falcon-40b-ft-alpaca-dolly-dutch / trainer_state.json
Bram Vanroy
init adapters
517adfd
{
"best_metric": 0.9153895974159241,
"best_model_checkpoint": "/dodrio/scratch/projects/2023_005/llm-finetuning/instruct-tuning/results/falcon-40b-ft-alpaca-dolly-dutch/2e-4lr+512tbs+5ep+32a+4r-v2/checkpoint-240",
"epoch": 2.5963488843813387,
"global_step": 320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 0.0,
"loss": 1.3493,
"step": 1
},
{
"epoch": 0.02,
"learning_rate": 2.7667037805630792e-05,
"loss": 1.1896,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 4.3851217427959356e-05,
"loss": 1.1459,
"step": 3
},
{
"epoch": 0.03,
"learning_rate": 5.5334075611261585e-05,
"loss": 1.2258,
"step": 4
},
{
"epoch": 0.04,
"learning_rate": 6.424087238320494e-05,
"loss": 1.246,
"step": 5
},
{
"epoch": 0.05,
"learning_rate": 7.151825523359015e-05,
"loss": 1.1355,
"step": 6
},
{
"epoch": 0.06,
"learning_rate": 7.767119476239142e-05,
"loss": 1.2275,
"step": 7
},
{
"epoch": 0.06,
"learning_rate": 8.300111341689237e-05,
"loss": 1.137,
"step": 8
},
{
"epoch": 0.07,
"learning_rate": 8.770243485591871e-05,
"loss": 1.103,
"step": 9
},
{
"epoch": 0.08,
"learning_rate": 9.190791018883575e-05,
"loss": 1.1777,
"step": 10
},
{
"epoch": 0.09,
"learning_rate": 9.571222537883264e-05,
"loss": 1.1354,
"step": 11
},
{
"epoch": 0.1,
"learning_rate": 9.918529303922093e-05,
"loss": 1.0573,
"step": 12
},
{
"epoch": 0.11,
"learning_rate": 0.00010238020557926735,
"loss": 1.1057,
"step": 13
},
{
"epoch": 0.11,
"learning_rate": 0.0001053382325680222,
"loss": 1.1141,
"step": 14
},
{
"epoch": 0.12,
"learning_rate": 0.0001080920898111643,
"loss": 1.0665,
"step": 15
},
{
"epoch": 0.13,
"learning_rate": 0.00011066815122252317,
"loss": 1.1191,
"step": 16
},
{
"epoch": 0.14,
"learning_rate": 0.0001130879889579842,
"loss": 1.1608,
"step": 17
},
{
"epoch": 0.15,
"learning_rate": 0.0001153694726615495,
"loss": 1.0334,
"step": 18
},
{
"epoch": 0.15,
"learning_rate": 0.00011752757111002288,
"loss": 1.0872,
"step": 19
},
{
"epoch": 0.16,
"learning_rate": 0.00011957494799446651,
"loss": 1.1656,
"step": 20
},
{
"epoch": 0.16,
"eval_loss": 1.0107293128967285,
"eval_runtime": 104.5604,
"eval_samples_per_second": 31.867,
"eval_steps_per_second": 0.258,
"step": 20
},
{
"epoch": 0.17,
"learning_rate": 0.00012152241219035079,
"loss": 1.0628,
"step": 21
},
{
"epoch": 0.18,
"learning_rate": 0.00012337926318446344,
"loss": 1.0393,
"step": 22
},
{
"epoch": 0.19,
"learning_rate": 0.00012515355965434256,
"loss": 1.1608,
"step": 23
},
{
"epoch": 0.19,
"learning_rate": 0.00012685233084485173,
"loss": 1.0166,
"step": 24
},
{
"epoch": 0.2,
"learning_rate": 0.00012848174476640987,
"loss": 1.0257,
"step": 25
},
{
"epoch": 0.21,
"learning_rate": 0.00013004724338489814,
"loss": 1.1549,
"step": 26
},
{
"epoch": 0.22,
"learning_rate": 0.00013155365228387805,
"loss": 1.0424,
"step": 27
},
{
"epoch": 0.23,
"learning_rate": 0.000133005270373653,
"loss": 1.0365,
"step": 28
},
{
"epoch": 0.24,
"learning_rate": 0.00013440594385123045,
"loss": 1.1665,
"step": 29
},
{
"epoch": 0.24,
"learning_rate": 0.0001357591276167951,
"loss": 1.0571,
"step": 30
},
{
"epoch": 0.25,
"learning_rate": 0.00013706793661599027,
"loss": 0.9924,
"step": 31
},
{
"epoch": 0.26,
"learning_rate": 0.00013833518902815398,
"loss": 1.1116,
"step": 32
},
{
"epoch": 0.27,
"learning_rate": 0.00013956344280679199,
"loss": 1.0776,
"step": 33
},
{
"epoch": 0.28,
"learning_rate": 0.000140755026763615,
"loss": 0.9747,
"step": 34
},
{
"epoch": 0.28,
"learning_rate": 0.00014191206714559635,
"loss": 1.0629,
"step": 35
},
{
"epoch": 0.29,
