849056603773585,
"grad_norm": 0.6931374073028564,
"learning_rate": 4.865480126133872e-05,
"loss": 0.6604,
"step": 93
},
{
"epoch": 0.5911949685534591,
"grad_norm": 0.7253872156143188,
"learning_rate": 4.859492293879574e-05,
"loss": 0.6573,
"step": 94
},
{
"epoch": 0.5974842767295597,
"grad_norm": 1.2044156789779663,
"learning_rate": 4.853377929214243e-05,
"loss": 0.651,
"step": 95
},
{
"epoch": 0.6037735849056604,
"grad_norm": 1.007851481437683,
"learning_rate": 4.8471373600326996e-05,
"loss": 0.657,
"step": 96
},
{
"epoch": 0.610062893081761,
"grad_norm": 0.8250524401664734,
"learning_rate": 4.8407709209977305e-05,
"loss": 0.7211,
"step": 97
},
{
"epoch": 0.6163522012578616,
"grad_norm": 1.0096355676651,
"learning_rate": 4.834278953522138e-05,
"loss": 0.6968,
"step": 98
},
{
"epoch": 0.6226415094339622,
"grad_norm": 0.8853664994239807,
"learning_rate": 4.8276618057504376e-05,
"loss": 0.7779,
"step": 99
},
{
"epoch": 0.6289308176100629,
"grad_norm": 0.6233201026916504,
"learning_rate": 4.8209198325401815e-05,
"loss": 0.6068,
"step": 100
},
{
"epoch": 0.6352201257861635,
"grad_norm": 1.0647857189178467,
"learning_rate": 4.8140533954429327e-05,
"loss": 0.6693,
"step": 101
},
{
"epoch": 0.6415094339622641,
"grad_norm": 0.7899872064590454,
"learning_rate": 4.8070628626848735e-05,
"loss": 0.6985,
"step": 102
},
{
"epoch": 0.6477987421383647,
"grad_norm": 0.7315335273742676,
"learning_rate": 4.799948609147061e-05,
"loss": 0.6299,
"step": 103
},
{
"epoch": 0.6540880503144654,
"grad_norm": 0.9970303177833557,
"learning_rate": 4.792711016345321e-05,
"loss": 0.7113,
"step": 104
},
{
"epoch": 0.660377358490566,
"grad_norm": 1.0047298669815063,
"learning_rate": 4.785350472409792e-05,
"loss": 0.6709,
"step": 105
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.7542774081230164,
"learning_rate": 4.777867372064105e-05,
"loss": 0.6476,
"step": 106
},
{
"epoch": 0.6729559748427673,
"grad_norm": 0.8545119166374207,
"learning_rate": 4.770262116604224e-05,
"loss": 0.6006,
"step": 107
},
{
"epoch": 0.6792452830188679,
"grad_norm": 0.8473503589630127,
"learning_rate": 4.762535113876917e-05,
"loss": 0.6766,
"step": 108
},
{
"epoch": 0.6855345911949685,
"grad_norm": 0.8628621697425842,
"learning_rate": 4.754686778257891e-05,
"loss": 0.6126,
"step": 109
},
{
"epoch": 0.6918238993710691,
"grad_norm": 0.7884383797645569,
"learning_rate": 4.7467175306295655e-05,
"loss": 0.7041,
"step": 110
},
{
"epoch": 0.6981132075471698,
"grad_norm": 0.8852160573005676,
"learning_rate": 4.738627798358506e-05,
"loss": 0.6032,
"step": 111
},
{
"epoch": 0.7044025157232704,
"grad_norm": 0.7209658026695251,
"learning_rate": 4.730418015272503e-05,
"loss": 0.7171,
"step": 112
},
{
"epoch": 0.710691823899371,
"grad_norm": 0.7943414449691772,
"learning_rate": 4.722088621637309e-05,
"loss": 0.6068,
"step": 113
},
{
"epoch": 0.7169811320754716,
"grad_norm": 1.1219490766525269,
"learning_rate": 4.713640064133025e-05,
"loss": 0.6797,
"step": 114
},
{
"epoch": 0.7232704402515723,
"grad_norm": 0.831230103969574,
"learning_rate": 4.7050727958301506e-05,
"loss": 0.5656,
"step": 115
},
{
"epoch": 0.7295597484276729,
"grad_norm": 0.7929229140281677,
"learning_rate": 4.6963872761652835e-05,
"loss": 0.6747,
"step": 116
},
{
"epoch": 0.7358490566037735,
"grad_norm": 0.7629885673522949,
"learning_rate": 4.687583970916487e-05,
"loss": 0.6348,
"step": 117
},
{
"epoch": 0.7421383647798742,
"grad_norm": 0.9467084407806396,
"learning_rate": 4.678663352178301e-05,
"loss": 0.6644,
"step": 118
},
{
"epoch": 0.7484276729559748,
"grad_norm": 0.8642951846122742,
"learning_rate": 4.6696258983364385e-05,
"loss": 0.6118,
"step": 119
},
{
"epoch": 0.7547169811320755,
"grad_norm": 1.1236990690231323,
"learning_rate": 4.660472094042121e-05,
"loss": 0.7634,
"step": 120
},
{
"epoch": 0.7610062893081762,
"grad_norm": 0.7560772895812988,
"learning_rate": 4.651202430186092e-05,
"loss": 0.6888,
"step": 121
},
{
"epoch": 0.7672955974842768,
"grad_norm": 0.7810654640197754,
"learning_rate": 4.641817403872293e-05,
"loss": 0.6577,
"step": 122
},
{
"epoch": 0.7735849056603774,
"grad_norm": 0.8273403644561768,
"learning_rate": 4.6323175183912024e-05,
"loss": 0.553,
"step": 123
},
{
"epoch": 0.779874213836478,
"grad_norm": 1.1021790504455566,
"learning_rate": 4.6227032831928484e-05,
"loss": 0.5661,
"step": 124
},
{
"epoch": 0.7861635220125787,
"grad_norm": 0.7645938992500305,
"learning_rate": 4.6129752138594874e-05,
"loss": 0.6227,
"step": 125
},
{
"epoch": 0.7924528301886793,
"grad_norm": 0.8086506128311157,
"learning_rate": 4.6031338320779534e-05,
"loss": 0.6925,
"step": 126
},
{
"epoch": 0.7987421383647799,
"grad_norm": 0.7653587460517883,
"learning_rate": 4.5931796656116846e-05,
"loss": 0.6714,
"step": 127
},
{
"epoch": 0.8050314465408805,
"grad_norm": 1.047590732574463,
"learning_rate": 4.5831132482724195e-05,
"loss": 0.7366,
"step": 128
},
{
"epoch": 0.8113207547169812,
"grad_norm": 0.8709453344345093,
"learning_rate": 4.572935119891571e-05,
"loss": 0.6268,
"step": 129
},
{
"epoch": 0.8176100628930818,
"grad_norm": 0.8644561767578125,
"learning_rate": 4.5626458262912745e-05,
"loss": 0.5889,
"step": 130
},
{
"epoch": 0.8238993710691824,
"grad_norm": 0.6348162293434143,
"learning_rate": 4.552245919255117e-05,
"loss": 0.5879,
"step": 131
},
{
"epoch": 0.8301886792452831,
"grad_norm": 0.6342223882675171,
"learning_rate": 4.541735956498554e-05,
"loss": 0.5568,
"step": 132
},
{
"epoch": 0.8364779874213837,
"grad_norm": 0.7982462048530579,
"learning_rate": 4.5311165016389916e-05,
"loss": 0.6787,
"step": 133
},
{
"epoch": 0.8427672955974843,
"grad_norm": 0.8147134780883789,
"learning_rate": 4.5203881241655644e-05,
"loss": 0.6351,
"step": 134
},
{
"epoch": 0.8490566037735849,
"grad_norm": 0.6963750720024109,
"learning_rate": 4.509551399408598e-05,
"loss": 0.7065,
"step": 135
},
{
"epoch": 0.8553459119496856,
"grad_norm": 0.7525794506072998,
"learning_rate": 4.498606908508754e-05,
"loss": 0.6583,
"step": 136
},
{
"epoch": 0.8616352201257862,
"grad_norm": 0.8069612979888916,
"learning_rate": 4.487555238385862e-05,
"loss": 0.6037,
"step": 137
},
{
"epoch": 0.8679245283018868,
"grad_norm": 0.773317813873291,
"learning_rate": 4.476396981707453e-05,
"loss": 0.6908,
"step": 138
},
{
"epoch": 0.8742138364779874,
"grad_norm": 0.8265529274940491,
"learning_rate": 4.465132736856969e-05,
"loss": 0.699,
"step": 139
},
{
"epoch": 0.8805031446540881,
"grad_norm": 0.8220138549804688,
"learning_rate": 4.453763107901675e-05,
"loss": 0.6086,
"step": 140
},
{
"epoch": 0.8867924528301887,
"grad_norm": 0.70572429895401,
