Enigma-70B-Entity / trainer_state.json
SeacowX's picture
Upload folder using huggingface_hub
41721e6 verified
Raw
History Blame Contribute Delete
83.3 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": 4.8737874031066895,
"learning_rate": 1.0416666666666667e-06,
"loss": 3.6915,
"step": 1
},
{
"epoch": 0.012578616352201259,
"grad_norm": 4.941007137298584,
"learning_rate": 2.0833333333333334e-06,
"loss": 3.6352,
"step": 2
},
{
"epoch": 0.018867924528301886,
"grad_norm": 5.258031368255615,
"learning_rate": 3.125e-06,
"loss": 3.5816,
"step": 3
},
{
"epoch": 0.025157232704402517,
"grad_norm": 5.544358253479004,
"learning_rate": 4.166666666666667e-06,
"loss": 3.6233,
"step": 4
},
{
"epoch": 0.031446540880503145,
"grad_norm": 5.457708835601807,
"learning_rate": 5.208333333333334e-06,
"loss": 3.6107,
"step": 5
},
{
"epoch": 0.03773584905660377,
"grad_norm": 5.185223579406738,
"learning_rate": 6.25e-06,
"loss": 3.6277,
"step": 6
},
{
"epoch": 0.0440251572327044,
"grad_norm": 5.433485507965088,
"learning_rate": 7.2916666666666674e-06,
"loss": 3.4974,
"step": 7
},
{
"epoch": 0.050314465408805034,
"grad_norm": 5.38355016708374,
"learning_rate": 8.333333333333334e-06,
"loss": 3.5021,
"step": 8
},
{
"epoch": 0.05660377358490566,
"grad_norm": 5.851156234741211,
"learning_rate": 9.375000000000001e-06,
"loss": 3.5888,
"step": 9
},
{
"epoch": 0.06289308176100629,
"grad_norm": 8.335844993591309,
"learning_rate": 1.0416666666666668e-05,
"loss": 3.3652,
"step": 10
},
{
"epoch": 0.06918238993710692,
"grad_norm": 5.598583698272705,
"learning_rate": 1.1458333333333333e-05,
"loss": 3.1476,
"step": 11
},
{
"epoch": 0.07547169811320754,
"grad_norm": 6.811880588531494,
"learning_rate": 1.25e-05,
"loss": 3.6846,
"step": 12
},
{
"epoch": 0.08176100628930817,
"grad_norm": 5.962316513061523,
"learning_rate": 1.3541666666666666e-05,
"loss": 3.2959,
"step": 13
},
{
"epoch": 0.0880503144654088,
"grad_norm": 6.434226989746094,
"learning_rate": 1.4583333333333335e-05,
"loss": 3.3629,
"step": 14
},
{
"epoch": 0.09433962264150944,
"grad_norm": 5.959097862243652,
"learning_rate": 1.5625e-05,
"loss": 2.9799,
"step": 15
},
{
"epoch": 0.10062893081761007,
"grad_norm": 5.713625907897949,
"learning_rate": 1.6666666666666667e-05,
"loss": 2.8375,
"step": 16
},
{
"epoch": 0.1069182389937107,
"grad_norm": 4.679070472717285,
"learning_rate": 1.7708333333333335e-05,
"loss": 2.5298,
"step": 17
},
{
"epoch": 0.11320754716981132,
"grad_norm": 4.000007152557373,
"learning_rate": 1.8750000000000002e-05,
"loss": 2.4099,
"step": 18
},
{
"epoch": 0.11949685534591195,
"grad_norm": 3.6054141521453857,
"learning_rate": 1.9791666666666665e-05,
"loss": 2.1316,
"step": 19
},
{
"epoch": 0.12578616352201258,
"grad_norm": 3.1357152462005615,
"learning_rate": 2.0833333333333336e-05,
"loss": 1.8814,
"step": 20
},
{
"epoch": 0.1320754716981132,
"grad_norm": 2.848811149597168,
"learning_rate": 2.1875e-05,
"loss": 1.7786,
"step": 21
},
{
"epoch": 0.13836477987421383,
"grad_norm": 2.4006130695343018,
"learning_rate": 2.2916666666666667e-05,
"loss": 1.5133,
"step": 22
},
{
"epoch": 0.14465408805031446,
"grad_norm": 2.2537267208099365,
"learning_rate": 2.3958333333333334e-05,
"loss": 1.4229,
"step": 23
},
{
"epoch": 0.1509433962264151,
"grad_norm": 2.1117539405822754,
"learning_rate": 2.5e-05,
"loss": 1.3577,
"step": 24
},
{
"epoch": 0.15723270440251572,
"grad_norm": 1.5717436075210571,
"learning_rate": 2.604166666666667e-05,
"loss": 1.1501,
"step": 25
},
{
"epoch": 0.16352201257861634,
"grad_norm": 1.691366195678711,
"learning_rate": 2.7083333333333332e-05,
"loss": 1.292,
"step": 26
},
{
"epoch": 0.16981132075471697,
"grad_norm": 1.4936853647232056,
"learning_rate": 2.8125000000000003e-05,
"loss": 1.1469,
"step": 27
},
{
"epoch": 0.1761006289308176,
"grad_norm": 1.2858690023422241,
"learning_rate": 2.916666666666667e-05,
"loss": 1.0433,
"step": 28
},
{
"epoch": 0.18238993710691823,
"grad_norm": 1.148773193359375,
"learning_rate": 3.0208333333333334e-05,
"loss": 0.9627,
"step": 29
},
{
"epoch": 0.18867924528301888,
"grad_norm": 1.0097932815551758,
"learning_rate": 3.125e-05,
"loss": 0.9103,
"step": 30
},
{
"epoch": 0.1949685534591195,
"grad_norm": 0.9332402348518372,
"learning_rate": 3.229166666666667e-05,
"loss": 0.979,
"step": 31
},
{
"epoch": 0.20125786163522014,
"grad_norm": 0.8993248343467712,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.9553,
"step": 32
},
{
"epoch": 0.20754716981132076,
"grad_norm": 0.7925082445144653,
"learning_rate": 3.4375e-05,
"loss": 0.8379,
"step": 33
},
{
"epoch": 0.2138364779874214,
"grad_norm": 0.7847669720649719,
"learning_rate": 3.541666666666667e-05,
"loss": 0.9038,
"step": 34
},
{
"epoch": 0.22012578616352202,
"grad_norm": 0.800736665725708,
"learning_rate": 3.6458333333333336e-05,
"loss": 0.8354,
"step": 35
},
{
"epoch": 0.22641509433962265,
"grad_norm": 0.6036879420280457,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.8517,
"step": 36
},
{
"epoch": 0.23270440251572327,
"grad_norm": 0.6031014919281006,
"learning_rate": 3.854166666666667e-05,
"loss": 0.8702,
"step": 37
},
{
"epoch": 0.2389937106918239,
"grad_norm": 0.584917426109314,
"learning_rate": 3.958333333333333e-05,
"loss": 0.8812,
"step": 38
},
{
"epoch": 0.24528301886792453,
"grad_norm": 0.5815232396125793,
"learning_rate": 4.0625000000000005e-05,
"loss": 0.8336,
"step": 39
},
{
"epoch": 0.25157232704402516,
"grad_norm": 0.5307980179786682,
"learning_rate": 4.166666666666667e-05,
"loss": 0.7162,
"step": 40
},
{
"epoch": 0.2578616352201258,
"grad_norm": 0.49309515953063965,
"learning_rate": 4.270833333333333e-05,
"loss": 0.7433,
"step": 41
},
{
"epoch": 0.2641509433962264,
"grad_norm": 0.5215100646018982,
"learning_rate": 4.375e-05,
"loss": 0.738,
"step": 42
},
{
"epoch": 0.27044025157232704,
"grad_norm": 0.5361147522926331,
"learning_rate": 4.4791666666666673e-05,
"loss": 0.7884,
"step": 43
},
{
"epoch": 0.27672955974842767,
"grad_norm": 0.5043961405754089,
"learning_rate": 4.5833333333333334e-05,
"loss": 0.814,
"step": 44
},
{
"epoch": 0.2830188679245283,
"grad_norm": 0.476026713848114,
"learning_rate": 4.6875e-05,
"loss": 0.8249,
"step": 45
},
{
"epoch": 0.2893081761006289,
"grad_norm": 0.5099363923072815,
"learning_rate": 4.791666666666667e-05,
"loss": 0.7205,
"step": 46
},
{
"epoch": 0.29559748427672955,
"grad_norm": 0.4583563208580017,
"learning_rate": 4.8958333333333335e-05,
"loss": 0.7646,
"step": 47
},
{
"epoch": 0.3018867924528302,
"grad_norm": 0.5203999876976013,
"learning_rate": 5e-05,
"loss": 0.6697,
"step": 48
},
{
"epoch": 0.3081761006289308,
"grad_norm": 0.5025014877319336,
"learning_rate": 4.9999329662935534e-05,
"loss": 0.8134,
"step": 49
},
{
"epoch": 0.31446540880503143,
"grad_norm": 0.5326013565063477,
"learning_rate": 4.999731868769027e-05,
"loss": 0.7665,
"step": 50
},
{
"epoch": 0.32075471698113206,
"grad_norm": 0.4914862811565399,
"learning_rate": 4.99939671821067e-05,
"loss": 0.7717,
"step": 51
},
{
"epoch": 0.3270440251572327,
"grad_norm": 0.567669689655304,
"learning_rate": 4.998927532591592e-05,
"loss": 0.677,
"step": 52
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.44353675842285156,
"learning_rate": 4.9983243370727914e-05,
"loss": 0.6829,
"step": 53
},
{
"epoch": 0.33962264150943394,
"grad_norm": 0.48537710309028625,
"learning_rate": 4.9975871640018154e-05,
"loss": 0.7817,
"step": 54
},
{
"epoch": 0.34591194968553457,
"grad_norm": 0.6286251544952393,
"learning_rate": 4.996716052911017e-05,
"loss": 0.6835,
"step": 55
},
{
"epoch": 0.3522012578616352,
"grad_norm": 0.44308704137802124,
"learning_rate": 4.99571105051544e-05,
"loss": 0.728,
"step": 56
},
{
"epoch": 0.3584905660377358,
"grad_norm": 0.4270545542240143,
"learning_rate": 4.994572210710315e-05,
"loss": 0.6448,
"step": 57
},
{
"epoch": 0.36477987421383645,
"grad_norm": 0.38012009859085083,
"learning_rate": 4.993299594568163e-05,
"loss": 0.675,
"step": 58
},
{
"epoch": 0.3710691823899371,
"grad_norm": 0.46653008460998535,
"learning_rate": 4.9918932703355256e-05,
"loss": 0.6821,
"step": 59
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.3612300753593445,
"learning_rate": 4.990353313429303e-05,