"learning_rate": 0.0001430365104671803,
"loss": 1.092,
"step": 36
},
{
"epoch": 0.3,
"learning_rate": 0.00014413014321352675,
"loss": 0.9619,
"step": 37
},
{
"epoch": 0.31,
"learning_rate": 0.00014519460891565368,
"loss": 1.0927,
"step": 38
},
{
"epoch": 0.32,
"learning_rate": 0.0001462314230072267,
"loss": 1.0615,
"step": 39
},
{
"epoch": 0.32,
"learning_rate": 0.0001472419858000973,
"loss": 0.9778,
"step": 40
},
{
"epoch": 0.32,
"eval_loss": 0.9711087346076965,
"eval_runtime": 105.6416,
"eval_samples_per_second": 31.541,
"eval_steps_per_second": 0.256,
"step": 40
},
{
"epoch": 0.33,
"learning_rate": 0.00014822759385740156,
"loss": 1.0388,
"step": 41
},
{
"epoch": 0.34,
"learning_rate": 0.00014918944999598157,
"loss": 1.1334,
"step": 42
},
{
"epoch": 0.35,
"learning_rate": 0.00015012867211170486,
"loss": 0.9814,
"step": 43
},
{
"epoch": 0.36,
"learning_rate": 0.00015104630099009423,
"loss": 1.0085,
"step": 44
},
{
"epoch": 0.37,
"learning_rate": 0.00015194330723912363,
"loss": 1.067,
"step": 45
},
{
"epoch": 0.37,
"learning_rate": 0.00015282059745997336,
"loss": 1.0041,
"step": 46
},
{
"epoch": 0.38,
"learning_rate": 0.00015367901975410053,
"loss": 0.957,
"step": 47
},
{
"epoch": 0.39,
"learning_rate": 0.00015451936865048253,
"loss": 1.1281,
"step": 48
},
{
"epoch": 0.4,
"learning_rate": 0.00015534238952478284,
"loss": 1.0121,
"step": 49
},
{
"epoch": 0.41,
"learning_rate": 0.00015614878257204067,
"loss": 0.9725,
"step": 50
},
{
"epoch": 0.41,
"learning_rate": 0.00015693920638594356,
"loss": 1.0977,
"step": 51
},
{
"epoch": 0.42,
"learning_rate": 0.00015771428119052897,
"loss": 0.9913,
"step": 52
},
{
"epoch": 0.43,
"learning_rate": 0.00015847459176404594,
"loss": 0.9657,
"step": 53
},
{
"epoch": 0.44,
"learning_rate": 0.00015922069008950885,
"loss": 1.0791,
"step": 54
},
{
"epoch": 0.45,
"learning_rate": 0.00015995309776203757,
"loss": 1.0275,
"step": 55
},
{
"epoch": 0.45,
"learning_rate": 0.00016067230817928382,
"loss": 0.9755,
"step": 56
},
{
"epoch": 0.46,
"learning_rate": 0.00016137878853798225,
"loss": 1.0315,
"step": 57
},
{
"epoch": 0.47,
"learning_rate": 0.00016207298165686122,
"loss": 1.03,
"step": 58
},
{
"epoch": 0.48,
"learning_rate": 0.0001627553076437253,
"loss": 0.9478,
"step": 59
},
{
"epoch": 0.49,
"learning_rate": 0.00016342616542242586,
"loss": 1.0424,
"step": 60
},
{
"epoch": 0.49,
"eval_loss": 0.9511741995811462,
"eval_runtime": 105.0633,
"eval_samples_per_second": 31.714,
"eval_steps_per_second": 0.257,
"step": 60
},
{
"epoch": 0.49,
"learning_rate": 0.0001640859341336185,
"loss": 1.0543,
"step": 61
},
{
"epoch": 0.5,
"learning_rate": 0.00016473497442162107,
"loss": 0.9464,
"step": 62
},
{
"epoch": 0.51,
"learning_rate": 0.00016537362961831014,
"loss": 1.0091,
"step": 63
},
{
"epoch": 0.52,
"learning_rate": 0.00016600222683378475,
"loss": 1.1283,
"step": 64
},
{
"epoch": 0.53,
"learning_rate": 0.0001666210779624723,
"loss": 0.9776,
"step": 65
},
{
"epoch": 0.54,
"learning_rate": 0.00016723048061242278,
"loss": 1.0544,
"step": 66
},
{
"epoch": 0.54,
"learning_rate": 0.00016783071896472347,
"loss": 1.0389,
"step": 67
},
{
"epoch": 0.55,
"learning_rate": 0.0001684220645692458,
"loss": 0.9661,
"step": 68
},
{
"epoch": 0.56,
"learning_rate": 0.00016900477708230193,
"loss": 0.9977,
"step": 69
},
{
"epoch": 0.57,
"learning_rate": 0.00016957910495122718,
"loss": 1.1473,
"step": 70
},
{
"epoch": 0.58,
"learning_rate": 0.00017014528605040512,
"loss": 0.9642,
"step": 71
},
{
"epoch": 0.58,
"learning_rate": 0.0001707035482728111,
"loss": 0.9816,
"step": 72
},
{
"epoch": 0.59,
"learning_rate": 0.00017125411008075538,