"learning_rate": 4.442288704560268e-05,
"loss": 0.6745,
"step": 141
},
{
"epoch": 0.8930817610062893,
"grad_norm": 0.8666156530380249,
"learning_rate": 4.430710142170176e-05,
"loss": 0.5736,
"step": 142
},
{
"epoch": 0.89937106918239,
"grad_norm": 0.8797711730003357,
"learning_rate": 4.419028041654559e-05,
"loss": 0.6619,
"step": 143
},
{
"epoch": 0.9056603773584906,
"grad_norm": 0.6992596387863159,
"learning_rate": 4.4072430294890174e-05,
"loss": 0.725,
"step": 144
},
{
"epoch": 0.9119496855345912,
"grad_norm": 0.7946738004684448,
"learning_rate": 4.3953557376679856e-05,
"loss": 0.5861,
"step": 145
},
{
"epoch": 0.9182389937106918,
"grad_norm": 0.8084328174591064,
"learning_rate": 4.383366803670849e-05,
"loss": 0.6168,
"step": 146
},
{
"epoch": 0.9245283018867925,
"grad_norm": 0.8041918277740479,
"learning_rate": 4.371276870427753e-05,
"loss": 0.6769,
"step": 147
},
{
"epoch": 0.9308176100628931,
"grad_norm": 0.7114235758781433,
"learning_rate": 4.359086586285127e-05,
"loss": 0.6775,
"step": 148
},
{
"epoch": 0.9371069182389937,
"grad_norm": 0.820907473564148,
"learning_rate": 4.346796604970912e-05,
"loss": 0.6901,
"step": 149
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.8040430545806885,
"learning_rate": 4.3344075855595104e-05,
"loss": 0.6304,
"step": 150
},
{
"epoch": 0.949685534591195,
"grad_norm": 0.7677490711212158,
"learning_rate": 4.321920192436433e-05,
"loss": 0.6333,
"step": 151
},
{
"epoch": 0.9559748427672956,
"grad_norm": 0.7418397068977356,
"learning_rate": 4.309335095262676e-05,
"loss": 0.6051,
"step": 152
},
{
"epoch": 0.9622641509433962,
"grad_norm": 0.8183079361915588,
"learning_rate": 4.296652968938807e-05,
"loss": 0.6408,
"step": 153
},
{
"epoch": 0.9685534591194969,
"grad_norm": 0.7114039063453674,
"learning_rate": 4.283874493568772e-05,
"loss": 0.6751,
"step": 154
},
{
"epoch": 0.9748427672955975,
"grad_norm": 0.733608603477478,
"learning_rate": 4.271000354423426e-05,
"loss": 0.6308,
"step": 155
},
{
"epoch": 0.9811320754716981,
"grad_norm": 0.8840383887290955,
"learning_rate": 4.258031241903778e-05,
"loss": 0.6405,
"step": 156
},
{
"epoch": 0.9874213836477987,
"grad_norm": 0.7877086400985718,
"learning_rate": 4.2449678515039747e-05,
"loss": 0.6391,
"step": 157
},
{
"epoch": 0.9937106918238994,
"grad_norm": 0.7041771411895752,
"learning_rate": 4.231810883773999e-05,
"loss": 0.6274,
"step": 158
},
{
"epoch": 1.0,
"grad_norm": 0.7660680413246155,
"learning_rate": 4.218561044282099e-05,
"loss": 0.6457,
"step": 159
},
{
"epoch": 1.0062893081761006,
"grad_norm": 0.6843758821487427,
"learning_rate": 4.2052190435769554e-05,
"loss": 0.611,
"step": 160
},
{
"epoch": 1.0125786163522013,
"grad_norm": 0.6859325766563416,
"learning_rate": 4.191785597149577e-05,
"loss": 0.5462,
"step": 161
},
{
"epoch": 1.0188679245283019,
"grad_norm": 0.7766200304031372,
"learning_rate": 4.178261425394926e-05,
"loss": 0.5949,
"step": 162
},
{
"epoch": 1.0251572327044025,
"grad_norm": 0.7334004044532776,
"learning_rate": 4.1646472535732895e-05,
"loss": 0.5932,
"step": 163
},
{
"epoch": 1.0314465408805031,
"grad_norm": 0.6686878204345703,
"learning_rate": 4.1509438117713866e-05,
"loss": 0.6281,
"step": 164
},
{
"epoch": 1.0377358490566038,
"grad_norm": 0.8288165330886841,
"learning_rate": 4.137151834863213e-05,
"loss": 0.6103,
"step": 165
},
{
"epoch": 1.0440251572327044,
"grad_norm": 0.6973861455917358,
"learning_rate": 4.123272062470633e-05,
"loss": 0.5275,
"step": 166
},
{
"epoch": 1.050314465408805,
"grad_norm": 0.8053054809570312,
"learning_rate": 4.109305238923718e-05,
"loss": 0.5642,
"step": 167
},
{
"epoch": 1.0566037735849056,
"grad_norm": 0.7544634342193604,
"learning_rate": 4.095252113220827e-05,
"loss": 0.585,
"step": 168
},
{
"epoch": 1.0628930817610063,
"grad_norm": 0.8335475325584412,
"learning_rate": 4.0811134389884433e-05,
"loss": 0.571,
"step": 169
},
{
"epoch": 1.069182389937107,
"grad_norm": 0.8077796101570129,
"learning_rate": 4.066889974440757e-05,
"loss": 0.5511,
"step": 170
},
{
"epoch": 1.0754716981132075,
"grad_norm": 0.8840991258621216,
"learning_rate": 4.0525824823390045e-05,
"loss": 0.6526,
"step": 171
},
{
"epoch": 1.0817610062893082,
"grad_norm": 0.9405627250671387,
"learning_rate": 4.038191729950569e-05,
"loss": 0.6236,
"step": 172
},
{
"epoch": 1.0880503144654088,
"grad_norm": 0.9547109603881836,
"learning_rate": 4.0237184890078245e-05,
"loss": 0.6448,
"step": 173
},
{
"epoch": 1.0943396226415094,
"grad_norm": 0.8912748098373413,
"learning_rate": 4.009163535666761e-05,
"loss": 0.6699,
"step": 174
},
{
"epoch": 1.10062893081761,
"grad_norm": 0.8230434656143188,
"learning_rate": 3.994527650465352e-05,
"loss": 0.619,
"step": 175
},
{
"epoch": 1.1069182389937107,
"grad_norm": 0.6602234244346619,
"learning_rate": 3.979811618281706e-05,
"loss": 0.6719,
"step": 176
},
{
"epoch": 1.1132075471698113,
"grad_norm": 1.181139588356018,
"learning_rate": 3.9650162282919655e-05,
"loss": 0.6399,
"step": 177
},
{
"epoch": 1.119496855345912,
"grad_norm": 0.8026434779167175,
"learning_rate": 3.9501422739279956e-05,
"loss": 0.6176,
"step": 178
},
{
"epoch": 1.1257861635220126,
"grad_norm": 1.0383801460266113,
"learning_rate": 3.9351905528348285e-05,
"loss": 0.6039,
"step": 179
},
{
"epoch": 1.1320754716981132,
"grad_norm": 0.8335999846458435,
"learning_rate": 3.920161866827889e-05,
"loss": 0.6061,
"step": 180
},
{
"epoch": 1.1383647798742138,
"grad_norm": 1.3670942783355713,
"learning_rate": 3.90505702185e-05,
"loss": 0.6929,
"step": 181
},
{
"epoch": 1.1446540880503144,
"grad_norm": 1.1789627075195312,
"learning_rate": 3.889876827928156e-05,
"loss": 0.6374,
"step": 182
},
{
"epoch": 1.150943396226415,
"grad_norm": 0.948314905166626,
"learning_rate": 3.874622099130087e-05,
"loss": 0.6136,
"step": 183
},
{
"epoch": 1.1572327044025157,
"grad_norm": 0.8498350977897644,
"learning_rate": 3.859293653520604e-05,
"loss": 0.6533,
"step": 184
},
{
"epoch": 1.1635220125786163,
"grad_norm": 0.7883232235908508,
"learning_rate": 3.843892313117724e-05,
"loss": 0.6433,
"step": 185
},
{
"epoch": 1.169811320754717,
"grad_norm": 1.0547206401824951,
"learning_rate": 3.8284189038485936e-05,
"loss": 0.6022,
"step": 186
},
{
"epoch": 1.1761006289308176,
"grad_norm": 0.8304398655891418,
"learning_rate": 3.812874255505191e-05,
"loss": 0.5888,
"step": 187
},
{
"epoch": 1.1823899371069182,
"grad_norm": 0.9702231287956238,
"learning_rate": 3.797259201699833e-05,
"loss": 0.6215,
"step": 188
},
{
"epoch": 1.1886792452830188,
"grad_norm": 0.6644477844238281,
"learning_rate": 3.781574579820464e-05,
"loss": 0.5924,
