PEFT
Safetensors
trrapi-16 / trainer_state.json
rizla's picture
Upload folder using huggingface_hub
287e233 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9744245524296675,
"eval_steps": 500,
"global_step": 390,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 2e-05,
"loss": 1.3382,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 4e-05,
"loss": 1.2632,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 6e-05,
"loss": 1.2406,
"step": 3
},
{
"epoch": 0.02,
"learning_rate": 8e-05,
"loss": 1.2408,
"step": 4
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 0.9425,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 0.00012,
"loss": 0.9388,
"step": 6
},
{
"epoch": 0.04,
"learning_rate": 0.00014,
"loss": 0.7308,
"step": 7
},
{
"epoch": 0.04,
"learning_rate": 0.00016,
"loss": 0.7467,
"step": 8
},
{
"epoch": 0.05,
"learning_rate": 0.00018,
"loss": 0.5595,
"step": 9
},
{
"epoch": 0.05,
"learning_rate": 0.0002,
"loss": 0.6003,
"step": 10
},
{
"epoch": 0.06,
"learning_rate": 0.00019999850743593963,
"loss": 0.5589,
"step": 11
},
{
"epoch": 0.06,
"learning_rate": 0.0001999940297883134,
"loss": 0.6077,
"step": 12
},
{
"epoch": 0.07,
"learning_rate": 0.00019998656719078482,
"loss": 0.5765,
"step": 13
},
{
"epoch": 0.07,
"learning_rate": 0.00019997611986612203,
"loss": 0.4975,
"step": 14
},
{
"epoch": 0.08,
"learning_rate": 0.00019996268812619107,
"loss": 0.4747,
"step": 15
},
{
"epoch": 0.08,
"learning_rate": 0.00019994627237194653,
"loss": 0.5549,
"step": 16
},
{
"epoch": 0.09,
"learning_rate": 0.00019992687309341976,
"loss": 0.5668,
"step": 17
},
{
"epoch": 0.09,
"learning_rate": 0.00019990449086970403,
"loss": 0.5841,
"step": 18
},
{
"epoch": 0.1,
"learning_rate": 0.00019987912636893745,
"loss": 0.6035,
"step": 19
},
{
"epoch": 0.1,
"learning_rate": 0.0001998507803482828,
"loss": 0.4897,
"step": 20
},
{
"epoch": 0.11,
"learning_rate": 0.00019981945365390516,
"loss": 0.4759,
"step": 21
},
{
"epoch": 0.11,
"learning_rate": 0.00019978514722094647,
"loss": 0.5182,
"step": 22
},
{
"epoch": 0.12,
"learning_rate": 0.00019974786207349775,
"loss": 0.9306,
"step": 23
},
{
"epoch": 0.12,
"learning_rate": 0.00019970759932456836,
"loss": 0.5365,
"step": 24
},
{
"epoch": 0.13,
"learning_rate": 0.00019966436017605297,
"loss": 0.5729,
"step": 25
},
{
"epoch": 0.13,
"learning_rate": 0.00019961814591869557,
"loss": 0.4515,
"step": 26
},
{
"epoch": 0.14,
"learning_rate": 0.0001995689579320509,
"loss": 0.5547,
"step": 27
},
{
"epoch": 0.14,
"learning_rate": 0.00019951679768444346,
"loss": 0.5144,
"step": 28
},
{
"epoch": 0.15,
"learning_rate": 0.00019946166673292344,
"loss": 0.4723,
"step": 29
},
{
"epoch": 0.15,
"learning_rate": 0.00019940356672322037,
"loss": 0.4073,
"step": 30
},
{
"epoch": 0.16,