"loss": 1.0252,
"step": 73
},
{
"epoch": 0.6,
"learning_rate": 0.00017179718101915754,
"loss": 0.931,
"step": 74
},
{
"epoch": 0.61,
"learning_rate": 0.00017233296219436924,
"loss": 0.9844,
"step": 75
},
{
"epoch": 0.62,
"learning_rate": 0.0001728616467212845,
"loss": 1.1343,
"step": 76
},
{
"epoch": 0.62,
"learning_rate": 0.00017338342014122407,
"loss": 0.9936,
"step": 77
},
{
"epoch": 0.63,
"learning_rate": 0.0001738984608128575,
"loss": 0.9236,
"step": 78
},
{
"epoch": 0.64,
"learning_rate": 0.0001744069402782235,
"loss": 1.0555,
"step": 79
},
{
"epoch": 0.65,
"learning_rate": 0.0001749090236057281,
"loss": 0.9858,
"step": 80
},
{
"epoch": 0.65,
"eval_loss": 0.9414578080177307,
"eval_runtime": 104.7003,
"eval_samples_per_second": 31.824,
"eval_steps_per_second": 0.258,
"step": 80
},
{
"epoch": 0.66,
"learning_rate": 0.00017540486971183742,
"loss": 0.9318,
"step": 81
},
{
"epoch": 0.67,
"learning_rate": 0.00017589463166303238,
"loss": 1.0657,
"step": 82
},
{
"epoch": 0.67,
"learning_rate": 0.00017637845695946241,
"loss": 0.999,
"step": 83
},
{
"epoch": 0.68,
"learning_rate": 0.00017685648780161234,
"loss": 0.9757,
"step": 84
},
{
"epoch": 0.69,
"learning_rate": 0.00017732886134118916,
"loss": 1.0634,
"step": 85
},
{
"epoch": 0.7,
"learning_rate": 0.00017779570991733563,
"loss": 1.0504,
"step": 86
},
{
"epoch": 0.71,
"learning_rate": 0.0001782571612791898,
"loss": 0.9327,
"step": 87
},
{
"epoch": 0.71,
"learning_rate": 0.00017871333879572503,
"loss": 1.0454,
"step": 88
},
{
"epoch": 0.72,
"learning_rate": 0.00017916436165373454,
"loss": 1.0217,
"step": 89
},
{
"epoch": 0.73,
"learning_rate": 0.00017961034504475446,
"loss": 0.9515,
"step": 90
},
{
"epoch": 0.74,
"learning_rate": 0.00018005140034165878,
"loss": 1.0084,
"step": 91
},
{
"epoch": 0.75,
"learning_rate": 0.00018048763526560415,
"loss": 1.0499,
"step": 92
},
{
"epoch": 0.75,
"learning_rate": 0.00018091915404394964,
"loss": 0.991,
"step": 93
},
{
"epoch": 0.76,
"learning_rate": 0.00018134605755973132,
"loss": 1.0103,
"step": 94
},
{
"epoch": 0.77,
"learning_rate": 0.00018176844349322783,
"loss": 1.0429,
"step": 95
},
{
"epoch": 0.78,
"learning_rate": 0.00018218640645611332,
"loss": 0.9532,
"step": 96
},
{
"epoch": 0.79,
"learning_rate": 0.00018260003811865945,
"loss": 0.9993,
"step": 97
},
{
"epoch": 0.8,
"learning_rate": 0.00018300942733041366,
"loss": 1.1194,
"step": 98
},
{
"epoch": 0.8,
"learning_rate": 0.00018341466023475133,
"loss": 0.9519,
"step": 99
},
{
"epoch": 0.81,
"learning_rate": 0.0001838158203776715,
"loss": 0.9457,
"step": 100
},
{
"epoch": 0.81,
"eval_loss": 0.9341189861297607,
"eval_runtime": 105.432,
"eval_samples_per_second": 31.603,
"eval_steps_per_second": 0.256,
"step": 100
},
{
"epoch": 0.82,
"learning_rate": 0.000184212988811179,
"loss": 1.1073,
"step": 101
},
{
"epoch": 0.83,
"learning_rate": 0.00018460624419157432,
"loss": 0.9791,
"step": 102
},
{
"epoch": 0.84,
"learning_rate": 0.00018499566287294834,
"loss": 0.9439,
"step": 103
},
{
"epoch": 0.84,
"learning_rate": 0.00018538131899615973,
"loss": 1.0692,
"step": 104
},
{
"epoch": 0.85,
"learning_rate": 0.0001857632845735557,
"loss": 1.0314,
"step": 105
},
{
"epoch": 0.86,
"learning_rate": 0.00018614162956967674,
"loss": 0.9366,
"step": 106
},
{
"epoch": 0.87,
"learning_rate": 0.00018651642197817242,
"loss": 1.0502,
"step": 107
},
{
"epoch": 0.88,
"learning_rate": 0.00018688772789513967,
"loss": 1.0201,
"step": 108
},
{
"epoch": 0.88,
"learning_rate": 0.00018725561158908094,
"loss": 0.9718,
"step": 109
},
{
"epoch": 0.89,
"learning_rate": 0.0001876201355676684,
"loss": 0.9949,
"step": 110
},