"step": 189
},
{
"epoch": 1.1949685534591195,
"grad_norm": 0.7866007685661316,
"learning_rate": 3.765821230985758e-05,
"loss": 0.5827,
"step": 190
},
{
"epoch": 1.20125786163522,
"grad_norm": 0.8880365490913391,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.6907,
"step": 191
},
{
"epoch": 1.2075471698113207,
"grad_norm": 0.7273608446121216,
"learning_rate": 3.7341117353077966e-05,
"loss": 0.6059,
"step": 192
},
{
"epoch": 1.2138364779874213,
"grad_norm": 0.7370182275772095,
"learning_rate": 3.718157288948563e-05,
"loss": 0.5832,
"step": 193
},
{
"epoch": 1.220125786163522,
"grad_norm": 0.7611547708511353,
"learning_rate": 3.702137516510838e-05,
"loss": 0.6606,
"step": 194
},
{
"epoch": 1.2264150943396226,
"grad_norm": 0.697180986404419,
"learning_rate": 3.686053277086401e-05,
"loss": 0.6076,
"step": 195
},
{
"epoch": 1.2327044025157232,
"grad_norm": 0.9172202348709106,
"learning_rate": 3.669905433224199e-05,
"loss": 0.5693,
"step": 196
},
{
"epoch": 1.2389937106918238,
"grad_norm": 0.7670121192932129,
"learning_rate": 3.653694850884091e-05,
"loss": 0.5446,
"step": 197
},
{
"epoch": 1.2452830188679245,
"grad_norm": 0.7891936898231506,
"learning_rate": 3.637422399390413e-05,
"loss": 0.5647,
"step": 198
},
{
"epoch": 1.251572327044025,
"grad_norm": 0.8863941431045532,
"learning_rate": 3.621088951385353e-05,
"loss": 0.6175,
"step": 199
},
{
"epoch": 1.2578616352201257,
"grad_norm": 0.9106245040893555,
"learning_rate": 3.604695382782159e-05,
"loss": 0.7153,
"step": 200
},
{
"epoch": 1.2641509433962264,
"grad_norm": 0.7456657290458679,
"learning_rate": 3.588242572718162e-05,
"loss": 0.6109,
"step": 201
},
{
"epoch": 1.270440251572327,
"grad_norm": 0.8054442405700684,
"learning_rate": 3.5717314035076355e-05,
"loss": 0.603,
"step": 202
},
{
"epoch": 1.2767295597484276,
"grad_norm": 1.0004535913467407,
"learning_rate": 3.5551627605944745e-05,
"loss": 0.6583,
"step": 203
},
{
"epoch": 1.2830188679245282,
"grad_norm": 0.8054452538490295,
"learning_rate": 3.5385375325047166e-05,
"loss": 0.5918,
"step": 204
},
{
"epoch": 1.2893081761006289,
"grad_norm": 0.7591025233268738,
"learning_rate": 3.521856610798887e-05,
"loss": 0.6072,
"step": 205
},
{
"epoch": 1.2955974842767295,
"grad_norm": 0.8540089726448059,
"learning_rate": 3.505120890024195e-05,
"loss": 0.635,
"step": 206
},
{
"epoch": 1.3018867924528301,
"grad_norm": 0.7391782402992249,
"learning_rate": 3.4883312676665536e-05,
"loss": 0.6589,
"step": 207
},
{
"epoch": 1.3081761006289307,
"grad_norm": 0.8612866997718811,
"learning_rate": 3.4714886441024574e-05,
"loss": 0.5573,
"step": 208
},
{
"epoch": 1.3144654088050314,
"grad_norm": 0.8475955128669739,
"learning_rate": 3.4545939225506934e-05,
"loss": 0.5782,
"step": 209
},
{
"epoch": 1.320754716981132,
"grad_norm": 0.8738673329353333,
"learning_rate": 3.437648009023905e-05,
"loss": 0.6133,
"step": 210
},
{
"epoch": 1.3270440251572326,
"grad_norm": 0.7458251714706421,
"learning_rate": 3.420651812280006e-05,
"loss": 0.6063,
"step": 211
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.76874178647995,
"learning_rate": 3.4036062437734484e-05,
"loss": 0.5561,
"step": 212
},
{
"epoch": 1.3396226415094339,
"grad_norm": 0.7305719256401062,
"learning_rate": 3.386512217606339e-05,
"loss": 0.5965,
"step": 213
},
{
"epoch": 1.3459119496855345,
"grad_norm": 0.7534472942352295,
"learning_rate": 3.369370650479425e-05,
"loss": 0.6081,
"step": 214
},
{
"epoch": 1.3522012578616351,
"grad_norm": 0.8087597489356995,
"learning_rate": 3.3521824616429285e-05,
"loss": 0.5653,
"step": 215
},
{
"epoch": 1.3584905660377358,
"grad_norm": 0.800054669380188,
"learning_rate": 3.3349485728472535e-05,
"loss": 0.6375,
"step": 216
},
{
"epoch": 1.3647798742138364,
"grad_norm": 0.7617058157920837,
"learning_rate": 3.3176699082935545e-05,
"loss": 0.6542,
"step": 217
},
{
"epoch": 1.371069182389937,
"grad_norm": 0.7484142780303955,
"learning_rate": 3.300347394584172e-05,
"loss": 0.6585,
"step": 218
},
{
"epoch": 1.3773584905660377,
"grad_norm": 0.7602940201759338,
"learning_rate": 3.282981960672948e-05,
"loss": 0.6099,
"step": 219
},
{
"epoch": 1.3836477987421385,
"grad_norm": 0.7106953263282776,
"learning_rate": 3.265574537815398e-05,
"loss": 0.5839,
"step": 220
},
{
"epoch": 1.389937106918239,
"grad_norm": 0.7194884419441223,
"learning_rate": 3.248126059518785e-05,
"loss": 0.5964,
"step": 221
},
{
"epoch": 1.3962264150943398,
"grad_norm": 0.7843396067619324,
"learning_rate": 3.230637461492043e-05,
"loss": 0.5906,
"step": 222
},
{
"epoch": 1.4025157232704402,
"grad_norm": 0.8458132147789001,
"learning_rate": 3.213109681595612e-05,
"loss": 0.6096,
"step": 223
},
{
"epoch": 1.408805031446541,
"grad_norm": 0.703157365322113,
"learning_rate": 3.195543659791132e-05,
"loss": 0.5957,
"step": 224
},
{
"epoch": 1.4150943396226414,
"grad_norm": 0.7627823948860168,
"learning_rate": 3.177940338091043e-05,
"loss": 0.5193,
"step": 225
},
{
"epoch": 1.4213836477987423,
"grad_norm": 0.835606575012207,
"learning_rate": 3.160300660508064e-05,
"loss": 0.6124,
"step": 226
},
{
"epoch": 1.4276729559748427,
"grad_norm": 0.693316638469696,
"learning_rate": 3.14262557300457e-05,
"loss": 0.5558,
"step": 227
},
{
"epoch": 1.4339622641509435,
"grad_norm": 0.8970513343811035,
"learning_rate": 3.124916023441865e-05,
"loss": 0.6101,
"step": 228
},
{
"epoch": 1.440251572327044,
"grad_norm": 0.7977336049079895,
"learning_rate": 3.107172961529343e-05,
"loss": 0.5399,
"step": 229
},
{
"epoch": 1.4465408805031448,
"grad_norm": 0.9715563654899597,
"learning_rate": 3.0893973387735687e-05,
"loss": 0.7266,
"step": 230
},
{
"epoch": 1.4528301886792452,
"grad_norm": 0.9589596390724182,
"learning_rate": 3.071590108427244e-05,
"loss": 0.6833,
"step": 231
},
{
"epoch": 1.459119496855346,
"grad_norm": 0.811178982257843,
"learning_rate": 3.0537522254380905e-05,
"loss": 0.5289,
"step": 232
},
{
"epoch": 1.4654088050314464,
"grad_norm": 0.8802748918533325,
"learning_rate": 3.0358846463976372e-05,
"loss": 0.6149,
"step": 233
},
{
"epoch": 1.4716981132075473,
"grad_norm": 0.8504813313484192,
"learning_rate": 3.017988329489923e-05,
"loss": 0.6165,
"step": 234
},
{
"epoch": 1.4779874213836477,
"grad_norm": 0.8345064520835876,
"learning_rate": 3.0000642344401113e-05,
"loss": 0.6646,
"step": 235
},
{
"epoch": 1.4842767295597485,
"grad_norm": 0.8303079009056091,
"learning_rate": 2.9821133224630226e-05,
"loss": 0.4849,
"step": 236
},
{
"epoch": 1.490566037735849,
"grad_norm": 0.734706461429596,
"learning_rate": 2.9641365562115887e-05,
"loss": 0.5404,
"step": 237
},
{
"epoch": 1.4968553459119498,