"loss": 0.7063,
"step": 60
},
{
"epoch": 0.3836477987421384,
"grad_norm": 0.5769712924957275,
"learning_rate": 4.988679806432712e-05,
"loss": 0.6941,
"step": 61
},
{
"epoch": 0.389937106918239,
"grad_norm": 0.431997150182724,
"learning_rate": 4.9868728390908526e-05,
"loss": 0.6544,
"step": 62
},
{
"epoch": 0.39622641509433965,
"grad_norm": 0.38126569986343384,
"learning_rate": 4.9849325083059e-05,
"loss": 0.6966,
"step": 63
},
{
"epoch": 0.4025157232704403,
"grad_norm": 0.40978091955184937,
"learning_rate": 4.982858918131906e-05,
"loss": 0.6866,
"step": 64
},
{
"epoch": 0.4088050314465409,
"grad_norm": 0.5318535566329956,
"learning_rate": 4.980652179769218e-05,
"loss": 0.7029,
"step": 65
},
{
"epoch": 0.41509433962264153,
"grad_norm": 0.5288330912590027,
"learning_rate": 4.978312411558518e-05,
"loss": 0.6553,
"step": 66
},
{
"epoch": 0.42138364779874216,
"grad_norm": 0.6496341824531555,
"learning_rate": 4.9758397389744734e-05,
"loss": 0.696,
"step": 67
},
{
"epoch": 0.4276729559748428,
"grad_norm": 0.5196901559829712,
"learning_rate": 4.973234294619011e-05,
"loss": 0.6676,
"step": 68
},
{
"epoch": 0.4339622641509434,
"grad_norm": 0.5431825518608093,
"learning_rate": 4.9704962182142044e-05,
"loss": 0.6337,
"step": 69
},
{
"epoch": 0.44025157232704404,
"grad_norm": 0.6981568932533264,
"learning_rate": 4.967625656594782e-05,
"loss": 0.7645,
"step": 70
},
{
"epoch": 0.44654088050314467,
"grad_norm": 0.4362429082393646,
"learning_rate": 4.9646227637002515e-05,
"loss": 0.647,
"step": 71
},
{
"epoch": 0.4528301886792453,
"grad_norm": 0.41873699426651,
"learning_rate": 4.961487700566646e-05,
"loss": 0.6342,
"step": 72
},
{
"epoch": 0.4591194968553459,
"grad_norm": 0.589474081993103,
"learning_rate": 4.958220635317886e-05,
"loss": 0.641,
"step": 73
},
{
"epoch": 0.46540880503144655,
"grad_norm": 0.5421208739280701,
"learning_rate": 4.9548217431567665e-05,
"loss": 0.8103,
"step": 74
},
{
"epoch": 0.4716981132075472,
"grad_norm": 0.48354727029800415,
"learning_rate": 4.95129120635556e-05,
"loss": 0.6205,
"step": 75
},
{
"epoch": 0.4779874213836478,
"grad_norm": 0.6121774911880493,
"learning_rate": 4.9476292142462374e-05,
"loss": 0.6678,
"step": 76
},
{
"epoch": 0.48427672955974843,
"grad_norm": 0.5440806150436401,
"learning_rate": 4.943835963210324e-05,
"loss": 0.6668,
"step": 77
},
{
"epoch": 0.49056603773584906,
"grad_norm": 0.40740126371383667,
"learning_rate": 4.939911656668361e-05,
"loss": 0.61,
"step": 78
},
{
"epoch": 0.4968553459119497,
"grad_norm": 0.5682631731033325,
"learning_rate": 4.9358565050689985e-05,
"loss": 0.6939,
"step": 79
},
{
"epoch": 0.5031446540880503,
"grad_norm": 0.43090444803237915,
"learning_rate": 4.93167072587771e-05,
"loss": 0.6698,
"step": 80
},
{
"epoch": 0.5094339622641509,
"grad_norm": 0.3859785795211792,
"learning_rate": 4.92735454356513e-05,
"loss": 0.5798,
"step": 81
},
{
"epoch": 0.5157232704402516,
"grad_norm": 0.4453634023666382,
"learning_rate": 4.922908189595018e-05,
"loss": 0.7936,
"step": 82
},
{
"epoch": 0.5220125786163522,
"grad_norm": 0.4291239380836487,
"learning_rate": 4.9183319024118415e-05,
"loss": 0.667,
"step": 83
},
{
"epoch": 0.5283018867924528,
"grad_norm": 0.44029664993286133,
"learning_rate": 4.9136259274279955e-05,
"loss": 0.6518,
"step": 84
},
{
"epoch": 0.5345911949685535,
"grad_norm": 0.44975465536117554,
"learning_rate": 4.908790517010636e-05,
"loss": 0.6105,
"step": 85
},
{
"epoch": 0.5408805031446541,
"grad_norm": 0.3954753577709198,
"learning_rate": 4.903825930468149e-05,
"loss": 0.6954,
"step": 86
},
{
"epoch": 0.5471698113207547,
"grad_norm": 0.3935467302799225,
"learning_rate": 4.898732434036244e-05,
"loss": 0.6614,
"step": 87
},
{
"epoch": 0.5534591194968553,
"grad_norm": 0.37667638063430786,
"learning_rate": 4.893510300863676e-05,
"loss": 0.6186,
"step": 88
},
{
"epoch": 0.559748427672956,
"grad_norm": 0.5198381543159485,
"learning_rate": 4.8881598109976004e-05,
"loss": 0.6153,
"step": 89
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.4063967764377594,
"learning_rate": 4.8826812513685487e-05,
"loss": 0.6472,
"step": 90
},
{
"epoch": 0.5723270440251572,
"grad_norm": 0.4114413857460022,
"learning_rate": 4.877074915775049e-05,
"loss": 0.6404,
"step": 91
},
{
"epoch": 0.5786163522012578,
"grad_norm": 0.37414512038230896,
"learning_rate": 4.8713411048678635e-05,
"loss": 0.6353,
"step": 92
},
{
"epoch": 0.5849056603773585,
"grad_norm": 0.3818528354167938,
"learning_rate": 4.865480126133872e-05,
"loss": 0.6289,
"step": 93
},
{
"epoch": 0.5911949685534591,
"grad_norm": 0.37088677287101746,
"learning_rate": 4.859492293879574e-05,
"loss": 0.6242,
"step": 94
},
{
"epoch": 0.5974842767295597,
"grad_norm": 0.43665146827697754,
"learning_rate": 4.853377929214243e-05,
"loss": 0.6055,
"step": 95
},
{
"epoch": 0.6037735849056604,
"grad_norm": 0.5932008028030396,
"learning_rate": 4.8471373600326996e-05,
"loss": 0.6354,
"step": 96
},
{
"epoch": 0.610062893081761,
"grad_norm": 0.46184536814689636,
"learning_rate": 4.8407709209977305e-05,
"loss": 0.7039,
"step": 97
},
{
"epoch": 0.6163522012578616,
"grad_norm": 0.44864600896835327,
"learning_rate": 4.834278953522138e-05,
"loss": 0.6679,
"step": 98
},
{
"epoch": 0.6226415094339622,
"grad_norm": 0.4060032069683075,
"learning_rate": 4.8276618057504376e-05,
"loss": 0.7479,
"step": 99
},
{
"epoch": 0.6289308176100629,
"grad_norm": 0.5088009834289551,
"learning_rate": 4.8209198325401815e-05,
"loss": 0.5923,
"step": 100
},
{
"epoch": 0.6352201257861635,
"grad_norm": 0.4294343590736389,
"learning_rate": 4.8140533954429327e-05,
"loss": 0.6313,
"step": 101
},
{
"epoch": 0.6415094339622641,
"grad_norm": 0.5710806846618652,
"learning_rate": 4.8070628626848735e-05,
"loss": 0.6736,
"step": 102
},
{
"epoch": 0.6477987421383647,
"grad_norm": 0.40560534596443176,
"learning_rate": 4.799948609147061e-05,
"loss": 0.6173,
"step": 103
},
{
"epoch": 0.6540880503144654,
"grad_norm": 0.46361806988716125,
"learning_rate": 4.792711016345321e-05,
"loss": 0.6756,
"step": 104
},
{
"epoch": 0.660377358490566,
"grad_norm": 0.5206082463264465,
"learning_rate": 4.785350472409792e-05,
"loss": 0.6413,
"step": 105
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.42583200335502625,
"learning_rate": 4.777867372064105e-05,
"loss": 0.6201,
"step": 106
},
{
"epoch": 0.6729559748427673,
"grad_norm": 0.49348241090774536,
"learning_rate": 4.770262116604224e-05,
"loss": 0.5777,
"step": 107
},
{
"epoch": 0.6792452830188679,
"grad_norm": 0.43750065565109253,
"learning_rate": 4.762535113876917e-05,
"loss": 0.6326,
"step": 108
},
{
"epoch": 0.6855345911949685,
"grad_norm": 0.431643545627594,
"learning_rate": 4.754686778257891e-05,
"loss": 0.5849,
"step": 109
},
{
"epoch": 0.6918238993710691,
"grad_norm": 0.41906824707984924,
"learning_rate": 4.7467175306295655e-05,
"loss": 0.6779,
"step": 110
},
{
"epoch": 0.6981132075471698,
"grad_norm": 0.4261593222618103,
"learning_rate": 4.738627798358506e-05,
"loss": 0.565,
"step": 111
},
{
"epoch": 0.7044025157232704,
"grad_norm": 0.42877814173698425,
"learning_rate": 4.730418015272503e-05,
"loss": 0.6836,
"step": 112
},
{
"epoch": 0.710691823899371,
"grad_norm": 0.40995511412620544,
"learning_rate": 4.722088621637309e-05,
"loss": 0.5849,
"step": 113
},
{
"epoch": 0.7169811320754716,
"grad_norm": 0.5606467723846436,
"learning_rate": 4.713640064133025e-05,
"loss": 0.644,
"step": 114
},
{
"epoch": 0.7232704402515723,
"grad_norm": 0.48454752564430237,
"learning_rate": 4.7050727958301506e-05,
"loss": 0.5418,
"step": 115
},
{
"epoch": 0.7295597484276729,
"grad_norm": 0.4099617898464203,
"learning_rate": 4.6963872761652835e-05,
"loss": 0.6524,
"step": 116
},
{
"epoch": 0.7358490566037735,
"grad_norm": 0.47192293405532837,
"learning_rate": 4.687583970916487e-05,
"loss": 0.6134,
"step": 117
},
{
"epoch": 0.7421383647798742,
"grad_norm": 0.4925817549228668,
"learning_rate": 4.678663352178301e-05,
"loss": 0.6517,
"step": 118
},
{
"epoch": 0.7484276729559748,
"grad_norm": 0.41517379879951477,
"learning_rate": 4.6696258983364385e-05,
"loss": 0.5763,
"step": 119
},
{
"epoch": 0.7547169811320755,
"grad_norm": 0.6227956414222717,
"learning_rate": 4.660472094042121e-05,