{
"epoch": 0.9,
"learning_rate": 0.0001879813606414861,
"loss": 0.9924,
"step": 111
},
{
"epoch": 0.91,
"learning_rate": 0.0001883393459849146,
"loss": 0.9327,
"step": 112
},
{
"epoch": 0.92,
"learning_rate": 0.0001886941491943088,
"loss": 0.9908,
"step": 113
},
{
"epoch": 0.92,
"learning_rate": 0.00018904582634361305,
"loss": 1.0408,
"step": 114
},
{
"epoch": 0.93,
"learning_rate": 0.00018939443203754748,
"loss": 0.9303,
"step": 115
},
{
"epoch": 0.94,
"learning_rate": 0.000189740019462492,
"loss": 1.0056,
"step": 116
},
{
"epoch": 0.95,
"learning_rate": 0.00019008264043518606,
"loss": 1.0623,
"step": 117
},
{
"epoch": 0.96,
"learning_rate": 0.0001904223454493561,
"loss": 0.9385,
"step": 118
},
{
"epoch": 0.97,
"learning_rate": 0.00019075918372037564,
"loss": 0.9983,
"step": 119
},
{
"epoch": 0.97,
"learning_rate": 0.00019109320322805665,
"loss": 1.0584,
"step": 120
},
{
"epoch": 0.97,
"eval_loss": 0.9276898503303528,
"eval_runtime": 104.6096,
"eval_samples_per_second": 31.852,
"eval_steps_per_second": 0.258,
"step": 120
},
{
"epoch": 0.98,
"learning_rate": 0.00019142445075766529,
"loss": 0.928,
"step": 121
},
{
"epoch": 0.99,
"learning_rate": 0.0001917529719392493,
"loss": 0.9414,
"step": 122
},
{
"epoch": 1.0,
"learning_rate": 0.0001920788112853609,
"loss": 1.0277,
"step": 123
},
{
"epoch": 1.01,
"learning_rate": 0.00019240201222725186,
"loss": 1.0717,
"step": 124
},
{
"epoch": 1.01,
"learning_rate": 0.00019272261714961485,
"loss": 0.9567,
"step": 125
},
{
"epoch": 1.02,
"learning_rate": 0.0001930406674239409,
"loss": 0.8984,
"step": 126
},
{
"epoch": 1.03,
"learning_rate": 0.0001933562034405585,
"loss": 0.9453,
"step": 127
},
{
"epoch": 1.04,
"learning_rate": 0.00019366926463941554,
"loss": 0.9943,
"step": 128
},
{
"epoch": 1.05,
"learning_rate": 0.0001939798895396642,
"loss": 0.8644,
"step": 129
},
{
"epoch": 1.05,
"learning_rate": 0.0001942881157681031,
"loss": 0.968,
"step": 130
},
{
"epoch": 1.06,
"learning_rate": 0.00019459398008652982,
"loss": 1.0798,
"step": 131
},
{
"epoch": 1.07,
"learning_rate": 0.00019489751841805358,
"loss": 0.9165,
"step": 132
},
{
"epoch": 1.08,
"learning_rate": 0.0001951987658724143,
"loss": 0.9391,
"step": 133
},
{
"epoch": 1.09,
"learning_rate": 0.00019549775677035427,
"loss": 1.0106,
"step": 134
},
{
"epoch": 1.1,
"learning_rate": 0.000195794524667083,
"loss": 0.8805,
"step": 135
},
{
"epoch": 1.1,
"learning_rate": 0.0001960891023748766,
"loss": 0.935,
"step": 136
},
{
"epoch": 1.11,
"learning_rate": 0.00019638152198484917,
"loss": 1.0291,
"step": 137
},
{
"epoch": 1.12,
"learning_rate": 0.00019667181488793273,
"loss": 0.8887,
"step": 138
},
{
"epoch": 1.13,
"learning_rate": 0.00019696001179509912,
"loss": 0.8827,
"step": 139
},
{
"epoch": 1.14,
"learning_rate": 0.00019724614275685795,
"loss": 1.0284,
"step": 140
},
{
"epoch": 1.14,
"eval_loss": 0.9371529817581177,
"eval_runtime": 104.4681,
"eval_samples_per_second": 31.895,
"eval_steps_per_second": 0.258,
"step": 140
},
{
"epoch": 1.14,
"learning_rate": 0.0001975302371820599,
"loss": 0.8687,
"step": 141
},
{
"epoch": 1.15,
"learning_rate": 0.0001978123238560359,
"loss": 0.8771,
"step": 142
},
{
"epoch": 1.16,
"learning_rate": 0.00019809243095809996,
"loss": 1.0197,
"step": 143
},
{
"epoch": 1.17,
"learning_rate": 0.00019837058607844187,
"loss": 0.8979,
"step": 144
},
{
"epoch": 1.18,
"learning_rate": 0.00019864681623443537,
"loss": 0.9128,
"step": 145
},
{
"epoch": 1.18,
"learning_rate": 0.00019892114788638618,
"loss": 1.0225,
"step": 146
},
{
"epoch": 1.19,
"learning_rate": 0.00019919360695274218,
"loss": 0.9081,
"step": 147
},
{