"learning_rate": 0.00019934249938969396,
"loss": 0.417,
"step": 31
},
{
"epoch": 0.16,
"learning_rate": 0.0001992784665552824,
"loss": 0.4371,
"step": 32
},
{
"epoch": 0.17,
"learning_rate": 0.0001992114701314478,
"loss": 0.4897,
"step": 33
},
{
"epoch": 0.17,
"learning_rate": 0.00019914151211811924,
"loss": 0.5069,
"step": 34
},
{
"epoch": 0.18,
"learning_rate": 0.00019906859460363307,
"loss": 0.511,
"step": 35
},
{
"epoch": 0.18,
"learning_rate": 0.00019899271976467055,
"loss": 0.4006,
"step": 36
},
{
"epoch": 0.19,
"learning_rate": 0.00019891388986619277,
"loss": 0.3999,
"step": 37
},
{
"epoch": 0.19,
"learning_rate": 0.00019883210726137326,
"loss": 0.5261,
"step": 38
},
{
"epoch": 0.2,
"learning_rate": 0.00019874737439152748,
"loss": 0.431,
"step": 39
},
{
"epoch": 0.2,
"learning_rate": 0.0001986596937860402,
"loss": 0.448,
"step": 40
},
{
"epoch": 0.21,
"learning_rate": 0.00019856906806228986,
"loss": 0.4472,
"step": 41
},
{
"epoch": 0.21,
"learning_rate": 0.00019847549992557038,
"loss": 0.4318,
"step": 42
},
{
"epoch": 0.22,
"learning_rate": 0.00019837899216901053,
"loss": 0.4525,
"step": 43
},
{
"epoch": 0.23,
"learning_rate": 0.00019827954767349048,
"loss": 0.4942,
"step": 44
},
{
"epoch": 0.23,
"learning_rate": 0.00019817716940755586,
"loss": 0.4034,
"step": 45
},
{
"epoch": 0.24,
"learning_rate": 0.00019807186042732907,
"loss": 0.472,
"step": 46
},
{
"epoch": 0.24,
"learning_rate": 0.00019796362387641806,
"loss": 0.4244,
"step": 47
},
{
"epoch": 0.25,
"learning_rate": 0.0001978524629858226,
"loss": 0.4703,
"step": 48
},
{
"epoch": 0.25,
"learning_rate": 0.00019773838107383767,
"loss": 0.3931,
"step": 49
},
{
"epoch": 0.26,
"learning_rate": 0.00019762138154595446,
"loss": 0.4041,
"step": 50
},
{
"epoch": 0.26,
"learning_rate": 0.00019750146789475885,
"loss": 0.4527,
"step": 51
},
{
"epoch": 0.27,
"learning_rate": 0.00019737864369982693,
"loss": 0.3539,
"step": 52
},
{
"epoch": 0.27,
"learning_rate": 0.00019725291262761828,
"loss": 0.3967,
"step": 53
},
{
"epoch": 0.28,
"learning_rate": 0.0001971242784313665,
"loss": 0.3604,
"step": 54
},
{
"epoch": 0.28,
"learning_rate": 0.00019699274495096712,
"loss": 0.4196,
"step": 55
},
{
"epoch": 0.29,
"learning_rate": 0.0001968583161128631,
"loss": 0.5119,
"step": 56
},
{
"epoch": 0.29,
"learning_rate": 0.0001967209959299275,
"loss": 0.3844,
"step": 57
},
{
"epoch": 0.3,
"learning_rate": 0.00019658078850134366,
"loss": 0.4271,
"step": 58
},
{
"epoch": 0.3,
"learning_rate": 0.000196437698012483,
"loss": 0.4148,
"step": 59
},
{
"epoch": 0.31,
"learning_rate": 0.00019629172873477995,
"loss": 0.3858,
"step": 60
},
{
"epoch": 0.31,
"learning_rate": 0.0001961428850256044,
"loss": 0.4999,
"step": 61
},
{
"epoch": 0.32,
"learning_rate": 0.00019599117132813184,