"loss": 0.7349,
"step": 120
},
{
"epoch": 0.7610062893081762,
"grad_norm": 0.39108285307884216,
"learning_rate": 4.651202430186092e-05,
"loss": 0.6548,
"step": 121
},
{
"epoch": 0.7672955974842768,
"grad_norm": 0.40210628509521484,
"learning_rate": 4.641817403872293e-05,
"loss": 0.6135,
"step": 122
},
{
"epoch": 0.7735849056603774,
"grad_norm": 0.42835289239883423,
"learning_rate": 4.6323175183912024e-05,
"loss": 0.5227,
"step": 123
},
{
"epoch": 0.779874213836478,
"grad_norm": 0.6215445399284363,
"learning_rate": 4.6227032831928484e-05,
"loss": 0.5265,
"step": 124
},
{
"epoch": 0.7861635220125787,
"grad_norm": 0.43423891067504883,
"learning_rate": 4.6129752138594874e-05,
"loss": 0.5972,
"step": 125
},
{
"epoch": 0.7924528301886793,
"grad_norm": 0.4693637788295746,
"learning_rate": 4.6031338320779534e-05,
"loss": 0.6758,
"step": 126
},
{
"epoch": 0.7987421383647799,
"grad_norm": 0.47571077942848206,
"learning_rate": 4.5931796656116846e-05,
"loss": 0.6419,
"step": 127
},
{
"epoch": 0.8050314465408805,
"grad_norm": 0.5828564167022705,
"learning_rate": 4.5831132482724195e-05,
"loss": 0.6952,
"step": 128
},
{
"epoch": 0.8113207547169812,
"grad_norm": 0.5325509309768677,
"learning_rate": 4.572935119891571e-05,
"loss": 0.6093,
"step": 129
},
{
"epoch": 0.8176100628930818,
"grad_norm": 0.4476761519908905,
"learning_rate": 4.5626458262912745e-05,
"loss": 0.5688,
"step": 130
},
{
"epoch": 0.8238993710691824,
"grad_norm": 0.36771783232688904,
"learning_rate": 4.552245919255117e-05,
"loss": 0.5778,
"step": 131
},
{
"epoch": 0.8301886792452831,
"grad_norm": 0.3695763647556305,
"learning_rate": 4.541735956498554e-05,
"loss": 0.517,
"step": 132
},
{
"epoch": 0.8364779874213837,
"grad_norm": 0.46155717968940735,
"learning_rate": 4.5311165016389916e-05,
"loss": 0.631,
"step": 133
},
{
"epoch": 0.8427672955974843,
"grad_norm": 0.5536386966705322,
"learning_rate": 4.5203881241655644e-05,
"loss": 0.6309,
"step": 134
},
{
"epoch": 0.8490566037735849,
"grad_norm": 0.42256519198417664,
"learning_rate": 4.509551399408598e-05,
"loss": 0.6754,
"step": 135
},
{
"epoch": 0.8553459119496856,
"grad_norm": 0.41634511947631836,
"learning_rate": 4.498606908508754e-05,
"loss": 0.6547,
"step": 136
},
{
"epoch": 0.8616352201257862,
"grad_norm": 0.38796696066856384,
"learning_rate": 4.487555238385862e-05,
"loss": 0.5693,
"step": 137
},
{
"epoch": 0.8679245283018868,
"grad_norm": 0.43052586913108826,
"learning_rate": 4.476396981707453e-05,
"loss": 0.6655,
"step": 138
},
{
"epoch": 0.8742138364779874,
"grad_norm": 0.44500571489334106,
"learning_rate": 4.465132736856969e-05,
"loss": 0.6563,
"step": 139
},
{
"epoch": 0.8805031446540881,
"grad_norm": 0.44693946838378906,
"learning_rate": 4.453763107901675e-05,
"loss": 0.5729,
"step": 140
},
{
"epoch": 0.8867924528301887,
"grad_norm": 0.3982055187225342,
"learning_rate": 4.442288704560268e-05,
"loss": 0.6425,
"step": 141
},
{
"epoch": 0.8930817610062893,
"grad_norm": 0.5706686973571777,
"learning_rate": 4.430710142170176e-05,
"loss": 0.5732,
"step": 142
},
{
"epoch": 0.89937106918239,
"grad_norm": 0.4493609070777893,
"learning_rate": 4.419028041654559e-05,
"loss": 0.6507,
"step": 143
},
{
"epoch": 0.9056603773584906,
"grad_norm": 0.3834284543991089,
"learning_rate": 4.4072430294890174e-05,
"loss": 0.6883,
"step": 144
},
{
"epoch": 0.9119496855345912,
"grad_norm": 0.4243861138820648,
"learning_rate": 4.3953557376679856e-05,
"loss": 0.5818,
"step": 145
},
{
"epoch": 0.9182389937106918,
"grad_norm": 0.42064470052719116,
"learning_rate": 4.383366803670849e-05,
"loss": 0.5963,
"step": 146
},
{
"epoch": 0.9245283018867925,
"grad_norm": 0.46670740842819214,
"learning_rate": 4.371276870427753e-05,
"loss": 0.65,
"step": 147
},
{
"epoch": 0.9308176100628931,
"grad_norm": 0.44945597648620605,
"learning_rate": 4.359086586285127e-05,
"loss": 0.6554,
"step": 148
},
{
"epoch": 0.9371069182389937,
"grad_norm": 0.48251762986183167,
"learning_rate": 4.346796604970912e-05,
"loss": 0.6681,
"step": 149
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.4577901065349579,
"learning_rate": 4.3344075855595104e-05,
"loss": 0.6108,
"step": 150
},
{
"epoch": 0.949685534591195,
"grad_norm": 0.4710671603679657,
"learning_rate": 4.321920192436433e-05,
"loss": 0.6177,
"step": 151
},
{
"epoch": 0.9559748427672956,
"grad_norm": 0.4275747835636139,
"learning_rate": 4.309335095262676e-05,
"loss": 0.5881,
"step": 152
},
{
"epoch": 0.9622641509433962,
"grad_norm": 0.4541204571723938,
"learning_rate": 4.296652968938807e-05,
"loss": 0.6142,
"step": 153
},
{
"epoch": 0.9685534591194969,
"grad_norm": 0.4179267883300781,
"learning_rate": 4.283874493568772e-05,
"loss": 0.6463,
"step": 154
},
{
"epoch": 0.9748427672955975,
"grad_norm": 0.42434728145599365,
"learning_rate": 4.271000354423426e-05,
"loss": 0.5873,
"step": 155
},
{
"epoch": 0.9811320754716981,
"grad_norm": 0.44202134013175964,
"learning_rate": 4.258031241903778e-05,
"loss": 0.6046,
"step": 156
},
{
"epoch": 0.9874213836477987,
"grad_norm": 0.47817060351371765,
"learning_rate": 4.2449678515039747e-05,
"loss": 0.5989,
"step": 157
},
{
"epoch": 0.9937106918238994,
"grad_norm": 0.43847018480300903,
"learning_rate": 4.231810883773999e-05,
"loss": 0.6156,
"step": 158
},
{
"epoch": 1.0,
"grad_norm": 0.48976290225982666,
"learning_rate": 4.218561044282099e-05,
"loss": 0.6177,
"step": 159
},
{
"epoch": 1.0062893081761006,
"grad_norm": 0.4213927984237671,
"learning_rate": 4.2052190435769554e-05,
"loss": 0.5881,
"step": 160
},
{
"epoch": 1.0125786163522013,
"grad_norm": 0.3923173248767853,
"learning_rate": 4.191785597149577e-05,
"loss": 0.5354,
"step": 161
},
{
"epoch": 1.0188679245283019,
"grad_norm": 0.41947099566459656,
"learning_rate": 4.178261425394926e-05,
"loss": 0.5745,
"step": 162
},
{
"epoch": 1.0251572327044025,
"grad_norm": 0.45812028646469116,
"learning_rate": 4.1646472535732895e-05,
"loss": 0.5836,
"step": 163
},
{
"epoch": 1.0314465408805031,
"grad_norm": 0.4519791007041931,
"learning_rate": 4.1509438117713866e-05,
"loss": 0.6118,
"step": 164
},
{
"epoch": 1.0377358490566038,
"grad_norm": 0.5628229379653931,
"learning_rate": 4.137151834863213e-05,
"loss": 0.5811,
"step": 165
},
{
"epoch": 1.0440251572327044,
"grad_norm": 0.3987003564834595,
"learning_rate": 4.123272062470633e-05,
"loss": 0.5179,
"step": 166
},
{
"epoch": 1.050314465408805,
"grad_norm": 0.4499664604663849,
"learning_rate": 4.109305238923718e-05,
"loss": 0.548,
"step": 167
},
{
"epoch": 1.0566037735849056,
"grad_norm": 0.4336049556732178,
"learning_rate": 4.095252113220827e-05,
"loss": 0.5659,
"step": 168
},
{
"epoch": 1.0628930817610063,
"grad_norm": 0.5409331321716309,
"learning_rate": 4.0811134389884433e-05,
"loss": 0.5522,
"step": 169
},
{
"epoch": 1.069182389937107,
"grad_norm": 0.41802865266799927,
"learning_rate": 4.066889974440757e-05,
"loss": 0.542,
"step": 170
},
{
"epoch": 1.0754716981132075,
"grad_norm": 0.48149824142456055,
"learning_rate": 4.0525824823390045e-05,
"loss": 0.6131,
"step": 171
},
{
"epoch": 1.0817610062893082,
"grad_norm": 0.5059295296669006,
"learning_rate": 4.038191729950569e-05,
"loss": 0.5823,
"step": 172
},
{
"epoch": 1.0880503144654088,
"grad_norm": 0.5310669541358948,
"learning_rate": 4.0237184890078245e-05,
"loss": 0.6066,
"step": 173
},
{
"epoch": 1.0943396226415094,
"grad_norm": 0.5470696091651917,
"learning_rate": 4.009163535666761e-05,
"loss": 0.6407,
"step": 174
},
{
"epoch": 1.10062893081761,
"grad_norm": 0.4407522678375244,
"learning_rate": 3.994527650465352e-05,
"loss": 0.573,
"step": 175
},
{
"epoch": 1.1069182389937107,
"grad_norm": 0.4072059094905853,
"learning_rate": 3.979811618281706e-05,
"loss": 0.6349,
"step": 176
},
{
"epoch": 1.1132075471698113,
"grad_norm": 0.6410146355628967,
"learning_rate": 3.9650162282919655e-05,
"loss": 0.6036,
"step": 177
},
{
"epoch": 1.119496855345912,
"grad_norm": 0.4664185047149658,
"learning_rate": 3.9501422739279956e-05,
"loss": 0.5855,
"step": 178
},
{
"epoch": 1.1257861635220126,
"grad_norm": 0.7204552888870239,
"learning_rate": 3.9351905528348285e-05,
"loss": 0.5719,
"step": 179
},
{
"epoch": 1.1320754716981132,
"grad_norm": 0.47248926758766174,