"grad_norm": 1.0077383518218994,
"learning_rate": 2.9461348997252265e-05,
"loss": 0.6095,
"step": 238
},
{
"epoch": 1.5031446540880502,
"grad_norm": 0.9368531107902527,
"learning_rate": 2.9281093183781403e-05,
"loss": 0.6215,
"step": 239
},
{
"epoch": 1.509433962264151,
"grad_norm": 0.9052350521087646,
"learning_rate": 2.9100607788275545e-05,
"loss": 0.5891,
"step": 240
},
{
"epoch": 1.5157232704402515,
"grad_norm": 0.891499400138855,
"learning_rate": 2.891990248961871e-05,
"loss": 0.5447,
"step": 241
},
{
"epoch": 1.5220125786163523,
"grad_norm": 0.7840732932090759,
"learning_rate": 2.8738986978487625e-05,
"loss": 0.5978,
"step": 242
},
{
"epoch": 1.5283018867924527,
"grad_norm": 0.7833483815193176,
"learning_rate": 2.8557870956832132e-05,
"loss": 0.5102,
"step": 243
},
{
"epoch": 1.5345911949685536,
"grad_norm": 0.7420686483383179,
"learning_rate": 2.8376564137354795e-05,
"loss": 0.5829,
"step": 244
},
{
"epoch": 1.540880503144654,
"grad_norm": 0.9215426445007324,
"learning_rate": 2.8195076242990122e-05,
"loss": 0.6041,
"step": 245
},
{
"epoch": 1.5471698113207548,
"grad_norm": 0.9076328277587891,
"learning_rate": 2.8013417006383076e-05,
"loss": 0.657,
"step": 246
},
{
"epoch": 1.5534591194968552,
"grad_norm": 1.0412029027938843,
"learning_rate": 2.783159616936723e-05,
"loss": 0.669,
"step": 247
},
{
"epoch": 1.559748427672956,
"grad_norm": 0.7854359149932861,
"learning_rate": 2.764962348244228e-05,
"loss": 0.5842,
"step": 248
},
{
"epoch": 1.5660377358490565,
"grad_norm": 0.8531441688537598,
"learning_rate": 2.7467508704251137e-05,
"loss": 0.6051,
"step": 249
},
{
"epoch": 1.5723270440251573,
"grad_norm": 0.9410574436187744,
"learning_rate": 2.7285261601056698e-05,
"loss": 0.5669,
"step": 250
},
{
"epoch": 1.5786163522012577,
"grad_norm": 0.8966511487960815,
"learning_rate": 2.7102891946217994e-05,
"loss": 0.6213,
"step": 251
},
{
"epoch": 1.5849056603773586,
"grad_norm": 0.9011272192001343,
"learning_rate": 2.6920409519666174e-05,
"loss": 0.6148,
"step": 252
},
{
"epoch": 1.591194968553459,
"grad_norm": 0.9505216479301453,
"learning_rate": 2.6737824107379948e-05,
"loss": 0.6813,
"step": 253
},
{
"epoch": 1.5974842767295598,
"grad_norm": 0.894383430480957,
"learning_rate": 2.655514550086086e-05,
"loss": 0.5579,
"step": 254
},
{
"epoch": 1.6037735849056602,
"grad_norm": 0.8340818285942078,
"learning_rate": 2.637238349660819e-05,
"loss": 0.7153,
"step": 255
},
{
"epoch": 1.610062893081761,
"grad_norm": 0.7479216456413269,
"learning_rate": 2.6189547895593562e-05,
"loss": 0.5377,
"step": 256
},
{
"epoch": 1.6163522012578615,
"grad_norm": 1.0334053039550781,
"learning_rate": 2.600664850273538e-05,
"loss": 0.6289,
"step": 257
},
{
"epoch": 1.6226415094339623,
"grad_norm": 0.9294121861457825,
"learning_rate": 2.582369512637302e-05,
"loss": 0.579,
"step": 258
},
{
"epoch": 1.6289308176100628,
"grad_norm": 0.8203226923942566,
"learning_rate": 2.564069757774082e-05,
"loss": 0.5132,
"step": 259
},
{
"epoch": 1.6352201257861636,
"grad_norm": 1.0053868293762207,
"learning_rate": 2.545766567044194e-05,
"loss": 0.6269,
"step": 260
},
{
"epoch": 1.641509433962264,
"grad_norm": 0.7887965440750122,
"learning_rate": 2.527460921992209e-05,
"loss": 0.5898,
"step": 261
},
{
"epoch": 1.6477987421383649,
"grad_norm": 0.8424786925315857,
"learning_rate": 2.509153804294318e-05,
"loss": 0.5945,
"step": 262
},
{
"epoch": 1.6540880503144653,
"grad_norm": 0.8613922595977783,
"learning_rate": 2.490846195705683e-05,
"loss": 0.5688,
"step": 263
},
{
"epoch": 1.6603773584905661,
"grad_norm": 0.9355970025062561,
"learning_rate": 2.4725390780077908e-05,
"loss": 0.6624,
"step": 264
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.7711520195007324,
"learning_rate": 2.4542334329558077e-05,
"loss": 0.5243,
"step": 265
},
{
"epoch": 1.6729559748427674,
"grad_norm": 0.8319422602653503,
"learning_rate": 2.435930242225919e-05,
"loss": 0.5546,
"step": 266
},
{
"epoch": 1.6792452830188678,
"grad_norm": 0.8753066658973694,
"learning_rate": 2.4176304873626985e-05,
"loss": 0.578,
"step": 267
},
{
"epoch": 1.6855345911949686,
"grad_norm": 0.879960298538208,
"learning_rate": 2.399335149726463e-05,
"loss": 0.6359,
"step": 268
},
{
"epoch": 1.691823899371069,
"grad_norm": 0.8653309345245361,
"learning_rate": 2.3810452104406444e-05,
"loss": 0.655,
"step": 269
},
{
"epoch": 1.6981132075471699,
"grad_norm": 0.9247555136680603,
"learning_rate": 2.3627616503391814e-05,
"loss": 0.6245,
"step": 270
},
{
"epoch": 1.7044025157232703,
"grad_norm": 0.8340408205986023,
"learning_rate": 2.3444854499139142e-05,
"loss": 0.5982,
"step": 271
},
{
"epoch": 1.7106918238993711,
"grad_norm": 0.9371923804283142,
"learning_rate": 2.3262175892620065e-05,
"loss": 0.5817,
"step": 272
},
{
"epoch": 1.7169811320754715,
"grad_norm": 0.809624195098877,
"learning_rate": 2.307959048033383e-05,
"loss": 0.5853,
"step": 273
},
{
"epoch": 1.7232704402515724,
"grad_norm": 1.0021802186965942,
"learning_rate": 2.2897108053782e-05,
"loss": 0.6333,
"step": 274
},
{
"epoch": 1.7295597484276728,
"grad_norm": 0.8453190922737122,
"learning_rate": 2.271473839894331e-05,
"loss": 0.5381,
"step": 275
},
{
"epoch": 1.7358490566037736,
"grad_norm": 0.7890949845314026,
"learning_rate": 2.253249129574887e-05,
"loss": 0.5658,
"step": 276
},
{
"epoch": 1.742138364779874,
"grad_norm": 0.8816287517547607,
"learning_rate": 2.2350376517557727e-05,
"loss": 0.5995,
"step": 277
},
{
"epoch": 1.748427672955975,
"grad_norm": 0.881336510181427,
"learning_rate": 2.216840383063277e-05,
"loss": 0.6086,
"step": 278
},
{
"epoch": 1.7547169811320755,
"grad_norm": 0.9889141917228699,
"learning_rate": 2.1986582993616926e-05,
"loss": 0.6055,
"step": 279
},
{
"epoch": 1.7610062893081762,
"grad_norm": 0.8085553646087646,
"learning_rate": 2.1804923757009884e-05,
"loss": 0.5715,
"step": 280
},
{
"epoch": 1.7672955974842768,
"grad_norm": 0.8645254969596863,
"learning_rate": 2.1623435862645204e-05,
"loss": 0.5559,
"step": 281
},
{
"epoch": 1.7735849056603774,
"grad_norm": 0.7751315236091614,
"learning_rate": 2.1442129043167874e-05,
"loss": 0.5651,
"step": 282
},
{
"epoch": 1.779874213836478,
"grad_norm": 0.8653814196586609,
"learning_rate": 2.126101302151238e-05,
"loss": 0.5811,
"step": 283
},
{
"epoch": 1.7861635220125787,
"grad_norm": 0.7941970229148865,
"learning_rate": 2.1080097510381298e-05,
"loss": 0.6164,
"step": 284
},
{
"epoch": 1.7924528301886793,
"grad_norm": 0.9225692749023438,
"learning_rate": 2.089939221172446e-05,
"loss": 0.6913,
"step": 285
},
{
"epoch": 1.79874213836478,
"grad_norm": 0.8978100419044495,