"epoch": 1.2,
"learning_rate": 0.00019946421882478829,
"loss": 0.8753,
"step": 148
},
{
"epoch": 1.21,
"learning_rate": 0.00019973300838084634,
"loss": 0.9424,
"step": 149
},
{
"epoch": 1.22,
"learning_rate": 0.0002,
"loss": 0.9569,
"step": 150
},
{
"epoch": 1.23,
"learning_rate": 0.0002,
"loss": 0.8741,
"step": 151
},
{
"epoch": 1.23,
"learning_rate": 0.0002,
"loss": 0.9871,
"step": 152
},
{
"epoch": 1.24,
"learning_rate": 0.0002,
"loss": 1.0033,
"step": 153
},
{
"epoch": 1.25,
"learning_rate": 0.0002,
"loss": 0.8893,
"step": 154
},
{
"epoch": 1.26,
"learning_rate": 0.0002,
"loss": 0.9914,
"step": 155
},
{
"epoch": 1.27,
"learning_rate": 0.0002,
"loss": 0.9987,
"step": 156
},
{
"epoch": 1.27,
"learning_rate": 0.0002,
"loss": 0.8921,
"step": 157
},
{
"epoch": 1.28,
"learning_rate": 0.0002,
"loss": 0.9267,
"step": 158
},
{
"epoch": 1.29,
"learning_rate": 0.0002,
"loss": 1.0002,
"step": 159
},
{
"epoch": 1.3,
"learning_rate": 0.0002,
"loss": 0.8781,
"step": 160
},
{
"epoch": 1.3,
"eval_loss": 0.9295092821121216,
"eval_runtime": 103.7825,
"eval_samples_per_second": 32.106,
"eval_steps_per_second": 0.26,
"step": 160
},
{
"epoch": 1.31,
"learning_rate": 0.0002,
"loss": 0.9044,
"step": 161
},
{
"epoch": 1.31,
"learning_rate": 0.0002,
"loss": 1.0411,
"step": 162
},
{
"epoch": 1.32,
"learning_rate": 0.0002,
"loss": 0.8899,
"step": 163
},
{
"epoch": 1.33,
"learning_rate": 0.0002,
"loss": 0.9184,
"step": 164
},
{
"epoch": 1.34,
"learning_rate": 0.0002,
"loss": 1.0897,
"step": 165
},
{
"epoch": 1.35,
"learning_rate": 0.0002,
"loss": 0.8789,
"step": 166
},
{
"epoch": 1.35,
"learning_rate": 0.0002,
"loss": 0.8745,
"step": 167
},
{
"epoch": 1.36,
"learning_rate": 0.0002,
"loss": 1.0253,
"step": 168
},
{
"epoch": 1.37,
"learning_rate": 0.0002,
"loss": 0.9096,
"step": 169
},
{
"epoch": 1.38,
"learning_rate": 0.0002,
"loss": 0.8611,
"step": 170
},
{
"epoch": 1.39,
"learning_rate": 0.0002,
"loss": 1.0522,
"step": 171
},
{
"epoch": 1.4,
"learning_rate": 0.0002,
"loss": 0.9484,
"step": 172
},
{
"epoch": 1.4,
"learning_rate": 0.0002,
"loss": 0.8581,
"step": 173
},
{
"epoch": 1.41,
"learning_rate": 0.0002,
"loss": 0.9673,
"step": 174
},
{
"epoch": 1.42,
"learning_rate": 0.0002,
"loss": 0.8891,
"step": 175
},
{
"epoch": 1.43,
"learning_rate": 0.0002,
"loss": 0.8821,
"step": 176
},
{
"epoch": 1.44,
"learning_rate": 0.0002,
"loss": 0.9771,
"step": 177
},
{
"epoch": 1.44,
"learning_rate": 0.0002,
"loss": 0.974,
"step": 178
},
{
"epoch": 1.45,
"learning_rate": 0.0002,
"loss": 0.8719,
"step": 179
},
{
"epoch": 1.46,
"learning_rate": 0.0002,
"loss": 0.9531,
"step": 180
},
{
"epoch": 1.46,
"eval_loss": 0.926723837852478,
"eval_runtime": 104.5098,
"eval_samples_per_second": 31.882,
"eval_steps_per_second": 0.258,
"step": 180
},
{
"epoch": 1.47,
"learning_rate": 0.0002,
"loss": 0.9568,
"step": 181
},
{
"epoch": 1.48,
"learning_rate": 0.0002,
"loss": 0.8702,
"step": 182
},
{
"epoch": 1.48,
"learning_rate": 0.0002,
"loss": 0.98,
"step": 183
},
{
"epoch": 1.49,
"learning_rate": 0.0002,
"loss": 0.9709,
"step": 184
},
{
"epoch": 1.5,
"learning_rate": 0.0002,
"loss": 0.8892,
"step": 185
},
{
"epoch": 1.51,
"learning_rate": 0.0002,
"loss": 0.8827,
"step": 186
},
{
"epoch": 1.52,
"learning_rate": 0.0002,
"loss": 0.9874,
"step": 187
},
{
"epoch": 1.53,
"learning_rate": 0.0002,
"loss": 0.8684,
"step": 188
},
{
"epoch": 1.53,
"learning_rate": 0.0002,
"loss": 0.8632,
"step": 189
},
{
"epoch": 1.54,
"learning_rate": 0.0002,
"loss": 1.0162,
"step": 190
},
{
"epoch": 1.55,
"learning_rate": 0.0002,
"loss": 0.9045,
"step": 191
},
{
"epoch": 1.56,
"learning_rate": 0.0002,