"learning_rate": 3.920161866827889e-05,
"loss": 0.5764,
"step": 180
},
{
"epoch": 1.1383647798742138,
"grad_norm": 0.64155113697052,
"learning_rate": 3.90505702185e-05,
"loss": 0.644,
"step": 181
},
{
"epoch": 1.1446540880503144,
"grad_norm": 0.5559830665588379,
"learning_rate": 3.889876827928156e-05,
"loss": 0.5883,
"step": 182
},
{
"epoch": 1.150943396226415,
"grad_norm": 0.46357718110084534,
"learning_rate": 3.874622099130087e-05,
"loss": 0.5942,
"step": 183
},
{
"epoch": 1.1572327044025157,
"grad_norm": 0.4527578055858612,
"learning_rate": 3.859293653520604e-05,
"loss": 0.6086,
"step": 184
},
{
"epoch": 1.1635220125786163,
"grad_norm": 0.45968765020370483,
"learning_rate": 3.843892313117724e-05,
"loss": 0.6105,
"step": 185
},
{
"epoch": 1.169811320754717,
"grad_norm": 0.5293806791305542,
"learning_rate": 3.8284189038485936e-05,
"loss": 0.5622,
"step": 186
},
{
"epoch": 1.1761006289308176,
"grad_norm": 0.4499124586582184,
"learning_rate": 3.812874255505191e-05,
"loss": 0.5451,
"step": 187
},
{
"epoch": 1.1823899371069182,
"grad_norm": 0.5140619277954102,
"learning_rate": 3.797259201699833e-05,
"loss": 0.5864,
"step": 188
},
{
"epoch": 1.1886792452830188,
"grad_norm": 0.40918025374412537,
"learning_rate": 3.781574579820464e-05,
"loss": 0.5602,
"step": 189
},
{
"epoch": 1.1949685534591195,
"grad_norm": 0.4417194128036499,
"learning_rate": 3.765821230985758e-05,
"loss": 0.5556,
"step": 190
},
{
"epoch": 1.20125786163522,
"grad_norm": 0.4697968363761902,
"learning_rate": 3.7500000000000003e-05,
"loss": 0.6491,
"step": 191
},
{
"epoch": 1.2075471698113207,
"grad_norm": 0.40187153220176697,
"learning_rate": 3.7341117353077966e-05,
"loss": 0.5797,
"step": 192
},
{
"epoch": 1.2138364779874213,
"grad_norm": 0.456836462020874,
"learning_rate": 3.718157288948563e-05,
"loss": 0.5613,
"step": 193
},
{
"epoch": 1.220125786163522,
"grad_norm": 0.45992621779441833,
"learning_rate": 3.702137516510838e-05,
"loss": 0.6301,
"step": 194
},
{
"epoch": 1.2264150943396226,
"grad_norm": 0.45000532269477844,
"learning_rate": 3.686053277086401e-05,
"loss": 0.5772,
"step": 195
},
{
"epoch": 1.2327044025157232,
"grad_norm": 0.4734075963497162,
"learning_rate": 3.669905433224199e-05,
"loss": 0.5452,
"step": 196
},
{
"epoch": 1.2389937106918238,
"grad_norm": 0.41561609506607056,
"learning_rate": 3.653694850884091e-05,
"loss": 0.5365,
"step": 197
},
{
"epoch": 1.2452830188679245,
"grad_norm": 0.469097763299942,
"learning_rate": 3.637422399390413e-05,
"loss": 0.5396,
"step": 198
},
{
"epoch": 1.251572327044025,
"grad_norm": 0.46705111861228943,
"learning_rate": 3.621088951385353e-05,
"loss": 0.6024,
"step": 199
},
{
"epoch": 1.2578616352201257,
"grad_norm": 0.5201407074928284,
"learning_rate": 3.604695382782159e-05,
"loss": 0.6905,
"step": 200
},
{
"epoch": 1.2641509433962264,
"grad_norm": 0.4639309346675873,
"learning_rate": 3.588242572718162e-05,
"loss": 0.5909,
"step": 201
},
{
"epoch": 1.270440251572327,
"grad_norm": 0.4756741225719452,
"learning_rate": 3.5717314035076355e-05,
"loss": 0.5745,
"step": 202
},
{
"epoch": 1.2767295597484276,
"grad_norm": 0.5243765115737915,
"learning_rate": 3.5551627605944745e-05,
"loss": 0.6439,
"step": 203
},
{
"epoch": 1.2830188679245282,
"grad_norm": 0.4495748281478882,
"learning_rate": 3.5385375325047166e-05,
"loss": 0.5611,
"step": 204
},
{
"epoch": 1.2893081761006289,
"grad_norm": 0.47336816787719727,
"learning_rate": 3.521856610798887e-05,
"loss": 0.5846,
"step": 205
},
{
"epoch": 1.2955974842767295,
"grad_norm": 0.4667642116546631,
"learning_rate": 3.505120890024195e-05,
"loss": 0.5982,
"step": 206
},
{
"epoch": 1.3018867924528301,
"grad_norm": 0.4477729797363281,
"learning_rate": 3.4883312676665536e-05,
"loss": 0.6265,
"step": 207
},
{
"epoch": 1.3081761006289307,
"grad_norm": 0.5078469514846802,
"learning_rate": 3.4714886441024574e-05,
"loss": 0.5161,
"step": 208
},
{
"epoch": 1.3144654088050314,
"grad_norm": 0.5072935819625854,
"learning_rate": 3.4545939225506934e-05,
"loss": 0.5408,
"step": 209
},
{
"epoch": 1.320754716981132,
"grad_norm": 0.5006597638130188,
"learning_rate": 3.437648009023905e-05,
"loss": 0.5715,
"step": 210
},
{
"epoch": 1.3270440251572326,
"grad_norm": 0.44206503033638,
"learning_rate": 3.420651812280006e-05,
"loss": 0.5694,
"step": 211
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.4964762032032013,
"learning_rate": 3.4036062437734484e-05,
"loss": 0.5203,
"step": 212
},
{
"epoch": 1.3396226415094339,
"grad_norm": 0.4342876076698303,
"learning_rate": 3.386512217606339e-05,
"loss": 0.5778,
"step": 213
},
{
"epoch": 1.3459119496855345,
"grad_norm": 0.46520864963531494,
"learning_rate": 3.369370650479425e-05,
"loss": 0.5757,
"step": 214
},
{
"epoch": 1.3522012578616351,
"grad_norm": 0.5179557204246521,
"learning_rate": 3.3521824616429285e-05,
"loss": 0.5348,
"step": 215
},
{
"epoch": 1.3584905660377358,
"grad_norm": 0.45922955870628357,
"learning_rate": 3.3349485728472535e-05,
"loss": 0.5982,
"step": 216
},
{
"epoch": 1.3647798742138364,
"grad_norm": 0.5132372975349426,
"learning_rate": 3.3176699082935545e-05,
"loss": 0.6131,
"step": 217
},
{
"epoch": 1.371069182389937,
"grad_norm": 0.4740884602069855,
"learning_rate": 3.300347394584172e-05,
"loss": 0.6194,
"step": 218
},
{
"epoch": 1.3773584905660377,
"grad_norm": 0.45677605271339417,
"learning_rate": 3.282981960672948e-05,
"loss": 0.5938,
"step": 219
},
{
"epoch": 1.3836477987421385,
"grad_norm": 0.4545290172100067,
"learning_rate": 3.265574537815398e-05,
"loss": 0.5634,
"step": 220
},
{
"epoch": 1.389937106918239,
"grad_norm": 0.4439068138599396,
"learning_rate": 3.248126059518785e-05,
"loss": 0.5543,
"step": 221
},
{
"epoch": 1.3962264150943398,
"grad_norm": 0.4973136782646179,
"learning_rate": 3.230637461492043e-05,
"loss": 0.549,
"step": 222
},
{
"epoch": 1.4025157232704402,
"grad_norm": 0.5018147826194763,
"learning_rate": 3.213109681595612e-05,
"loss": 0.5799,
"step": 223
},
{
"epoch": 1.408805031446541,
"grad_norm": 0.4906245768070221,
"learning_rate": 3.195543659791132e-05,
"loss": 0.5788,
"step": 224
},
{
"epoch": 1.4150943396226414,
"grad_norm": 0.5228400826454163,
"learning_rate": 3.177940338091043e-05,
"loss": 0.4883,
"step": 225
},
{
"epoch": 1.4213836477987423,
"grad_norm": 0.5177304148674011,
"learning_rate": 3.160300660508064e-05,
"loss": 0.5737,
"step": 226
},
{
"epoch": 1.4276729559748427,
"grad_norm": 0.4641788899898529,
"learning_rate": 3.14262557300457e-05,
"loss": 0.5416,
"step": 227
},
{
"epoch": 1.4339622641509435,
"grad_norm": 0.49331700801849365,
"learning_rate": 3.124916023441865e-05,
"loss": 0.5747,
"step": 228
},
{
"epoch": 1.440251572327044,
"grad_norm": 0.4671120345592499,
"learning_rate": 3.107172961529343e-05,
"loss": 0.508,
"step": 229
},
{
"epoch": 1.4465408805031448,
"grad_norm": 0.5811720490455627,
"learning_rate": 3.0893973387735687e-05,
"loss": 0.6929,
"step": 230
},
{
"epoch": 1.4528301886792452,
"grad_norm": 0.48253801465034485,
"learning_rate": 3.071590108427244e-05,
"loss": 0.6467,
"step": 231
},
{
"epoch": 1.459119496855346,
"grad_norm": 0.5044742226600647,
"learning_rate": 3.0537522254380905e-05,
"loss": 0.5067,
"step": 232
},
{
"epoch": 1.4654088050314464,
"grad_norm": 0.5015671849250793,
"learning_rate": 3.0358846463976372e-05,
"loss": 0.6052,
"step": 233
},
{
"epoch": 1.4716981132075473,
"grad_norm": 0.5023432374000549,
"learning_rate": 3.017988329489923e-05,
"loss": 0.5961,
"step": 234
},
{
"epoch": 1.4779874213836477,
"grad_norm": 0.5157131552696228,
"learning_rate": 3.0000642344401113e-05,
"loss": 0.6285,
"step": 235
},
{
"epoch": 1.4842767295597485,
"grad_norm": 0.5081720948219299,
"learning_rate": 2.9821133224630226e-05,
"loss": 0.4634,
"step": 236
},
{
"epoch": 1.490566037735849,
"grad_norm": 0.4940662086009979,
"learning_rate": 2.9641365562115887e-05,
"loss": 0.5288,
"step": 237
},
{
"epoch": 1.4968553459119498,
"grad_norm": 0.6005929708480835,
"learning_rate": 2.9461348997252265e-05,
"loss": 0.5748,
"step": 238
},
{
"epoch": 1.5031446540880502,
"grad_norm": 0.5340356230735779,
"learning_rate": 2.9281093183781403e-05,
"loss": 0.5944,
"step": 239
},
{
"epoch": 1.509433962264151,