"learning_rate": 2.07189068162186e-05,
"loss": 0.6091,
"step": 286
},
{
"epoch": 1.8050314465408805,
"grad_norm": 0.7802568078041077,
"learning_rate": 2.0538651002747744e-05,
"loss": 0.5726,
"step": 287
},
{
"epoch": 1.8113207547169812,
"grad_norm": 0.920458197593689,
"learning_rate": 2.0358634437884112e-05,
"loss": 0.6116,
"step": 288
},
{
"epoch": 1.8176100628930818,
"grad_norm": 0.8483282327651978,
"learning_rate": 2.0178866775369777e-05,
"loss": 0.6037,
"step": 289
},
{
"epoch": 1.8238993710691824,
"grad_norm": 0.7381348013877869,
"learning_rate": 1.9999357655598893e-05,
"loss": 0.6082,
"step": 290
},
{
"epoch": 1.830188679245283,
"grad_norm": 0.781938374042511,
"learning_rate": 1.9820116705100777e-05,
"loss": 0.6105,
"step": 291
},
{
"epoch": 1.8364779874213837,
"grad_norm": 0.9458007216453552,
"learning_rate": 1.9641153536023644e-05,
"loss": 0.622,
"step": 292
},
{
"epoch": 1.8427672955974843,
"grad_norm": 0.8598377108573914,
"learning_rate": 1.9462477745619108e-05,
"loss": 0.5713,
"step": 293
},
{
"epoch": 1.849056603773585,
"grad_norm": 0.9226989150047302,
"learning_rate": 1.928409891572757e-05,
"loss": 0.6458,
"step": 294
},
{
"epoch": 1.8553459119496856,
"grad_norm": 0.7962223887443542,
"learning_rate": 1.9106026612264316e-05,
"loss": 0.6411,
"step": 295
},
{
"epoch": 1.8616352201257862,
"grad_norm": 0.8416233658790588,
"learning_rate": 1.8928270384706584e-05,
"loss": 0.6442,
"step": 296
},
{
"epoch": 1.8679245283018868,
"grad_norm": 0.8757483959197998,
"learning_rate": 1.875083976558136e-05,
"loss": 0.5864,
"step": 297
},
{
"epoch": 1.8742138364779874,
"grad_norm": 0.9159104824066162,
"learning_rate": 1.8573744269954298e-05,
"loss": 0.6433,
"step": 298
},
{
"epoch": 1.880503144654088,
"grad_norm": 0.8131967782974243,
"learning_rate": 1.8396993394919372e-05,
"loss": 0.5951,
"step": 299
},
{
"epoch": 1.8867924528301887,
"grad_norm": 0.8790562152862549,
"learning_rate": 1.8220596619089576e-05,
"loss": 0.733,
"step": 300
},
{
"epoch": 1.8930817610062893,
"grad_norm": 0.8865863680839539,
"learning_rate": 1.8044563402088684e-05,
"loss": 0.6455,
"step": 301
},
{
"epoch": 1.89937106918239,
"grad_norm": 0.8361108303070068,
"learning_rate": 1.7868903184043887e-05,
"loss": 0.5735,
"step": 302
},
{
"epoch": 1.9056603773584906,
"grad_norm": 1.0793273448944092,
"learning_rate": 1.7693625385079577e-05,
"loss": 0.575,
"step": 303
},
{
"epoch": 1.9119496855345912,
"grad_norm": 0.9358295202255249,
"learning_rate": 1.7518739404812155e-05,
"loss": 0.6469,
"step": 304
},
{
"epoch": 1.9182389937106918,
"grad_norm": 0.7908286452293396,
"learning_rate": 1.7344254621846016e-05,
"loss": 0.5925,
"step": 305
},
{
"epoch": 1.9245283018867925,
"grad_norm": 0.8354206085205078,
"learning_rate": 1.7170180393270532e-05,
"loss": 0.5952,
"step": 306
},
{
"epoch": 1.930817610062893,
"grad_norm": 0.7533159852027893,
"learning_rate": 1.699652605415828e-05,
"loss": 0.5285,
"step": 307
},
{
"epoch": 1.9371069182389937,
"grad_norm": 0.8126991987228394,
"learning_rate": 1.682330091706446e-05,
"loss": 0.6097,
"step": 308
},
{
"epoch": 1.9433962264150944,
"grad_norm": 0.7812114953994751,
"learning_rate": 1.6650514271527468e-05,
"loss": 0.5392,
"step": 309
},
{
"epoch": 1.949685534591195,
"grad_norm": 0.8412722945213318,
"learning_rate": 1.647817538357072e-05,
"loss": 0.577,
"step": 310
},
{
"epoch": 1.9559748427672956,
"grad_norm": 0.9080637693405151,
"learning_rate": 1.630629349520576e-05,
"loss": 0.5666,
"step": 311
},
{
"epoch": 1.9622641509433962,
"grad_norm": 0.8412050008773804,
"learning_rate": 1.613487782393661e-05,
"loss": 0.5836,
"step": 312
},
{
"epoch": 1.9685534591194969,
"grad_norm": 0.8351119160652161,
"learning_rate": 1.5963937562265525e-05,
"loss": 0.5527,
"step": 313
},
{
"epoch": 1.9748427672955975,
"grad_norm": 0.8003585338592529,
"learning_rate": 1.5793481877199946e-05,
"loss": 0.5683,
"step": 314
},
{
"epoch": 1.9811320754716981,
"grad_norm": 0.8032575249671936,
"learning_rate": 1.5623519909760954e-05,
"loss": 0.6225,
"step": 315
},
{
"epoch": 1.9874213836477987,
"grad_norm": 0.9350284337997437,
"learning_rate": 1.5454060774493068e-05,
"loss": 0.6303,
"step": 316
},
{
"epoch": 1.9937106918238994,
"grad_norm": 0.8984176516532898,
"learning_rate": 1.528511355897543e-05,
"loss": 0.57,
"step": 317
},
{
"epoch": 2.0,
"grad_norm": 1.0674775838851929,
"learning_rate": 1.5116687323334467e-05,
"loss": 0.6437,
"step": 318
},
{
"epoch": 2.006289308176101,
"grad_norm": 0.8595516085624695,
"learning_rate": 1.4948791099758052e-05,
"loss": 0.5796,
"step": 319
},
{
"epoch": 2.0125786163522013,
"grad_norm": 0.8340380191802979,
"learning_rate": 1.4781433892011131e-05,
"loss": 0.5662,
"step": 320
},
{
"epoch": 2.018867924528302,
"grad_norm": 0.7874680757522583,
"learning_rate": 1.4614624674952842e-05,
"loss": 0.5754,
"step": 321
},
{
"epoch": 2.0251572327044025,
"grad_norm": 1.0123980045318604,
"learning_rate": 1.4448372394055249e-05,
"loss": 0.5358,
"step": 322
},
{
"epoch": 2.0314465408805034,
"grad_norm": 0.9125322699546814,
"learning_rate": 1.4282685964923642e-05,
"loss": 0.5623,
"step": 323
},
{
"epoch": 2.0377358490566038,
"grad_norm": 0.9834954142570496,
"learning_rate": 1.4117574272818388e-05,
"loss": 0.5465,
"step": 324
},
{
"epoch": 2.0440251572327046,
"grad_norm": 0.7798795700073242,
"learning_rate": 1.3953046172178414e-05,
"loss": 0.4833,
"step": 325
},
{
"epoch": 2.050314465408805,
"grad_norm": 0.7403257489204407,
"learning_rate": 1.378911048614647e-05,
"loss": 0.5041,
"step": 326
},
{
"epoch": 2.056603773584906,
"grad_norm": 0.9695855379104614,
"learning_rate": 1.3625776006095881e-05,
"loss": 0.614,
"step": 327
},
{
"epoch": 2.0628930817610063,
"grad_norm": 0.8196994066238403,
"learning_rate": 1.3463051491159096e-05,
"loss": 0.5469,
"step": 328
},
{
"epoch": 2.069182389937107,
"grad_norm": 0.9879975318908691,
"learning_rate": 1.3300945667758014e-05,
"loss": 0.5432,
"step": 329
},
{
"epoch": 2.0754716981132075,
"grad_norm": 0.940609335899353,
"learning_rate": 1.3139467229135999e-05,
"loss": 0.5998,
"step": 330
},
{
"epoch": 2.0817610062893084,
"grad_norm": 0.9598972201347351,
"learning_rate": 1.2978624834891628e-05,
"loss": 0.5617,
"step": 331
},
{
"epoch": 2.088050314465409,
"grad_norm": 0.9208015203475952,
"learning_rate": 1.281842711051438e-05,
"loss": 0.5593,
"step": 332
},
{
"epoch": 2.0943396226415096,
"grad_norm": 0.9645323157310486,
"learning_rate": 1.2658882646922034e-05,
"loss": 0.5613,
"step": 333
},
{
"epoch": 2.10062893081761,
"grad_norm": 0.9999901652336121,
"learning_rate": 1.2500000000000006e-05,
"loss": 0.5894,