"loss": 0.9085,
"step": 192
},
{
"epoch": 1.57,
"learning_rate": 0.0002,
"loss": 1.0432,
"step": 193
},
{
"epoch": 1.57,
"learning_rate": 0.0002,
"loss": 0.8893,
"step": 194
},
{
"epoch": 1.58,
"learning_rate": 0.0002,
"loss": 0.9048,
"step": 195
},
{
"epoch": 1.59,
"learning_rate": 0.0002,
"loss": 0.9976,
"step": 196
},
{
"epoch": 1.6,
"learning_rate": 0.0002,
"loss": 0.9437,
"step": 197
},
{
"epoch": 1.61,
"learning_rate": 0.0002,
"loss": 0.8922,
"step": 198
},
{
"epoch": 1.61,
"learning_rate": 0.0002,
"loss": 1.0123,
"step": 199
},
{
"epoch": 1.62,
"learning_rate": 0.0002,
"loss": 0.9496,
"step": 200
},
{
"epoch": 1.62,
"eval_loss": 0.9225737452507019,
"eval_runtime": 104.9516,
"eval_samples_per_second": 31.748,
"eval_steps_per_second": 0.257,
"step": 200
},
{
"epoch": 1.63,
"learning_rate": 0.0002,
"loss": 0.88,
"step": 201
},
{
"epoch": 1.64,
"learning_rate": 0.0002,
"loss": 0.9854,
"step": 202
},
{
"epoch": 1.65,
"learning_rate": 0.0002,
"loss": 0.9728,
"step": 203
},
{
"epoch": 1.66,
"learning_rate": 0.0002,
"loss": 0.8941,
"step": 204
},
{
"epoch": 1.66,
"learning_rate": 0.0002,
"loss": 0.9427,
"step": 205
},
{
"epoch": 1.67,
"learning_rate": 0.0002,
"loss": 0.9838,
"step": 206
},
{
"epoch": 1.68,
"learning_rate": 0.0002,
"loss": 0.8835,
"step": 207
},
{
"epoch": 1.69,
"learning_rate": 0.0002,
"loss": 0.9632,
"step": 208
},
{
"epoch": 1.7,
"learning_rate": 0.0002,
"loss": 1.063,
"step": 209
},
{
"epoch": 1.7,
"learning_rate": 0.0002,
"loss": 0.8781,
"step": 210
},
{
"epoch": 1.71,
"learning_rate": 0.0002,
"loss": 0.9306,
"step": 211
},
{
"epoch": 1.72,
"learning_rate": 0.0002,
"loss": 1.0085,
"step": 212
},
{
"epoch": 1.73,
"learning_rate": 0.0002,
"loss": 0.8721,
"step": 213
},
{
"epoch": 1.74,
"learning_rate": 0.0002,
"loss": 0.9305,
"step": 214
},
{
"epoch": 1.74,
"learning_rate": 0.0002,
"loss": 1.0009,
"step": 215
},
{
"epoch": 1.75,
"learning_rate": 0.0002,
"loss": 0.8773,
"step": 216
},
{
"epoch": 1.76,
"learning_rate": 0.0002,
"loss": 0.8806,
"step": 217
},
{
"epoch": 1.77,
"learning_rate": 0.0002,
"loss": 1.0875,
"step": 218
},
{
"epoch": 1.78,
"learning_rate": 0.0002,
"loss": 0.9055,
"step": 219
},
{
"epoch": 1.78,
"learning_rate": 0.0002,
"loss": 0.9178,
"step": 220
},
{
"epoch": 1.78,
"eval_loss": 0.9191879630088806,
"eval_runtime": 104.0108,
"eval_samples_per_second": 32.035,
"eval_steps_per_second": 0.26,
"step": 220
},
{
"epoch": 1.79,
"learning_rate": 0.0002,
"loss": 1.0323,
"step": 221
},
{
"epoch": 1.8,
"learning_rate": 0.0002,
"loss": 0.9546,
"step": 222
},
{
"epoch": 1.81,
"learning_rate": 0.0002,
"loss": 0.9247,
"step": 223
},
{
"epoch": 1.82,
"learning_rate": 0.0002,
"loss": 0.9444,
"step": 224
},
{
"epoch": 1.83,
"learning_rate": 0.0002,
"loss": 0.9241,
"step": 225
},
{
"epoch": 1.83,
"learning_rate": 0.0002,
"loss": 0.8956,
"step": 226
},
{
"epoch": 1.84,
"learning_rate": 0.0002,
"loss": 0.9338,
"step": 227
},
{
"epoch": 1.85,
"learning_rate": 0.0002,
"loss": 0.995,
"step": 228
},
{
"epoch": 1.86,
"learning_rate": 0.0002,
"loss": 0.8993,
"step": 229
},
{
"epoch": 1.87,
"learning_rate": 0.0002,
"loss": 0.9504,
"step": 230
},
{
"epoch": 1.87,
"learning_rate": 0.0002,
"loss": 0.9759,
"step": 231
},
{
"epoch": 1.88,
"learning_rate": 0.0002,
"loss": 0.8997,
"step": 232
},
{
"epoch": 1.89,
"learning_rate": 0.0002,
"loss": 0.9527,
"step": 233
},
{
"epoch": 1.9,
"learning_rate": 0.0002,
"loss": 1.0184,
"step": 234
},
{
"epoch": 1.91,
"learning_rate": 0.0002,
"loss": 0.899,
"step": 235
},
{
"epoch": 1.91,
"learning_rate": 0.0002,
"loss": 0.9286,
"step": 236
},
{
"epoch": 1.92,
"learning_rate": 0.0002,