"loss": 0.3914,
"step": 62
},
{
"epoch": 0.32,
"learning_rate": 0.0001958365921712105,
"loss": 0.4424,
"step": 63
},
{
"epoch": 0.33,
"learning_rate": 0.00019567915216922623,
"loss": 0.3736,
"step": 64
},
{
"epoch": 0.33,
"learning_rate": 0.0001955188560219648,
"loss": 0.4515,
"step": 65
},
{
"epoch": 0.34,
"learning_rate": 0.00019535570851447165,
"loss": 0.3831,
"step": 66
},
{
"epoch": 0.34,
"learning_rate": 0.00019518971451690885,
"loss": 0.3598,
"step": 67
},
{
"epoch": 0.35,
"learning_rate": 0.00019502087898440987,
"loss": 0.3964,
"step": 68
},
{
"epoch": 0.35,
"learning_rate": 0.00019484920695693174,
"loss": 0.3708,
"step": 69
},
{
"epoch": 0.36,
"learning_rate": 0.00019467470355910438,
"loss": 0.7938,
"step": 70
},
{
"epoch": 0.36,
"learning_rate": 0.0001944973740000778,
"loss": 0.3971,
"step": 71
},
{
"epoch": 0.37,
"learning_rate": 0.00019431722357336656,
"loss": 0.4269,
"step": 72
},
{
"epoch": 0.37,
"learning_rate": 0.00019413425765669166,
"loss": 0.4206,
"step": 73
},
{
"epoch": 0.38,
"learning_rate": 0.0001939484817118202,
"loss": 0.4172,
"step": 74
},
{
"epoch": 0.38,
"learning_rate": 0.00019375990128440204,
"loss": 0.4395,
"step": 75
},
{
"epoch": 0.39,
"learning_rate": 0.00019356852200380463,
"loss": 0.4173,
"step": 76
},
{
"epoch": 0.39,
"learning_rate": 0.00019337434958294471,
"loss": 0.3666,
"step": 77
},
{
"epoch": 0.4,
"learning_rate": 0.00019317738981811778,
"loss": 0.3745,
"step": 78
},
{
"epoch": 0.4,
"learning_rate": 0.00019297764858882514,
"loss": 0.3125,
"step": 79
},
{
"epoch": 0.41,
"learning_rate": 0.00019277513185759844,
"loss": 0.4018,
"step": 80
},
{
"epoch": 0.41,
"learning_rate": 0.0001925698456698216,
"loss": 0.3906,
"step": 81
},
{
"epoch": 0.42,
"learning_rate": 0.00019236179615355026,
"loss": 0.4224,
"step": 82
},
{
"epoch": 0.42,
"learning_rate": 0.00019215098951932906,
"loss": 0.4389,
"step": 83
},
{
"epoch": 0.43,
"learning_rate": 0.00019193743206000617,
"loss": 0.3637,
"step": 84
},
{
"epoch": 0.43,
"learning_rate": 0.00019172113015054532,
"loss": 0.462,
"step": 85
},
{
"epoch": 0.44,
"learning_rate": 0.00019150209024783562,
"loss": 0.346,
"step": 86
},
{
"epoch": 0.45,
"learning_rate": 0.00019128031889049883,
"loss": 0.381,
"step": 87
},
{
"epoch": 0.45,
"learning_rate": 0.00019105582269869412,
"loss": 0.3668,
"step": 88
},
{
"epoch": 0.46,
"learning_rate": 0.00019082860837392037,
"loss": 0.3271,
"step": 89
},
{
"epoch": 0.46,
"learning_rate": 0.0001905986826988164,
"loss": 0.4114,
"step": 90
},
{
"epoch": 0.47,
"learning_rate": 0.00019036605253695802,
"loss": 0.3432,
"step": 91
},
{
"epoch": 0.47,
"learning_rate": 0.00019013072483265377,
"loss": 0.908,
"step": 92
},
{
"epoch": 0.48,
"learning_rate": 0.0001898927066107371,
"loss": 0.4049,