"grad_norm": 0.5056872367858887,
"learning_rate": 2.9100607788275545e-05,
"loss": 0.5549,
"step": 240
},
{
"epoch": 1.5157232704402515,
"grad_norm": 0.5385233759880066,
"learning_rate": 2.891990248961871e-05,
"loss": 0.5345,
"step": 241
},
{
"epoch": 1.5220125786163523,
"grad_norm": 0.48144999146461487,
"learning_rate": 2.8738986978487625e-05,
"loss": 0.5839,
"step": 242
},
{
"epoch": 1.5283018867924527,
"grad_norm": 0.5069999098777771,
"learning_rate": 2.8557870956832132e-05,
"loss": 0.4931,
"step": 243
},
{
"epoch": 1.5345911949685536,
"grad_norm": 0.4660912752151489,
"learning_rate": 2.8376564137354795e-05,
"loss": 0.5598,
"step": 244
},
{
"epoch": 1.540880503144654,
"grad_norm": 0.5416334271430969,
"learning_rate": 2.8195076242990122e-05,
"loss": 0.5894,
"step": 245
},
{
"epoch": 1.5471698113207548,
"grad_norm": 0.49134501814842224,
"learning_rate": 2.8013417006383076e-05,
"loss": 0.6345,
"step": 246
},
{
"epoch": 1.5534591194968552,
"grad_norm": 0.5861812829971313,
"learning_rate": 2.783159616936723e-05,
"loss": 0.6519,
"step": 247
},
{
"epoch": 1.559748427672956,
"grad_norm": 0.47188055515289307,
"learning_rate": 2.764962348244228e-05,
"loss": 0.5658,
"step": 248
},
{
"epoch": 1.5660377358490565,
"grad_norm": 0.4838746190071106,
"learning_rate": 2.7467508704251137e-05,
"loss": 0.5934,
"step": 249
},
{
"epoch": 1.5723270440251573,
"grad_norm": 0.516981840133667,
"learning_rate": 2.7285261601056698e-05,
"loss": 0.5341,
"step": 250
},
{
"epoch": 1.5786163522012577,
"grad_norm": 0.4871889352798462,
"learning_rate": 2.7102891946217994e-05,
"loss": 0.5885,
"step": 251
},
{
"epoch": 1.5849056603773586,
"grad_norm": 0.47121965885162354,
"learning_rate": 2.6920409519666174e-05,
"loss": 0.5988,
"step": 252
},
{
"epoch": 1.591194968553459,
"grad_norm": 0.5421566367149353,
"learning_rate": 2.6737824107379948e-05,
"loss": 0.6635,
"step": 253
},
{
"epoch": 1.5974842767295598,
"grad_norm": 0.5449573993682861,
"learning_rate": 2.655514550086086e-05,
"loss": 0.5365,
"step": 254
},
{
"epoch": 1.6037735849056602,
"grad_norm": 0.5301252007484436,
"learning_rate": 2.637238349660819e-05,
"loss": 0.6854,
"step": 255
},
{
"epoch": 1.610062893081761,
"grad_norm": 0.41109421849250793,
"learning_rate": 2.6189547895593562e-05,
"loss": 0.5234,
"step": 256
},
{
"epoch": 1.6163522012578615,
"grad_norm": 0.5346055626869202,
"learning_rate": 2.600664850273538e-05,
"loss": 0.5962,
"step": 257
},
{
"epoch": 1.6226415094339623,
"grad_norm": 0.5180861949920654,
"learning_rate": 2.582369512637302e-05,
"loss": 0.5396,
"step": 258
},
{
"epoch": 1.6289308176100628,
"grad_norm": 0.49982312321662903,
"learning_rate": 2.564069757774082e-05,
"loss": 0.4889,
"step": 259
},
{
"epoch": 1.6352201257861636,
"grad_norm": 0.5226275324821472,
"learning_rate": 2.545766567044194e-05,
"loss": 0.5893,
"step": 260
},
{
"epoch": 1.641509433962264,
"grad_norm": 0.47939932346343994,
"learning_rate": 2.527460921992209e-05,
"loss": 0.566,
"step": 261
},
{
"epoch": 1.6477987421383649,
"grad_norm": 0.536264955997467,
"learning_rate": 2.509153804294318e-05,
"loss": 0.5598,
"step": 262
},
{
"epoch": 1.6540880503144653,
"grad_norm": 0.5297203660011292,
"learning_rate": 2.490846195705683e-05,
"loss": 0.552,
"step": 263
},
{
"epoch": 1.6603773584905661,
"grad_norm": 0.5188955664634705,
"learning_rate": 2.4725390780077908e-05,
"loss": 0.6233,
"step": 264
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.4677969217300415,
"learning_rate": 2.4542334329558077e-05,
"loss": 0.4968,
"step": 265
},
{
"epoch": 1.6729559748427674,
"grad_norm": 0.5716769099235535,
"learning_rate": 2.435930242225919e-05,
"loss": 0.5315,
"step": 266
},
{
"epoch": 1.6792452830188678,
"grad_norm": 0.5071538090705872,
"learning_rate": 2.4176304873626985e-05,
"loss": 0.5622,
"step": 267
},
{
"epoch": 1.6855345911949686,
"grad_norm": 0.523229718208313,
"learning_rate": 2.399335149726463e-05,
"loss": 0.5999,
"step": 268
},
{
"epoch": 1.691823899371069,
"grad_norm": 0.5254403948783875,
"learning_rate": 2.3810452104406444e-05,
"loss": 0.6351,
"step": 269
},
{
"epoch": 1.6981132075471699,
"grad_norm": 0.53160560131073,
"learning_rate": 2.3627616503391814e-05,
"loss": 0.5754,
"step": 270
},
{
"epoch": 1.7044025157232703,
"grad_norm": 0.5396912693977356,
"learning_rate": 2.3444854499139142e-05,
"loss": 0.5635,
"step": 271
},
{
"epoch": 1.7106918238993711,
"grad_norm": 0.5528764724731445,
"learning_rate": 2.3262175892620065e-05,
"loss": 0.5368,
"step": 272
},
{
"epoch": 1.7169811320754715,
"grad_norm": 0.4812920093536377,
"learning_rate": 2.307959048033383e-05,
"loss": 0.5508,
"step": 273
},
{
"epoch": 1.7232704402515724,
"grad_norm": 0.5467996001243591,
"learning_rate": 2.2897108053782e-05,
"loss": 0.6073,
"step": 274
},
{
"epoch": 1.7295597484276728,
"grad_norm": 0.5203157663345337,
"learning_rate": 2.271473839894331e-05,
"loss": 0.5326,
"step": 275
},
{
"epoch": 1.7358490566037736,
"grad_norm": 0.4970155358314514,
"learning_rate": 2.253249129574887e-05,
"loss": 0.5444,
"step": 276
},
{
"epoch": 1.742138364779874,
"grad_norm": 0.5279742479324341,
"learning_rate": 2.2350376517557727e-05,
"loss": 0.5778,
"step": 277
},
{
"epoch": 1.748427672955975,
"grad_norm": 0.5293897986412048,
"learning_rate": 2.216840383063277e-05,
"loss": 0.5786,
"step": 278
},
{
"epoch": 1.7547169811320755,
"grad_norm": 0.565350353717804,
"learning_rate": 2.1986582993616926e-05,
"loss": 0.5877,
"step": 279
},
{
"epoch": 1.7610062893081762,
"grad_norm": 0.47762683033943176,
"learning_rate": 2.1804923757009884e-05,
"loss": 0.5419,
"step": 280
},
{
"epoch": 1.7672955974842768,
"grad_norm": 0.5171676278114319,
"learning_rate": 2.1623435862645204e-05,
"loss": 0.5418,
"step": 281
},
{
"epoch": 1.7735849056603774,
"grad_norm": 0.45764029026031494,
"learning_rate": 2.1442129043167874e-05,
"loss": 0.5487,
"step": 282
},
{
"epoch": 1.779874213836478,
"grad_norm": 0.5272729396820068,
"learning_rate": 2.126101302151238e-05,
"loss": 0.5317,
"step": 283
},
{
"epoch": 1.7861635220125787,
"grad_norm": 0.49889740347862244,
"learning_rate": 2.1080097510381298e-05,
"loss": 0.6046,
"step": 284
},
{
"epoch": 1.7924528301886793,
"grad_norm": 0.5512136220932007,
"learning_rate": 2.089939221172446e-05,
"loss": 0.6541,
"step": 285
},
{
"epoch": 1.79874213836478,
"grad_norm": 0.5187536478042603,
"learning_rate": 2.07189068162186e-05,
"loss": 0.5646,
"step": 286
},
{
"epoch": 1.8050314465408805,
"grad_norm": 0.4920225143432617,
"learning_rate": 2.0538651002747744e-05,
"loss": 0.5561,
"step": 287
},
{
"epoch": 1.8113207547169812,
"grad_norm": 0.5807557702064514,
"learning_rate": 2.0358634437884112e-05,
"loss": 0.6087,
"step": 288
},
{
"epoch": 1.8176100628930818,
"grad_norm": 0.48779597878456116,
"learning_rate": 2.0178866775369777e-05,
"loss": 0.5801,
"step": 289
},
{
"epoch": 1.8238993710691824,
"grad_norm": 0.47587329149246216,
"learning_rate": 1.9999357655598893e-05,
"loss": 0.6003,
"step": 290
},
{
"epoch": 1.830188679245283,
"grad_norm": 0.5073649883270264,
"learning_rate": 1.9820116705100777e-05,
"loss": 0.5838,
"step": 291
},
{
"epoch": 1.8364779874213837,
"grad_norm": 0.5661832690238953,
"learning_rate": 1.9641153536023644e-05,
"loss": 0.6054,
"step": 292
},
{
"epoch": 1.8427672955974843,
"grad_norm": 0.4822792410850525,
"learning_rate": 1.9462477745619108e-05,
"loss": 0.5515,
"step": 293
},
{
"epoch": 1.849056603773585,
"grad_norm": 0.5438389182090759,
"learning_rate": 1.928409891572757e-05,
"loss": 0.6225,
"step": 294
},
{
"epoch": 1.8553459119496856,
"grad_norm": 0.4644525647163391,
"learning_rate": 1.9106026612264316e-05,
"loss": 0.6214,
"step": 295
},
{
"epoch": 1.8616352201257862,
"grad_norm": 0.477500855922699,
"learning_rate": 1.8928270384706584e-05,
"loss": 0.6164,
"step": 296
},
{
"epoch": 1.8679245283018868,
"grad_norm": 0.4944886267185211,
"learning_rate": 1.875083976558136e-05,
"loss": 0.5533,
"step": 297
},
{
"epoch": 1.8742138364779874,
"grad_norm": 0.5480766892433167,
"learning_rate": 1.8573744269954298e-05,
"loss": 0.6305,
"step": 298
},
{
"epoch": 1.880503144654088,
"grad_norm": 0.5053893327713013,
"learning_rate": 1.8396993394919372e-05,
"loss": 0.574,
"step": 299
},
{