"loss": 1.026,
"step": 237
},
{
"epoch": 1.93,
"learning_rate": 0.0002,
"loss": 0.8862,
"step": 238
},
{
"epoch": 1.94,
"learning_rate": 0.0002,
"loss": 0.9016,
"step": 239
},
{
"epoch": 1.95,
"learning_rate": 0.0002,
"loss": 1.0763,
"step": 240
},
{
"epoch": 1.95,
"eval_loss": 0.9153895974159241,
"eval_runtime": 103.9918,
"eval_samples_per_second": 32.041,
"eval_steps_per_second": 0.26,
"step": 240
},
{
"epoch": 1.96,
"learning_rate": 0.0002,
"loss": 0.8882,
"step": 241
},
{
"epoch": 1.96,
"learning_rate": 0.0002,
"loss": 0.8924,
"step": 242
},
{
"epoch": 1.97,
"learning_rate": 0.0002,
"loss": 1.0703,
"step": 243
},
{
"epoch": 1.98,
"learning_rate": 0.0002,
"loss": 0.8981,
"step": 244
},
{
"epoch": 1.99,
"learning_rate": 0.0002,
"loss": 0.9054,
"step": 245
},
{
"epoch": 2.0,
"learning_rate": 0.0002,
"loss": 0.9789,
"step": 246
},
{
"epoch": 2.0,
"learning_rate": 0.0002,
"loss": 0.9292,
"step": 247
},
{
"epoch": 2.01,
"learning_rate": 0.0002,
"loss": 0.8948,
"step": 248
},
{
"epoch": 2.02,
"learning_rate": 0.0002,
"loss": 0.7865,
"step": 249
},
{
"epoch": 2.03,
"learning_rate": 0.0002,
"loss": 0.8547,
"step": 250
},
{
"epoch": 2.04,
"learning_rate": 0.0002,
"loss": 0.9293,
"step": 251
},
{
"epoch": 2.04,
"learning_rate": 0.0002,
"loss": 0.7914,
"step": 252
},
{
"epoch": 2.05,
"learning_rate": 0.0002,
"loss": 0.8331,
"step": 253
},
{
"epoch": 2.06,
"learning_rate": 0.0002,
"loss": 0.9477,
"step": 254
},
{
"epoch": 2.07,
"learning_rate": 0.0002,
"loss": 0.8101,
"step": 255
},
{
"epoch": 2.08,
"learning_rate": 0.0002,
"loss": 0.8236,
"step": 256
},
{
"epoch": 2.09,
"learning_rate": 0.0002,
"loss": 0.93,
"step": 257
},
{
"epoch": 2.09,
"learning_rate": 0.0002,
"loss": 0.8066,
"step": 258
},
{
"epoch": 2.1,
"learning_rate": 0.0002,
"loss": 0.7846,
"step": 259
},
{
"epoch": 2.11,
"learning_rate": 0.0002,
"loss": 0.9561,
"step": 260
},
{
"epoch": 2.11,
"eval_loss": 0.9422550201416016,
"eval_runtime": 104.5769,
"eval_samples_per_second": 31.862,
"eval_steps_per_second": 0.258,
"step": 260
},
{
"epoch": 2.12,
"learning_rate": 0.0002,
"loss": 0.8292,
"step": 261
},
{
"epoch": 2.13,
"learning_rate": 0.0002,
"loss": 0.7639,
"step": 262
},
{
"epoch": 2.13,
"learning_rate": 0.0002,
"loss": 0.895,
"step": 263
},
{
"epoch": 2.14,
"learning_rate": 0.0002,
"loss": 0.8531,
"step": 264
},
{
"epoch": 2.15,
"learning_rate": 0.0002,
"loss": 0.7822,
"step": 265
},
{
"epoch": 2.16,
"learning_rate": 0.0002,
"loss": 0.9096,
"step": 266
},
{
"epoch": 2.17,
"learning_rate": 0.0002,
"loss": 0.8751,
"step": 267
},
{
"epoch": 2.17,
"learning_rate": 0.0002,
"loss": 0.7933,
"step": 268
},
{
"epoch": 2.18,
"learning_rate": 0.0002,
"loss": 0.8531,
"step": 269
},
{
"epoch": 2.19,
"learning_rate": 0.0002,
"loss": 0.8781,
"step": 270
},
{
"epoch": 2.2,
"learning_rate": 0.0002,
"loss": 0.7974,
"step": 271
},
{
"epoch": 2.21,
"learning_rate": 0.0002,
"loss": 0.8571,
"step": 272
},
{
"epoch": 2.22,
"learning_rate": 0.0002,
"loss": 0.893,
"step": 273
},
{
"epoch": 2.22,
"learning_rate": 0.0002,
"loss": 0.7881,
"step": 274
},
{
"epoch": 2.23,
"learning_rate": 0.0002,
"loss": 0.8831,
"step": 275
},
{
"epoch": 2.24,
"learning_rate": 0.0002,
"loss": 0.9272,
"step": 276
},
{
"epoch": 2.25,
"learning_rate": 0.0002,
"loss": 0.7924,
"step": 277
},
{
"epoch": 2.26,
"learning_rate": 0.0002,
"loss": 0.8152,
"step": 278
},
{
"epoch": 2.26,
"learning_rate": 0.0002,
"loss": 0.9071,
"step": 279
},
{
"epoch": 2.27,
"learning_rate": 0.0002,
"loss": 0.7991,
"step": 280
},
{
"epoch": 2.27,
"eval_loss": 0.9367637634277344,
"eval_runtime": 104.4183,
"eval_samples_per_second": 31.91,