"step": 334
},
{
"epoch": 2.106918238993711,
"grad_norm": 0.9951053857803345,
"learning_rate": 1.2341787690142437e-05,
"loss": 0.6147,
"step": 335
},
{
"epoch": 2.1132075471698113,
"grad_norm": 0.8768017292022705,
"learning_rate": 1.2184254201795365e-05,
"loss": 0.5767,
"step": 336
},
{
"epoch": 2.119496855345912,
"grad_norm": 1.089101791381836,
"learning_rate": 1.2027407983001681e-05,
"loss": 0.548,
"step": 337
},
{
"epoch": 2.1257861635220126,
"grad_norm": 1.0265827178955078,
"learning_rate": 1.1871257444948098e-05,
"loss": 0.6397,
"step": 338
},
{
"epoch": 2.1320754716981134,
"grad_norm": 1.3154712915420532,
"learning_rate": 1.1715810961514073e-05,
"loss": 0.6168,
"step": 339
},
{
"epoch": 2.138364779874214,
"grad_norm": 0.9637579321861267,
"learning_rate": 1.1561076868822756e-05,
"loss": 0.6011,
"step": 340
},
{
"epoch": 2.1446540880503147,
"grad_norm": 1.0628957748413086,
"learning_rate": 1.1407063464793966e-05,
"loss": 0.5153,
"step": 341
},
{
"epoch": 2.150943396226415,
"grad_norm": 0.9873301982879639,
"learning_rate": 1.1253779008699131e-05,
"loss": 0.5839,
"step": 342
},
{
"epoch": 2.157232704402516,
"grad_norm": 0.8832650780677795,
"learning_rate": 1.1101231720718442e-05,
"loss": 0.5196,
"step": 343
},
{
"epoch": 2.1635220125786163,
"grad_norm": 0.9620294570922852,
"learning_rate": 1.09494297815e-05,
"loss": 0.5057,
"step": 344
},
{
"epoch": 2.169811320754717,
"grad_norm": 0.861911416053772,
"learning_rate": 1.0798381331721109e-05,
"loss": 0.5579,
"step": 345
},
{
"epoch": 2.1761006289308176,
"grad_norm": 1.0682628154754639,
"learning_rate": 1.0648094471651724e-05,
"loss": 0.5442,
"step": 346
},
{
"epoch": 2.1823899371069184,
"grad_norm": 0.9263089895248413,
"learning_rate": 1.049857726072005e-05,
"loss": 0.534,
"step": 347
},
{
"epoch": 2.188679245283019,
"grad_norm": 1.1437605619430542,
"learning_rate": 1.034983771708035e-05,
"loss": 0.6663,
"step": 348
},
{
"epoch": 2.1949685534591197,
"grad_norm": 1.0065386295318604,
"learning_rate": 1.020188381718295e-05,
"loss": 0.5785,
"step": 349
},
{
"epoch": 2.20125786163522,
"grad_norm": 0.9750943183898926,
"learning_rate": 1.0054723495346482e-05,
"loss": 0.6158,
"step": 350
},
{
"epoch": 2.207547169811321,
"grad_norm": 0.8873117566108704,
"learning_rate": 9.908364643332399e-06,
"loss": 0.5427,
"step": 351
},
{
"epoch": 2.2138364779874213,
"grad_norm": 0.8856627345085144,
"learning_rate": 9.762815109921761e-06,
"loss": 0.5662,
"step": 352
},
{
"epoch": 2.220125786163522,
"grad_norm": 1.0704549551010132,
"learning_rate": 9.618082700494319e-06,
"loss": 0.5534,
"step": 353
},
{
"epoch": 2.2264150943396226,
"grad_norm": 0.8676052689552307,
"learning_rate": 9.474175176609956e-06,
"loss": 0.59,
"step": 354
},
{
"epoch": 2.2327044025157234,
"grad_norm": 0.9761330485343933,
"learning_rate": 9.331100255592437e-06,
"loss": 0.4916,
"step": 355
},
{
"epoch": 2.238993710691824,
"grad_norm": 0.97756028175354,
"learning_rate": 9.18886561011557e-06,
"loss": 0.588,
"step": 356
},
{
"epoch": 2.2452830188679247,
"grad_norm": 0.841642439365387,
"learning_rate": 9.047478867791732e-06,
"loss": 0.5049,
"step": 357
},
{
"epoch": 2.251572327044025,
"grad_norm": 0.8155388236045837,
"learning_rate": 8.906947610762825e-06,
"loss": 0.5161,
"step": 358
},
{
"epoch": 2.257861635220126,
"grad_norm": 0.9711548089981079,
"learning_rate": 8.767279375293672e-06,
"loss": 0.586,
"step": 359
},
{
"epoch": 2.2641509433962264,
"grad_norm": 0.9738316535949707,
"learning_rate": 8.628481651367876e-06,
"loss": 0.5676,
"step": 360
},
{
"epoch": 2.270440251572327,
"grad_norm": 1.0838391780853271,
"learning_rate": 8.490561882286136e-06,
"loss": 0.5541,
"step": 361
},
{
"epoch": 2.2767295597484276,
"grad_norm": 0.8917685151100159,
"learning_rate": 8.353527464267104e-06,
"loss": 0.5196,
"step": 362
},
{
"epoch": 2.2830188679245285,
"grad_norm": 1.020671010017395,
"learning_rate": 8.217385746050742e-06,
"loss": 0.5488,
"step": 363
},
{
"epoch": 2.289308176100629,
"grad_norm": 0.9560969471931458,
"learning_rate": 8.082144028504233e-06,
"loss": 0.6031,
"step": 364
},
{
"epoch": 2.2955974842767297,
"grad_norm": 0.9737635254859924,
"learning_rate": 7.947809564230445e-06,
"loss": 0.5155,
"step": 365
},
{
"epoch": 2.30188679245283,
"grad_norm": 0.9711655974388123,
"learning_rate": 7.814389557179017e-06,
"loss": 0.5838,
"step": 366
},
{
"epoch": 2.308176100628931,
"grad_norm": 0.9898532032966614,
"learning_rate": 7.681891162260015e-06,
"loss": 0.5451,
"step": 367
},
{
"epoch": 2.3144654088050314,
"grad_norm": 1.035120964050293,
"learning_rate": 7.5503214849602516e-06,
"loss": 0.5736,
"step": 368
},
{
"epoch": 2.3207547169811322,
"grad_norm": 1.108474850654602,
"learning_rate": 7.419687580962223e-06,
"loss": 0.5196,
"step": 369
},
{
"epoch": 2.3270440251572326,
"grad_norm": 1.0157012939453125,
"learning_rate": 7.289996455765749e-06,
"loss": 0.51,
"step": 370
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.894260823726654,
"learning_rate": 7.161255064312283e-06,
"loss": 0.5139,
"step": 371
},
{
"epoch": 2.339622641509434,
"grad_norm": 0.9161332845687866,
"learning_rate": 7.033470310611945e-06,
"loss": 0.6037,
"step": 372
},
{
"epoch": 2.3459119496855347,
"grad_norm": 1.046667218208313,
"learning_rate": 6.906649047373246e-06,
"loss": 0.5373,
"step": 373
},
{
"epoch": 2.352201257861635,
"grad_norm": 1.0308722257614136,
"learning_rate": 6.780798075635675e-06,
"loss": 0.4893,
"step": 374
},
{
"epoch": 2.358490566037736,
"grad_norm": 0.9212855696678162,
"learning_rate": 6.655924144404907e-06,
"loss": 0.4754,
"step": 375
},
{
"epoch": 2.3647798742138364,
"grad_norm": 0.9538508057594299,
"learning_rate": 6.532033950290886e-06,
"loss": 0.5595,
"step": 376
},
{
"epoch": 2.3710691823899372,
"grad_norm": 0.9696520566940308,
"learning_rate": 6.409134137148737e-06,
"loss": 0.5625,
"step": 377
},
{
"epoch": 2.3773584905660377,
"grad_norm": 1.0636050701141357,
"learning_rate": 6.28723129572247e-06,
"loss": 0.5553,
"step": 378
},
{
"epoch": 2.3836477987421385,
"grad_norm": 1.0527743101119995,
"learning_rate": 6.166331963291519e-06,
"loss": 0.5119,
"step": 379
},
{
"epoch": 2.389937106918239,
"grad_norm": 0.9810575246810913,
"learning_rate": 6.046442623320145e-06,
"loss": 0.5445,
"step": 380
},
{
"epoch": 2.3962264150943398,
"grad_norm": 1.3199591636657715,
"learning_rate": 5.927569705109828e-06,
"loss": 0.5123,
"step": 381
},
{
"epoch": 2.40251572327044,
"grad_norm": 1.0990368127822876,
"learning_rate": 5.809719583454415e-06,
"loss": 0.5272,
"step": 382
},
{
"epoch": 2.408805031446541,
"grad_norm": 0.9801096320152283,