"epoch": 1.8867924528301887,
"grad_norm": 0.559573233127594,
"learning_rate": 1.8220596619089576e-05,
"loss": 0.7138,
"step": 300
},
{
"epoch": 1.8930817610062893,
"grad_norm": 0.5412449836730957,
"learning_rate": 1.8044563402088684e-05,
"loss": 0.6108,
"step": 301
},
{
"epoch": 1.89937106918239,
"grad_norm": 0.49840930104255676,
"learning_rate": 1.7868903184043887e-05,
"loss": 0.546,
"step": 302
},
{
"epoch": 1.9056603773584906,
"grad_norm": 0.6073680520057678,
"learning_rate": 1.7693625385079577e-05,
"loss": 0.5597,
"step": 303
},
{
"epoch": 1.9119496855345912,
"grad_norm": 0.5452404022216797,
"learning_rate": 1.7518739404812155e-05,
"loss": 0.6107,
"step": 304
},
{
"epoch": 1.9182389937106918,
"grad_norm": 0.48530417680740356,
"learning_rate": 1.7344254621846016e-05,
"loss": 0.5671,
"step": 305
},
{
"epoch": 1.9245283018867925,
"grad_norm": 0.47276559472084045,
"learning_rate": 1.7170180393270532e-05,
"loss": 0.5503,
"step": 306
},
{
"epoch": 1.930817610062893,
"grad_norm": 0.4949149191379547,
"learning_rate": 1.699652605415828e-05,
"loss": 0.4958,
"step": 307
},
{
"epoch": 1.9371069182389937,
"grad_norm": 0.5216560363769531,
"learning_rate": 1.682330091706446e-05,
"loss": 0.58,
"step": 308
},
{
"epoch": 1.9433962264150944,
"grad_norm": 0.5286461710929871,
"learning_rate": 1.6650514271527468e-05,
"loss": 0.5293,
"step": 309
},
{
"epoch": 1.949685534591195,
"grad_norm": 0.4912795424461365,
"learning_rate": 1.647817538357072e-05,
"loss": 0.5666,
"step": 310
},
{
"epoch": 1.9559748427672956,
"grad_norm": 0.5797849297523499,
"learning_rate": 1.630629349520576e-05,
"loss": 0.5493,
"step": 311
},
{
"epoch": 1.9622641509433962,
"grad_norm": 0.5302536487579346,
"learning_rate": 1.613487782393661e-05,
"loss": 0.5516,
"step": 312
},
{
"epoch": 1.9685534591194969,
"grad_norm": 0.5543875694274902,
"learning_rate": 1.5963937562265525e-05,
"loss": 0.5358,
"step": 313
},
{
"epoch": 1.9748427672955975,
"grad_norm": 0.5673121809959412,
"learning_rate": 1.5793481877199946e-05,
"loss": 0.5332,
"step": 314
},
{
"epoch": 1.9811320754716981,
"grad_norm": 0.508794367313385,
"learning_rate": 1.5623519909760954e-05,
"loss": 0.5968,
"step": 315
},
{
"epoch": 1.9874213836477987,
"grad_norm": 0.5409150719642639,
"learning_rate": 1.5454060774493068e-05,
"loss": 0.6077,
"step": 316
},
{
"epoch": 1.9937106918238994,
"grad_norm": 0.5045109987258911,
"learning_rate": 1.528511355897543e-05,
"loss": 0.5503,
"step": 317
},
{
"epoch": 2.0,
"grad_norm": 0.6411352157592773,
"learning_rate": 1.5116687323334467e-05,
"loss": 0.6122,
"step": 318
},
{
"epoch": 2.006289308176101,
"grad_norm": 0.5026600360870361,
"learning_rate": 1.4948791099758052e-05,
"loss": 0.5369,
"step": 319
},
{
"epoch": 2.0125786163522013,
"grad_norm": 0.511720597743988,
"learning_rate": 1.4781433892011131e-05,
"loss": 0.5291,
"step": 320
},
{
"epoch": 2.018867924528302,
"grad_norm": 0.48618605732917786,
"learning_rate": 1.4614624674952842e-05,
"loss": 0.5378,
"step": 321
},
{
"epoch": 2.0251572327044025,
"grad_norm": 0.545242190361023,
"learning_rate": 1.4448372394055249e-05,
"loss": 0.5048,
"step": 322
},
{
"epoch": 2.0314465408805034,
"grad_norm": 0.5334264636039734,
"learning_rate": 1.4282685964923642e-05,
"loss": 0.5401,
"step": 323
},
{
"epoch": 2.0377358490566038,
"grad_norm": 0.5524183511734009,
"learning_rate": 1.4117574272818388e-05,
"loss": 0.5099,
"step": 324
},
{
"epoch": 2.0440251572327046,
"grad_norm": 0.48006343841552734,
"learning_rate": 1.3953046172178414e-05,
"loss": 0.4566,
"step": 325
},
{
"epoch": 2.050314465408805,
"grad_norm": 0.45758774876594543,
"learning_rate": 1.378911048614647e-05,
"loss": 0.4861,
"step": 326
},
{
"epoch": 2.056603773584906,
"grad_norm": 0.583585262298584,
"learning_rate": 1.3625776006095881e-05,
"loss": 0.5896,
"step": 327
},
{
"epoch": 2.0628930817610063,
"grad_norm": 0.5123708844184875,
"learning_rate": 1.3463051491159096e-05,
"loss": 0.5174,
"step": 328
},
{
"epoch": 2.069182389937107,
"grad_norm": 0.5508242845535278,
"learning_rate": 1.3300945667758014e-05,
"loss": 0.5042,
"step": 329
},
{
"epoch": 2.0754716981132075,
"grad_norm": 0.5526241660118103,
"learning_rate": 1.3139467229135999e-05,
"loss": 0.5457,
"step": 330
},
{
"epoch": 2.0817610062893084,
"grad_norm": 0.5903204083442688,
"learning_rate": 1.2978624834891628e-05,
"loss": 0.5299,
"step": 331
},
{
"epoch": 2.088050314465409,
"grad_norm": 0.5764947533607483,
"learning_rate": 1.281842711051438e-05,
"loss": 0.512,
"step": 332
},
{
"epoch": 2.0943396226415096,
"grad_norm": 0.5653167963027954,
"learning_rate": 1.2658882646922034e-05,
"loss": 0.5131,
"step": 333
},
{
"epoch": 2.10062893081761,
"grad_norm": 0.5843393206596375,
"learning_rate": 1.2500000000000006e-05,
"loss": 0.5383,
"step": 334
},
{
"epoch": 2.106918238993711,
"grad_norm": 0.6287145614624023,
"learning_rate": 1.2341787690142437e-05,
"loss": 0.5715,
"step": 335
},
{
"epoch": 2.1132075471698113,
"grad_norm": 0.5734768509864807,
"learning_rate": 1.2184254201795365e-05,
"loss": 0.548,
"step": 336
},
{
"epoch": 2.119496855345912,
"grad_norm": 0.5723345279693604,
"learning_rate": 1.2027407983001681e-05,
"loss": 0.4989,
"step": 337
},
{
"epoch": 2.1257861635220126,
"grad_norm": 0.6128413677215576,
"learning_rate": 1.1871257444948098e-05,
"loss": 0.5775,
"step": 338
},
{
"epoch": 2.1320754716981134,
"grad_norm": 0.7310327887535095,
"learning_rate": 1.1715810961514073e-05,
"loss": 0.5642,
"step": 339
},
{
"epoch": 2.138364779874214,
"grad_norm": 0.6413530707359314,
"learning_rate": 1.1561076868822756e-05,
"loss": 0.571,
"step": 340
},
{
"epoch": 2.1446540880503147,
"grad_norm": 0.6232360601425171,
"learning_rate": 1.1407063464793966e-05,
"loss": 0.4891,
"step": 341
},
{
"epoch": 2.150943396226415,
"grad_norm": 0.5760077238082886,
"learning_rate": 1.1253779008699131e-05,
"loss": 0.5196,
"step": 342
},
{
"epoch": 2.157232704402516,
"grad_norm": 0.5411508083343506,
"learning_rate": 1.1101231720718442e-05,
"loss": 0.479,
"step": 343
},
{
"epoch": 2.1635220125786163,
"grad_norm": 0.5932305455207825,
"learning_rate": 1.09494297815e-05,
"loss": 0.4888,
"step": 344
},
{
"epoch": 2.169811320754717,
"grad_norm": 0.5784667730331421,
"learning_rate": 1.0798381331721109e-05,
"loss": 0.5119,
"step": 345
},
{
"epoch": 2.1761006289308176,
"grad_norm": 0.663508415222168,
"learning_rate": 1.0648094471651724e-05,
"loss": 0.497,
"step": 346
},
{
"epoch": 2.1823899371069184,
"grad_norm": 0.5773563981056213,
"learning_rate": 1.049857726072005e-05,
"loss": 0.5013,
"step": 347
},
{
"epoch": 2.188679245283019,
"grad_norm": 0.697320282459259,
"learning_rate": 1.034983771708035e-05,
"loss": 0.6453,
"step": 348
},
{
"epoch": 2.1949685534591197,
"grad_norm": 0.5786930918693542,
"learning_rate": 1.020188381718295e-05,
"loss": 0.5044,
"step": 349
},
{
"epoch": 2.20125786163522,
"grad_norm": 0.6295438408851624,
"learning_rate": 1.0054723495346482e-05,
"loss": 0.594,
"step": 350
},
{
"epoch": 2.207547169811321,
"grad_norm": 0.6185433864593506,
"learning_rate": 9.908364643332399e-06,
"loss": 0.5059,
"step": 351
},
{
"epoch": 2.2138364779874213,
"grad_norm": 0.6029584407806396,
"learning_rate": 9.762815109921761e-06,
"loss": 0.5069,
"step": 352
},
{
"epoch": 2.220125786163522,
"grad_norm": 0.6533452272415161,
"learning_rate": 9.618082700494319e-06,
"loss": 0.5232,
"step": 353
},
{
"epoch": 2.2264150943396226,
"grad_norm": 0.6175459027290344,
"learning_rate": 9.474175176609956e-06,
"loss": 0.5466,
"step": 354
},
{
"epoch": 2.2327044025157234,
"grad_norm": 0.5657399892807007,
"learning_rate": 9.331100255592437e-06,
"loss": 0.4564,
"step": 355
},
{
"epoch": 2.238993710691824,
"grad_norm": 0.6779638528823853,
"learning_rate": 9.18886561011557e-06,
"loss": 0.5621,
"step": 356
},
{
"epoch": 2.2452830188679247,
"grad_norm": 0.5873716473579407,
"learning_rate": 9.047478867791732e-06,
"loss": 0.4575,
"step": 357
},
{
"epoch": 2.251572327044025,
"grad_norm": 0.5723773837089539,
"learning_rate": 8.906947610762825e-06,
"loss": 0.4867,
"step": 358
},
{
"epoch": 2.257861635220126,
"grad_norm": 0.6322399377822876,
"learning_rate": 8.767279375293672e-06,
"loss": 0.5453,
"step": 359
},
{