"eval_steps_per_second": 0.259,
"step": 280
},
{
"epoch": 2.28,
"learning_rate": 0.0002,
"loss": 0.8638,
"step": 281
},
{
"epoch": 2.29,
"learning_rate": 0.0002,
"loss": 0.9924,
"step": 282
},
{
"epoch": 2.3,
"learning_rate": 0.0002,
"loss": 0.8228,
"step": 283
},
{
"epoch": 2.3,
"learning_rate": 0.0002,
"loss": 0.8012,
"step": 284
},
{
"epoch": 2.31,
"learning_rate": 0.0002,
"loss": 0.9366,
"step": 285
},
{
"epoch": 2.32,
"learning_rate": 0.0002,
"loss": 0.8264,
"step": 286
},
{
"epoch": 2.33,
"learning_rate": 0.0002,
"loss": 0.7954,
"step": 287
},
{
"epoch": 2.34,
"learning_rate": 0.0002,
"loss": 0.9375,
"step": 288
},
{
"epoch": 2.34,
"learning_rate": 0.0002,
"loss": 0.8495,
"step": 289
},
{
"epoch": 2.35,
"learning_rate": 0.0002,
"loss": 0.8029,
"step": 290
},
{
"epoch": 2.36,
"learning_rate": 0.0002,
"loss": 0.9309,
"step": 291
},
{
"epoch": 2.37,
"learning_rate": 0.0002,
"loss": 0.895,
"step": 292
},
{
"epoch": 2.38,
"learning_rate": 0.0002,
"loss": 0.7859,
"step": 293
},
{
"epoch": 2.39,
"learning_rate": 0.0002,
"loss": 0.9474,
"step": 294
},
{
"epoch": 2.39,
"learning_rate": 0.0002,
"loss": 0.8534,
"step": 295
},
{
"epoch": 2.4,
"learning_rate": 0.0002,
"loss": 0.8233,
"step": 296
},
{
"epoch": 2.41,
"learning_rate": 0.0002,
"loss": 0.85,
"step": 297
},
{
"epoch": 2.42,
"learning_rate": 0.0002,
"loss": 0.9265,
"step": 298
},
{
"epoch": 2.43,
"learning_rate": 0.0002,
"loss": 0.8192,
"step": 299
},
{
"epoch": 2.43,
"learning_rate": 0.0002,
"loss": 0.8503,
"step": 300
},
{
"epoch": 2.43,
"eval_loss": 0.9363464117050171,
"eval_runtime": 104.4808,
"eval_samples_per_second": 31.891,
"eval_steps_per_second": 0.258,
"step": 300
},
{
"epoch": 2.44,
"learning_rate": 0.0002,
"loss": 0.9259,
"step": 301
},
{
"epoch": 2.45,
"learning_rate": 0.0002,
"loss": 0.8339,
"step": 302
},
{
"epoch": 2.46,
"learning_rate": 0.0002,
"loss": 0.8471,
"step": 303
},
{
"epoch": 2.47,
"learning_rate": 0.0002,
"loss": 0.9809,
"step": 304
},
{
"epoch": 2.47,
"learning_rate": 0.0002,
"loss": 0.8134,
"step": 305
},
{
"epoch": 2.48,
"learning_rate": 0.0002,
"loss": 0.8459,
"step": 306
},
{
"epoch": 2.49,
"learning_rate": 0.0002,
"loss": 0.9366,
"step": 307
},
{
"epoch": 2.5,
"learning_rate": 0.0002,
"loss": 0.8253,
"step": 308
},
{
"epoch": 2.51,
"learning_rate": 0.0002,
"loss": 0.786,
"step": 309
},
{
"epoch": 2.52,
"learning_rate": 0.0002,
"loss": 0.9092,
"step": 310
},
{
"epoch": 2.52,
"learning_rate": 0.0002,
"loss": 0.8864,
"step": 311
},
{
"epoch": 2.53,
"learning_rate": 0.0002,
"loss": 0.8244,
"step": 312
},
{
"epoch": 2.54,
"learning_rate": 0.0002,
"loss": 0.9474,
"step": 313
},
{
"epoch": 2.55,
"learning_rate": 0.0002,
"loss": 0.818,
"step": 314
},
{
"epoch": 2.56,
"learning_rate": 0.0002,
"loss": 0.8447,
"step": 315
},
{
"epoch": 2.56,
"learning_rate": 0.0002,
"loss": 0.9267,
"step": 316
},
{
"epoch": 2.57,
"learning_rate": 0.0002,
"loss": 0.8774,
"step": 317
},
{
"epoch": 2.58,
"learning_rate": 0.0002,
"loss": 0.8042,
"step": 318
},
{
"epoch": 2.59,
"learning_rate": 0.0002,
"loss": 0.8845,
"step": 319
},
{
"epoch": 2.6,
"learning_rate": 0.0002,
"loss": 0.8749,
"step": 320
},
{
"epoch": 2.6,
"eval_loss": 0.9298563003540039,
"eval_runtime": 104.4409,
"eval_samples_per_second": 31.903,
"eval_steps_per_second": 0.259,
"step": 320
},
{
"epoch": 2.6,
"step": 320,
"total_flos": 1.8893507662761165e+19,
"train_loss": 0.9615403175354004,
"train_runtime": 23129.1803,
"train_samples_per_second": 13.641,
"train_steps_per_second": 0.027
}
],
"max_steps": 615,
"num_train_epochs": 5,
"total_flos": 1.8893507662761165e+19,
"trial_name": null,
"trial_params": null
}