"learning_rate": 5.692898578298253e-06,
"loss": 0.5352,
"step": 383
},
{
"epoch": 2.4150943396226414,
"grad_norm": 0.9592974185943604,
"learning_rate": 5.577112954397321e-06,
"loss": 0.5599,
"step": 384
},
{
"epoch": 2.4213836477987423,
"grad_norm": 1.0604385137557983,
"learning_rate": 5.462368920983249e-06,
"loss": 0.6035,
"step": 385
},
{
"epoch": 2.4276729559748427,
"grad_norm": 1.105743408203125,
"learning_rate": 5.348672631430318e-06,
"loss": 0.5487,
"step": 386
},
{
"epoch": 2.4339622641509435,
"grad_norm": 1.0688592195510864,
"learning_rate": 5.236030182925475e-06,
"loss": 0.4962,
"step": 387
},
{
"epoch": 2.440251572327044,
"grad_norm": 1.0261510610580444,
"learning_rate": 5.124447616141381e-06,
"loss": 0.4908,
"step": 388
},
{
"epoch": 2.4465408805031448,
"grad_norm": 0.9577153921127319,
"learning_rate": 5.013930914912476e-06,
"loss": 0.5737,
"step": 389
},
{
"epoch": 2.452830188679245,
"grad_norm": 1.094467282295227,
"learning_rate": 4.9044860059140275e-06,
"loss": 0.545,
"step": 390
},
{
"epoch": 2.459119496855346,
"grad_norm": 1.0198063850402832,
"learning_rate": 4.796118758344354e-06,
"loss": 0.5605,
"step": 391
},
{
"epoch": 2.4654088050314464,
"grad_norm": 1.099183201789856,
"learning_rate": 4.6888349836100825e-06,
"loss": 0.613,
"step": 392
},
{
"epoch": 2.4716981132075473,
"grad_norm": 0.9367800354957581,
"learning_rate": 4.58264043501446e-06,
"loss": 0.5336,
"step": 393
},
{
"epoch": 2.4779874213836477,
"grad_norm": 1.0678683519363403,
"learning_rate": 4.477540807448832e-06,
"loss": 0.5223,
"step": 394
},
{
"epoch": 2.4842767295597485,
"grad_norm": 0.9716767072677612,
"learning_rate": 4.373541737087264e-06,
"loss": 0.6065,
"step": 395
},
{
"epoch": 2.490566037735849,
"grad_norm": 0.9876382946968079,
"learning_rate": 4.270648801084296e-06,
"loss": 0.6004,
"step": 396
},
{
"epoch": 2.49685534591195,
"grad_norm": 1.13174307346344,
"learning_rate": 4.1688675172758064e-06,
"loss": 0.7008,
"step": 397
},
{
"epoch": 2.50314465408805,
"grad_norm": 0.9936075806617737,
"learning_rate": 4.068203343883159e-06,
"loss": 0.5324,
"step": 398
},
{
"epoch": 2.509433962264151,
"grad_norm": 1.0240283012390137,
"learning_rate": 3.968661679220468e-06,
"loss": 0.5039,
"step": 399
},
{
"epoch": 2.5157232704402515,
"grad_norm": 0.9729213714599609,
"learning_rate": 3.8702478614051355e-06,
"loss": 0.6125,
"step": 400
},
{
"epoch": 2.5220125786163523,
"grad_norm": 1.0876688957214355,
"learning_rate": 3.772967168071517e-06,
"loss": 0.4964,
"step": 401
},
{
"epoch": 2.5283018867924527,
"grad_norm": 0.9535592198371887,
"learning_rate": 3.6768248160879787e-06,
"loss": 0.5612,
"step": 402
},
{
"epoch": 2.5345911949685536,
"grad_norm": 1.031736135482788,
"learning_rate": 3.5818259612770744e-06,
"loss": 0.5821,
"step": 403
},
{
"epoch": 2.540880503144654,
"grad_norm": 1.1026850938796997,
"learning_rate": 3.487975698139084e-06,
"loss": 0.5122,
"step": 404
},
{
"epoch": 2.547169811320755,
"grad_norm": 1.2503842115402222,
"learning_rate": 3.3952790595787987e-06,
"loss": 0.5523,
"step": 405
},
{
"epoch": 2.5534591194968552,
"grad_norm": 0.9601788520812988,
"learning_rate": 3.3037410166356143e-06,
"loss": 0.5387,
"step": 406
},
{
"epoch": 2.559748427672956,
"grad_norm": 1.054858922958374,
"learning_rate": 3.2133664782169948e-06,
"loss": 0.5589,
"step": 407
},
{
"epoch": 2.5660377358490565,
"grad_norm": 1.025303840637207,
"learning_rate": 3.1241602908351404e-06,
"loss": 0.4714,
"step": 408
},
{
"epoch": 2.5723270440251573,
"grad_norm": 1.1350743770599365,
"learning_rate": 3.036127238347164e-06,
"loss": 0.6239,
"step": 409
},
{
"epoch": 2.5786163522012577,
"grad_norm": 1.0093308687210083,
"learning_rate": 2.9492720416985e-06,
"loss": 0.4931,
"step": 410
},
{
"epoch": 2.5849056603773586,
"grad_norm": 0.9581872820854187,
"learning_rate": 2.8635993586697553e-06,
"loss": 0.5811,
"step": 411
},
{
"epoch": 2.591194968553459,
"grad_norm": 0.9479830265045166,
"learning_rate": 2.779113783626916e-06,
"loss": 0.5309,
"step": 412
},
{
"epoch": 2.59748427672956,
"grad_norm": 1.0558853149414062,
"learning_rate": 2.6958198472749717e-06,
"loss": 0.5677,
"step": 413
},
{
"epoch": 2.6037735849056602,
"grad_norm": 1.0703245401382446,
"learning_rate": 2.6137220164149435e-06,
"loss": 0.5454,
"step": 414
},
{
"epoch": 2.610062893081761,
"grad_norm": 1.0348020792007446,
"learning_rate": 2.5328246937043526e-06,
"loss": 0.5347,
"step": 415
},
{
"epoch": 2.6163522012578615,
"grad_norm": 1.0240215063095093,
"learning_rate": 2.4531322174210975e-06,
"loss": 0.5439,
"step": 416
},
{
"epoch": 2.6226415094339623,
"grad_norm": 1.0210105180740356,
"learning_rate": 2.37464886123083e-06,
"loss": 0.5493,
"step": 417
},
{
"epoch": 2.6289308176100628,
"grad_norm": 0.9978793859481812,
"learning_rate": 2.2973788339577613e-06,
"loss": 0.5734,
"step": 418
},
{
"epoch": 2.6352201257861636,
"grad_norm": 0.9935867786407471,
"learning_rate": 2.2213262793589484e-06,
"loss": 0.5357,
"step": 419
},
{
"epoch": 2.641509433962264,
"grad_norm": 1.1258329153060913,
"learning_rate": 2.1464952759020855e-06,
"loss": 0.5981,
"step": 420
},
{
"epoch": 2.647798742138365,
"grad_norm": 1.0054690837860107,
"learning_rate": 2.0728898365467903e-06,
"loss": 0.5475,
"step": 421
},
{
"epoch": 2.6540880503144653,
"grad_norm": 0.992341935634613,
"learning_rate": 2.0005139085293945e-06,
"loss": 0.5183,
"step": 422
},
{
"epoch": 2.660377358490566,
"grad_norm": 1.1585534811019897,
"learning_rate": 1.9293713731512673e-06,
"loss": 0.5818,
"step": 423
},
{
"epoch": 2.6666666666666665,
"grad_norm": 1.0974853038787842,
"learning_rate": 1.8594660455706763e-06,
"loss": 0.5459,
"step": 424
},
{
"epoch": 2.6729559748427674,
"grad_norm": 1.0383155345916748,
"learning_rate": 1.790801674598186e-06,
"loss": 0.5606,
"step": 425
},
{
"epoch": 2.6792452830188678,
"grad_norm": 1.0877803564071655,
"learning_rate": 1.7233819424956248e-06,
"loss": 0.6469,
"step": 426
},
{
"epoch": 2.6855345911949686,
"grad_norm": 0.9916452765464783,
"learning_rate": 1.6572104647786247e-06,
"loss": 0.5502,
"step": 427
},
{
"epoch": 2.691823899371069,
"grad_norm": 1.0767418146133423,
"learning_rate": 1.5922907900227018e-06,
"loss": 0.4795,
"step": 428
},
{
"epoch": 2.69811320754717,
"grad_norm": 0.943517804145813,
"learning_rate": 1.5286263996730026e-06,
"loss": 0.5133,
"step": 429
},
{
"epoch": 2.7044025157232703,
"grad_norm": 0.9800017476081848,
"learning_rate": 1.4662207078575684e-06,
"loss": 0.5562,
"step": 430
},
{
"epoch": 2.710691823899371,
"grad_norm": 1.1385478973388672,
"learning_rate": 1.40507706120426e-06,
"loss": 0.6125,
"step": 431