"epoch": 2.2641509433962264,
"grad_norm": 0.6071794033050537,
"learning_rate": 8.628481651367876e-06,
"loss": 0.5465,
"step": 360
},
{
"epoch": 2.270440251572327,
"grad_norm": 0.7954049110412598,
"learning_rate": 8.490561882286136e-06,
"loss": 0.5327,
"step": 361
},
{
"epoch": 2.2767295597484276,
"grad_norm": 0.5613401532173157,
"learning_rate": 8.353527464267104e-06,
"loss": 0.4759,
"step": 362
},
{
"epoch": 2.2830188679245285,
"grad_norm": 0.677384614944458,
"learning_rate": 8.217385746050742e-06,
"loss": 0.5186,
"step": 363
},
{
"epoch": 2.289308176100629,
"grad_norm": 0.6257275938987732,
"learning_rate": 8.082144028504233e-06,
"loss": 0.5786,
"step": 364
},
{
"epoch": 2.2955974842767297,
"grad_norm": 0.6544857025146484,
"learning_rate": 7.947809564230445e-06,
"loss": 0.4792,
"step": 365
},
{
"epoch": 2.30188679245283,
"grad_norm": 0.6507861614227295,
"learning_rate": 7.814389557179017e-06,
"loss": 0.5567,
"step": 366
},
{
"epoch": 2.308176100628931,
"grad_norm": 0.6414310336112976,
"learning_rate": 7.681891162260015e-06,
"loss": 0.5227,
"step": 367
},
{
"epoch": 2.3144654088050314,
"grad_norm": 0.6640331149101257,
"learning_rate": 7.5503214849602516e-06,
"loss": 0.5339,
"step": 368
},
{
"epoch": 2.3207547169811322,
"grad_norm": 0.6685804128646851,
"learning_rate": 7.419687580962223e-06,
"loss": 0.4857,
"step": 369
},
{
"epoch": 2.3270440251572326,
"grad_norm": 0.6468353271484375,
"learning_rate": 7.289996455765749e-06,
"loss": 0.4588,
"step": 370
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.5642632842063904,
"learning_rate": 7.161255064312283e-06,
"loss": 0.4676,
"step": 371
},
{
"epoch": 2.339622641509434,
"grad_norm": 0.583641529083252,
"learning_rate": 7.033470310611945e-06,
"loss": 0.5441,
"step": 372
},
{
"epoch": 2.3459119496855347,
"grad_norm": 0.6600673198699951,
"learning_rate": 6.906649047373246e-06,
"loss": 0.5144,
"step": 373
},
{
"epoch": 2.352201257861635,
"grad_norm": 0.6383036971092224,
"learning_rate": 6.780798075635675e-06,
"loss": 0.454,
"step": 374
},
{
"epoch": 2.358490566037736,
"grad_norm": 0.5476061105728149,
"learning_rate": 6.655924144404907e-06,
"loss": 0.4471,
"step": 375
},
{
"epoch": 2.3647798742138364,
"grad_norm": 0.6322336792945862,
"learning_rate": 6.532033950290886e-06,
"loss": 0.525,
"step": 376
},
{
"epoch": 2.3710691823899372,
"grad_norm": 0.6242497563362122,
"learning_rate": 6.409134137148737e-06,
"loss": 0.5181,
"step": 377
},
{
"epoch": 2.3773584905660377,
"grad_norm": 0.6529743075370789,
"learning_rate": 6.28723129572247e-06,
"loss": 0.5105,
"step": 378
},
{
"epoch": 2.3836477987421385,
"grad_norm": 0.6641661524772644,
"learning_rate": 6.166331963291519e-06,
"loss": 0.4859,
"step": 379
},
{
"epoch": 2.389937106918239,
"grad_norm": 0.6310985088348389,
"learning_rate": 6.046442623320145e-06,
"loss": 0.5136,
"step": 380
},
{
"epoch": 2.3962264150943398,
"grad_norm": 0.7812454104423523,
"learning_rate": 5.927569705109828e-06,
"loss": 0.4837,
"step": 381
},
{
"epoch": 2.40251572327044,
"grad_norm": 0.7382704019546509,
"learning_rate": 5.809719583454415e-06,
"loss": 0.4962,
"step": 382
},
{
"epoch": 2.408805031446541,
"grad_norm": 0.6249539256095886,
"learning_rate": 5.692898578298253e-06,
"loss": 0.5138,
"step": 383
},
{
"epoch": 2.4150943396226414,
"grad_norm": 0.6593284010887146,
"learning_rate": 5.577112954397321e-06,
"loss": 0.5097,
"step": 384
},
{
"epoch": 2.4213836477987423,
"grad_norm": 0.6821117997169495,
"learning_rate": 5.462368920983249e-06,
"loss": 0.5655,
"step": 385
},
{
"epoch": 2.4276729559748427,
"grad_norm": 0.7341055274009705,
"learning_rate": 5.348672631430318e-06,
"loss": 0.511,
"step": 386
},
{
"epoch": 2.4339622641509435,
"grad_norm": 0.676941454410553,
"learning_rate": 5.236030182925475e-06,
"loss": 0.4494,
"step": 387
},
{
"epoch": 2.440251572327044,
"grad_norm": 0.6467084288597107,
"learning_rate": 5.124447616141381e-06,
"loss": 0.4407,
"step": 388
},
{
"epoch": 2.4465408805031448,
"grad_norm": 0.6168628931045532,
"learning_rate": 5.013930914912476e-06,
"loss": 0.5424,
"step": 389
},
{
"epoch": 2.452830188679245,
"grad_norm": 0.6204650402069092,
"learning_rate": 4.9044860059140275e-06,
"loss": 0.5222,
"step": 390
},
{
"epoch": 2.459119496855346,
"grad_norm": 0.6808157563209534,
"learning_rate": 4.796118758344354e-06,
"loss": 0.5181,
"step": 391
},
{
"epoch": 2.4654088050314464,
"grad_norm": 0.6607468128204346,
"learning_rate": 4.6888349836100825e-06,
"loss": 0.5849,
"step": 392
},
{
"epoch": 2.4716981132075473,
"grad_norm": 0.6367841362953186,
"learning_rate": 4.58264043501446e-06,
"loss": 0.4938,
"step": 393
},
{
"epoch": 2.4779874213836477,
"grad_norm": 0.7117082476615906,
"learning_rate": 4.477540807448832e-06,
"loss": 0.4862,
"step": 394
},
{
"epoch": 2.4842767295597485,
"grad_norm": 0.662056565284729,
"learning_rate": 4.373541737087264e-06,
"loss": 0.5753,
"step": 395
},
{
"epoch": 2.490566037735849,
"grad_norm": 0.6686725616455078,
"learning_rate": 4.270648801084296e-06,
"loss": 0.555,
"step": 396
},
{
"epoch": 2.49685534591195,
"grad_norm": 0.7704941034317017,
"learning_rate": 4.1688675172758064e-06,
"loss": 0.6513,
"step": 397
},
{
"epoch": 2.50314465408805,
"grad_norm": 0.6105798482894897,
"learning_rate": 4.068203343883159e-06,
"loss": 0.4985,
"step": 398
},
{
"epoch": 2.509433962264151,
"grad_norm": 0.6414272785186768,
"learning_rate": 3.968661679220468e-06,
"loss": 0.4736,
"step": 399
},
{
"epoch": 2.5157232704402515,
"grad_norm": 0.6939327120780945,
"learning_rate": 3.8702478614051355e-06,
"loss": 0.567,
"step": 400
},
{
"epoch": 2.5220125786163523,
"grad_norm": 0.6539825797080994,
"learning_rate": 3.772967168071517e-06,
"loss": 0.4567,
"step": 401
},
{
"epoch": 2.5283018867924527,
"grad_norm": 0.6413926482200623,
"learning_rate": 3.6768248160879787e-06,
"loss": 0.5228,
"step": 402
},
{
"epoch": 2.5345911949685536,
"grad_norm": 0.6808522939682007,
"learning_rate": 3.5818259612770744e-06,
"loss": 0.5499,
"step": 403
},
{
"epoch": 2.540880503144654,
"grad_norm": 0.704112708568573,
"learning_rate": 3.487975698139084e-06,
"loss": 0.4901,
"step": 404
},
{
"epoch": 2.547169811320755,
"grad_norm": 0.7583079934120178,
"learning_rate": 3.3952790595787987e-06,
"loss": 0.5253,
"step": 405
},
{
"epoch": 2.5534591194968552,
"grad_norm": 0.6490894556045532,
"learning_rate": 3.3037410166356143e-06,
"loss": 0.5036,
"step": 406
},
{
"epoch": 2.559748427672956,
"grad_norm": 0.6950412392616272,
"learning_rate": 3.2133664782169948e-06,
"loss": 0.5155,
"step": 407
},
{
"epoch": 2.5660377358490565,
"grad_norm": 0.6417565941810608,
"learning_rate": 3.1241602908351404e-06,
"loss": 0.4473,
"step": 408
},
{
"epoch": 2.5723270440251573,
"grad_norm": 0.7289451360702515,
"learning_rate": 3.036127238347164e-06,
"loss": 0.5926,
"step": 409
},
{
"epoch": 2.5786163522012577,
"grad_norm": 0.6481583118438721,
"learning_rate": 2.9492720416985e-06,
"loss": 0.4606,
"step": 410
},
{
"epoch": 2.5849056603773586,
"grad_norm": 0.6715993881225586,
"learning_rate": 2.8635993586697553e-06,
"loss": 0.5701,
"step": 411
},
{
"epoch": 2.591194968553459,
"grad_norm": 0.6455919146537781,
"learning_rate": 2.779113783626916e-06,
"loss": 0.506,
"step": 412
},
{
"epoch": 2.59748427672956,
"grad_norm": 0.7350935935974121,
"learning_rate": 2.6958198472749717e-06,
"loss": 0.5624,
"step": 413
},
{
"epoch": 2.6037735849056602,
"grad_norm": 0.7309563755989075,
"learning_rate": 2.6137220164149435e-06,
"loss": 0.5052,
"step": 414
},
{
"epoch": 2.610062893081761,
"grad_norm": 0.6338467597961426,
"learning_rate": 2.5328246937043526e-06,
"loss": 0.5104,
"step": 415
},
{
"epoch": 2.6163522012578615,
"grad_norm": 0.6638092398643494,
"learning_rate": 2.4531322174210975e-06,
"loss": 0.5077,
"step": 416
},
{
"epoch": 2.6226415094339623,
"grad_norm": 0.6149672269821167,
"learning_rate": 2.37464886123083e-06,
"loss": 0.5086,
"step": 417
},
{
"epoch": 2.6289308176100628,
"grad_norm": 0.676372766494751,
"learning_rate": 2.2973788339577613e-06,
"loss": 0.5189,
"step": 418
},
{
"epoch": 2.6352201257861636,
"grad_norm": 0.6536609530448914,
"learning_rate": 2.2213262793589484e-06,
"loss": 0.4974,
"step": 419
},
{
"epoch": 2.641509433962264,