},
{
"epoch": 2.7169811320754715,
"grad_norm": 1.0203759670257568,
"learning_rate": 1.3451987386612851e-06,
"loss": 0.4892,
"step": 432
},
{
"epoch": 2.7232704402515724,
"grad_norm": 1.0325732231140137,
"learning_rate": 1.286588951321363e-06,
"loss": 0.5683,
"step": 433
},
{
"epoch": 2.729559748427673,
"grad_norm": 1.0057847499847412,
"learning_rate": 1.2292508422495158e-06,
"loss": 0.5591,
"step": 434
},
{
"epoch": 2.7358490566037736,
"grad_norm": 1.0370678901672363,
"learning_rate": 1.1731874863145143e-06,
"loss": 0.5453,
"step": 435
},
{
"epoch": 2.742138364779874,
"grad_norm": 0.9245080947875977,
"learning_rate": 1.1184018900240011e-06,
"loss": 0.4967,
"step": 436
},
{
"epoch": 2.748427672955975,
"grad_norm": 0.9071083068847656,
"learning_rate": 1.06489699136324e-06,
"loss": 0.5167,
"step": 437
},
{
"epoch": 2.7547169811320753,
"grad_norm": 1.0417287349700928,
"learning_rate": 1.0126756596375686e-06,
"loss": 0.5565,
"step": 438
},
{
"epoch": 2.761006289308176,
"grad_norm": 1.044604778289795,
"learning_rate": 9.617406953185138e-07,
"loss": 0.5653,
"step": 439
},
{
"epoch": 2.767295597484277,
"grad_norm": 1.0384100675582886,
"learning_rate": 9.120948298936421e-07,
"loss": 0.526,
"step": 440
},
{
"epoch": 2.7735849056603774,
"grad_norm": 1.1161764860153198,
"learning_rate": 8.637407257200497e-07,
"loss": 0.543,
"step": 441
},
{
"epoch": 2.779874213836478,
"grad_norm": 1.095703363418579,
"learning_rate": 8.166809758815896e-07,
"loss": 0.5837,
"step": 442
},
{
"epoch": 2.7861635220125787,
"grad_norm": 1.0628708600997925,
"learning_rate": 7.709181040498254e-07,
"loss": 0.5353,
"step": 443
},
{
"epoch": 2.7924528301886795,
"grad_norm": 1.014708161354065,
"learning_rate": 7.264545643486997e-07,
"loss": 0.5941,
"step": 444
},
{
"epoch": 2.79874213836478,
"grad_norm": 0.9685489535331726,
"learning_rate": 6.832927412229018e-07,
"loss": 0.5311,
"step": 445
},
{
"epoch": 2.8050314465408803,
"grad_norm": 0.9544473886489868,
"learning_rate": 6.41434949310013e-07,
"loss": 0.5635,
"step": 446
},
{
"epoch": 2.811320754716981,
"grad_norm": 1.2015351057052612,
"learning_rate": 6.008834333163876e-07,
"loss": 0.5833,
"step": 447
},
{
"epoch": 2.817610062893082,
"grad_norm": 0.9438028931617737,
"learning_rate": 5.616403678967624e-07,
"loss": 0.4877,
"step": 448
},
{
"epoch": 2.8238993710691824,
"grad_norm": 0.9827612638473511,
"learning_rate": 5.237078575376336e-07,
"loss": 0.5336,
"step": 449
},
{
"epoch": 2.830188679245283,
"grad_norm": 1.0386241674423218,
"learning_rate": 4.870879364444109e-07,
"loss": 0.5439,
"step": 450
},
{
"epoch": 2.8364779874213837,
"grad_norm": 1.0768251419067383,
"learning_rate": 4.517825684323324e-07,
"loss": 0.5641,
"step": 451
},
{
"epoch": 2.8427672955974845,
"grad_norm": 0.9616789221763611,
"learning_rate": 4.1779364682113796e-07,
"loss": 0.5653,
"step": 452
},
{
"epoch": 2.849056603773585,
"grad_norm": 1.0044620037078857,
"learning_rate": 3.851229943335394e-07,
"loss": 0.5828,
"step": 453
},
{
"epoch": 2.8553459119496853,
"grad_norm": 0.9143046140670776,
"learning_rate": 3.537723629974815e-07,
"loss": 0.5,
"step": 454
},
{
"epoch": 2.861635220125786,
"grad_norm": 1.1500418186187744,
"learning_rate": 3.237434340521789e-07,
"loss": 0.6026,
"step": 455
},
{
"epoch": 2.867924528301887,
"grad_norm": 1.0443059206008911,
"learning_rate": 2.9503781785795713e-07,
"loss": 0.5329,
"step": 456
},
{
"epoch": 2.8742138364779874,
"grad_norm": 1.0647647380828857,
"learning_rate": 2.6765705380989437e-07,
"loss": 0.5699,
"step": 457
},
{
"epoch": 2.880503144654088,
"grad_norm": 0.9754281640052795,
"learning_rate": 2.416026102552732e-07,
"loss": 0.5928,
"step": 458
},
{
"epoch": 2.8867924528301887,
"grad_norm": 0.9932259917259216,
"learning_rate": 2.168758844148272e-07,
"loss": 0.591,
"step": 459
},
{
"epoch": 2.8930817610062896,
"grad_norm": 1.1134239435195923,
"learning_rate": 1.93478202307823e-07,
"loss": 0.5862,
"step": 460
},
{
"epoch": 2.89937106918239,
"grad_norm": 1.1245214939117432,
"learning_rate": 1.7141081868094212e-07,
"loss": 0.5749,
"step": 461
},
{
"epoch": 2.9056603773584904,
"grad_norm": 0.9979386925697327,
"learning_rate": 1.5067491694100154e-07,
"loss": 0.5694,
"step": 462
},
{
"epoch": 2.911949685534591,
"grad_norm": 1.018856406211853,
"learning_rate": 1.3127160909147672e-07,
"loss": 0.5602,
"step": 463
},
{
"epoch": 2.918238993710692,
"grad_norm": 1.103405237197876,
"learning_rate": 1.1320193567288529e-07,
"loss": 0.6153,
"step": 464
},
{
"epoch": 2.9245283018867925,
"grad_norm": 1.1032378673553467,
"learning_rate": 9.646686570697061e-08,
"loss": 0.5614,
"step": 465
},
{
"epoch": 2.930817610062893,
"grad_norm": 1.0579830408096313,
"learning_rate": 8.106729664475176e-08,
"loss": 0.4928,
"step": 466
},
{
"epoch": 2.9371069182389937,
"grad_norm": 1.0092562437057495,
"learning_rate": 6.700405431837587e-08,
"loss": 0.5309,
"step": 467
},
{
"epoch": 2.9433962264150946,
"grad_norm": 0.923986554145813,
"learning_rate": 5.4277892896853476e-08,
"loss": 0.5312,
"step": 468
},
{
"epoch": 2.949685534591195,
"grad_norm": 1.07733154296875,
"learning_rate": 4.2889494845599344e-08,
"loss": 0.5962,
"step": 469
},
{
"epoch": 2.9559748427672954,
"grad_norm": 0.9723581671714783,
"learning_rate": 3.283947088983663e-08,
"loss": 0.479,
"step": 470
},
{
"epoch": 2.9622641509433962,
"grad_norm": 0.8795850872993469,
"learning_rate": 2.4128359981850924e-08,
"loss": 0.5356,
"step": 471
},
{
"epoch": 2.968553459119497,
"grad_norm": 1.0015406608581543,
"learning_rate": 1.6756629272085545e-08,
"loss": 0.641,
"step": 472
},
{
"epoch": 2.9748427672955975,
"grad_norm": 0.9870020747184753,
"learning_rate": 1.0724674084083841e-08,
"loss": 0.5117,
"step": 473
},
{
"epoch": 2.981132075471698,
"grad_norm": 1.0196905136108398,
"learning_rate": 6.032817893297793e-09,
"loss": 0.5305,
"step": 474
},
{
"epoch": 2.9874213836477987,
"grad_norm": 1.0410959720611572,
"learning_rate": 2.681312309735229e-09,
"loss": 0.5545,
"step": 475
},
{
"epoch": 2.9937106918238996,
"grad_norm": 1.0021966695785522,
"learning_rate": 6.703370644706164e-10,
"loss": 0.547,
"step": 476
},
{
"epoch": 3.0,
"grad_norm": 1.0657471418380737,
"learning_rate": 0.0,
"loss": 0.457,
"step": 477
},
{
"epoch": 3.0,
"step": 477,
"total_flos": 5.5856876418308506e+17,
"train_loss": 0.6629369193278019,
"train_runtime": 4969.9444,
"train_samples_per_second": 12.276,
"train_steps_per_second": 0.096
}
],
"logging_steps": 1,
"max_steps": 477,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.5856876418308506e+17,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}