"grad_norm": 0.6942263245582581,
"learning_rate": 2.1464952759020855e-06,
"loss": 0.5648,
"step": 420
},
{
"epoch": 2.647798742138365,
"grad_norm": 0.642825722694397,
"learning_rate": 2.0728898365467903e-06,
"loss": 0.5149,
"step": 421
},
{
"epoch": 2.6540880503144653,
"grad_norm": 0.6079385876655579,
"learning_rate": 2.0005139085293945e-06,
"loss": 0.472,
"step": 422
},
{
"epoch": 2.660377358490566,
"grad_norm": 0.715664267539978,
"learning_rate": 1.9293713731512673e-06,
"loss": 0.5368,
"step": 423
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.655768871307373,
"learning_rate": 1.8594660455706763e-06,
"loss": 0.4948,
"step": 424
},
{
"epoch": 2.6729559748427674,
"grad_norm": 0.6798547506332397,
"learning_rate": 1.790801674598186e-06,
"loss": 0.5213,
"step": 425
},
{
"epoch": 2.6792452830188678,
"grad_norm": 0.7034783959388733,
"learning_rate": 1.7233819424956248e-06,
"loss": 0.6128,
"step": 426
},
{
"epoch": 2.6855345911949686,
"grad_norm": 0.6610020995140076,
"learning_rate": 1.6572104647786247e-06,
"loss": 0.5228,
"step": 427
},
{
"epoch": 2.691823899371069,
"grad_norm": 0.664659857749939,
"learning_rate": 1.5922907900227018e-06,
"loss": 0.4266,
"step": 428
},
{
"epoch": 2.69811320754717,
"grad_norm": 0.6232119202613831,
"learning_rate": 1.5286263996730026e-06,
"loss": 0.476,
"step": 429
},
{
"epoch": 2.7044025157232703,
"grad_norm": 0.653411328792572,
"learning_rate": 1.4662207078575684e-06,
"loss": 0.5145,
"step": 430
},
{
"epoch": 2.710691823899371,
"grad_norm": 0.7129037380218506,
"learning_rate": 1.40507706120426e-06,
"loss": 0.5546,
"step": 431
},
{
"epoch": 2.7169811320754715,
"grad_norm": 0.6307116150856018,
"learning_rate": 1.3451987386612851e-06,
"loss": 0.4598,
"step": 432
},
{
"epoch": 2.7232704402515724,
"grad_norm": 0.6527603268623352,
"learning_rate": 1.286588951321363e-06,
"loss": 0.5322,
"step": 433
},
{
"epoch": 2.729559748427673,
"grad_norm": 0.6609143018722534,
"learning_rate": 1.2292508422495158e-06,
"loss": 0.5306,
"step": 434
},
{
"epoch": 2.7358490566037736,
"grad_norm": 0.6827059388160706,
"learning_rate": 1.1731874863145143e-06,
"loss": 0.5109,
"step": 435
},
{
"epoch": 2.742138364779874,
"grad_norm": 0.6002165079116821,
"learning_rate": 1.1184018900240011e-06,
"loss": 0.4579,
"step": 436
},
{
"epoch": 2.748427672955975,
"grad_norm": 0.6269481182098389,
"learning_rate": 1.06489699136324e-06,
"loss": 0.4888,
"step": 437
},
{
"epoch": 2.7547169811320753,
"grad_norm": 0.6706013679504395,
"learning_rate": 1.0126756596375686e-06,
"loss": 0.5225,
"step": 438
},
{
"epoch": 2.761006289308176,
"grad_norm": 0.6613484025001526,
"learning_rate": 9.617406953185138e-07,
"loss": 0.5241,
"step": 439
},
{
"epoch": 2.767295597484277,
"grad_norm": 0.7121806740760803,
"learning_rate": 9.120948298936421e-07,
"loss": 0.4887,
"step": 440
},
{
"epoch": 2.7735849056603774,
"grad_norm": 0.6842261552810669,
"learning_rate": 8.637407257200497e-07,
"loss": 0.4912,
"step": 441
},
{
"epoch": 2.779874213836478,
"grad_norm": 0.70417320728302,
"learning_rate": 8.166809758815896e-07,
"loss": 0.5405,
"step": 442
},
{
"epoch": 2.7861635220125787,
"grad_norm": 0.6980416178703308,
"learning_rate": 7.709181040498254e-07,
"loss": 0.5058,
"step": 443
},
{
"epoch": 2.7924528301886795,
"grad_norm": 0.6587446331977844,
"learning_rate": 7.264545643486997e-07,
"loss": 0.5572,
"step": 444
},
{
"epoch": 2.79874213836478,
"grad_norm": 0.6442897319793701,
"learning_rate": 6.832927412229018e-07,
"loss": 0.5118,
"step": 445
},
{
"epoch": 2.8050314465408803,
"grad_norm": 0.6052460670471191,
"learning_rate": 6.41434949310013e-07,
"loss": 0.536,
"step": 446
},
{
"epoch": 2.811320754716981,
"grad_norm": 0.7546602487564087,
"learning_rate": 6.008834333163876e-07,
"loss": 0.5397,
"step": 447
},
{
"epoch": 2.817610062893082,
"grad_norm": 0.6202664375305176,
"learning_rate": 5.616403678967624e-07,
"loss": 0.4516,
"step": 448
},
{
"epoch": 2.8238993710691824,
"grad_norm": 0.6362593173980713,
"learning_rate": 5.237078575376336e-07,
"loss": 0.5106,
"step": 449
},
{
"epoch": 2.830188679245283,
"grad_norm": 0.6845952272415161,
"learning_rate": 4.870879364444109e-07,
"loss": 0.5153,
"step": 450
},
{
"epoch": 2.8364779874213837,
"grad_norm": 0.659001886844635,
"learning_rate": 4.517825684323324e-07,
"loss": 0.5075,
"step": 451
},
{
"epoch": 2.8427672955974845,
"grad_norm": 0.6748611927032471,
"learning_rate": 4.1779364682113796e-07,
"loss": 0.5354,
"step": 452
},
{
"epoch": 2.849056603773585,
"grad_norm": 0.6511921286582947,
"learning_rate": 3.851229943335394e-07,
"loss": 0.5468,
"step": 453
},
{
"epoch": 2.8553459119496853,
"grad_norm": 0.6114699244499207,
"learning_rate": 3.537723629974815e-07,
"loss": 0.4799,
"step": 454
},
{
"epoch": 2.861635220125786,
"grad_norm": 0.717573344707489,
"learning_rate": 3.237434340521789e-07,
"loss": 0.5604,
"step": 455
},
{
"epoch": 2.867924528301887,
"grad_norm": 0.6609371900558472,
"learning_rate": 2.9503781785795713e-07,
"loss": 0.4958,
"step": 456
},
{
"epoch": 2.8742138364779874,
"grad_norm": 0.7015553712844849,
"learning_rate": 2.6765705380989437e-07,
"loss": 0.551,
"step": 457
},
{
"epoch": 2.880503144654088,
"grad_norm": 0.6695305109024048,
"learning_rate": 2.416026102552732e-07,
"loss": 0.5316,
"step": 458
},
{
"epoch": 2.8867924528301887,
"grad_norm": 0.6004621386528015,
"learning_rate": 2.168758844148272e-07,
"loss": 0.5416,
"step": 459
},
{
"epoch": 2.8930817610062896,
"grad_norm": 0.6895103454589844,
"learning_rate": 1.93478202307823e-07,
"loss": 0.5319,
"step": 460
},
{
"epoch": 2.89937106918239,
"grad_norm": 0.6955282092094421,
"learning_rate": 1.7141081868094212e-07,
"loss": 0.5208,
"step": 461
},
{
"epoch": 2.9056603773584904,
"grad_norm": 0.6640160083770752,
"learning_rate": 1.5067491694100154e-07,
"loss": 0.5133,
"step": 462
},
{
"epoch": 2.911949685534591,
"grad_norm": 0.6678076982498169,
"learning_rate": 1.3127160909147672e-07,
"loss": 0.5107,
"step": 463
},
{
"epoch": 2.918238993710692,
"grad_norm": 0.7161384224891663,
"learning_rate": 1.1320193567288529e-07,
"loss": 0.5842,
"step": 464
},
{
"epoch": 2.9245283018867925,
"grad_norm": 0.712955117225647,
"learning_rate": 9.646686570697061e-08,
"loss": 0.5275,
"step": 465
},
{
"epoch": 2.930817610062893,
"grad_norm": 0.711218535900116,
"learning_rate": 8.106729664475176e-08,
"loss": 0.4815,
"step": 466
},
{
"epoch": 2.9371069182389937,
"grad_norm": 0.6556484699249268,
"learning_rate": 6.700405431837587e-08,
"loss": 0.4802,
"step": 467
},
{
"epoch": 2.9433962264150946,
"grad_norm": 0.6334600448608398,
"learning_rate": 5.4277892896853476e-08,
"loss": 0.5044,
"step": 468
},
{
"epoch": 2.949685534591195,
"grad_norm": 0.6841341257095337,
"learning_rate": 4.2889494845599344e-08,
"loss": 0.5601,
"step": 469
},
{
"epoch": 2.9559748427672954,
"grad_norm": 0.6439110040664673,
"learning_rate": 3.283947088983663e-08,
"loss": 0.4279,
"step": 470
},
{
"epoch": 2.9622641509433962,
"grad_norm": 0.593614935874939,
"learning_rate": 2.4128359981850924e-08,
"loss": 0.483,
"step": 471
},
{
"epoch": 2.968553459119497,
"grad_norm": 0.6519297957420349,
"learning_rate": 1.6756629272085545e-08,
"loss": 0.5921,
"step": 472
},
{
"epoch": 2.9748427672955975,
"grad_norm": 0.6468663215637207,
"learning_rate": 1.0724674084083841e-08,
"loss": 0.4751,
"step": 473
},
{
"epoch": 2.981132075471698,
"grad_norm": 0.6676617860794067,
"learning_rate": 6.032817893297793e-09,
"loss": 0.4994,
"step": 474
},
{
"epoch": 2.9874213836477987,
"grad_norm": 0.6803402900695801,
"learning_rate": 2.681312309735229e-09,
"loss": 0.5308,
"step": 475
},
{
"epoch": 2.9937106918238996,
"grad_norm": 0.6511867642402649,
"learning_rate": 6.703370644706164e-10,
"loss": 0.5177,
"step": 476
},
{
"epoch": 3.0,
"grad_norm": 0.7002833485603333,
"learning_rate": 0.0,
"loss": 0.4149,
"step": 477
},
{
"epoch": 3.0,
"step": 477,
"total_flos": 4.746880321608352e+18,
"train_loss": 0.7064950698576633,
"train_runtime": 33423.533,
"train_samples_per_second": 1.825,
"train_steps_per_second": 0.014
}
],
"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": 4.746880321608352e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}