"step": 93
},
{
"epoch": 0.48,
"learning_rate": 0.0001896520049763568,
"loss": 0.3991,
"step": 94
},
{
"epoch": 0.49,
"learning_rate": 0.00018940862711476513,
"loss": 0.3593,
"step": 95
},
{
"epoch": 0.49,
"learning_rate": 0.00018916258029110305,
"loss": 0.4136,
"step": 96
},
{
"epoch": 0.5,
"learning_rate": 0.00018891387185018346,
"loss": 0.3989,
"step": 97
},
{
"epoch": 0.5,
"learning_rate": 0.000188662509216272,
"loss": 0.3329,
"step": 98
},
{
"epoch": 0.51,
"learning_rate": 0.00018840849989286532,
"loss": 0.3622,
"step": 99
},
{
"epoch": 0.51,
"learning_rate": 0.00018815185146246716,
"loss": 0.3152,
"step": 100
},
{
"epoch": 0.52,
"learning_rate": 0.00018789257158636203,
"loss": 0.3909,
"step": 101
},
{
"epoch": 0.52,
"learning_rate": 0.00018763066800438636,
"loss": 0.3701,
"step": 102
},
{
"epoch": 0.53,
"learning_rate": 0.00018736614853469768,
"loss": 0.3669,
"step": 103
},
{
"epoch": 0.53,
"learning_rate": 0.00018709902107354103,
"loss": 0.374,
"step": 104
},
{
"epoch": 0.54,
"learning_rate": 0.00018682929359501338,
"loss": 0.3207,
"step": 105
},
{
"epoch": 0.54,
"learning_rate": 0.00018655697415082556,
"loss": 0.3689,
"step": 106
},
{
"epoch": 0.55,
"learning_rate": 0.00018628207087006184,
"loss": 0.4283,
"step": 107
},
{
"epoch": 0.55,
"learning_rate": 0.00018600459195893738,
"loss": 0.3416,
"step": 108
},
{
"epoch": 0.56,
"learning_rate": 0.0001857245457005532,
"loss": 0.3555,
"step": 109
},
{
"epoch": 0.56,
"learning_rate": 0.00018544194045464886,
"loss": 0.3423,
"step": 110
},
{
"epoch": 0.57,
"learning_rate": 0.00018515678465735308,
"loss": 0.4063,
"step": 111
},
{
"epoch": 0.57,
"learning_rate": 0.00018486908682093173,
"loss": 0.3741,
"step": 112
},
{
"epoch": 0.58,
"learning_rate": 0.00018457885553353385,
"loss": 0.3447,
"step": 113
},
{
"epoch": 0.58,
"learning_rate": 0.00018428609945893518,
"loss": 0.3673,
"step": 114
},
{
"epoch": 0.59,
"learning_rate": 0.00018399082733627965,
"loss": 0.3729,
"step": 115
},
{
"epoch": 0.59,
"learning_rate": 0.00018369304797981843,
"loss": 0.3534,
"step": 116
},
{
"epoch": 0.6,
"learning_rate": 0.00018339277027864682,
"loss": 0.4007,
"step": 117
},
{
"epoch": 0.6,
"learning_rate": 0.00018309000319643892,
"loss": 0.327,
"step": 118
},
{
"epoch": 0.61,
"learning_rate": 0.00018278475577118,
"loss": 0.4156,
"step": 119
},
{
"epoch": 0.61,
"learning_rate": 0.00018247703711489686,
"loss": 0.3576,
"step": 120
},
{
"epoch": 0.62,
"learning_rate": 0.0001821668564133856,
"loss": 0.3178,
"step": 121
},
{
"epoch": 0.62,
"learning_rate": 0.0001818542229259376,
"loss": 0.3592,
"step": 122
},
{
"epoch": 0.63,
"learning_rate": 0.00018153914598506297,
"loss": 0.3595,
"step": 123
},
{
"epoch": 0.63,
"learning_rate": 0.00018122163499621207,
"loss": 0.3384,
"step": 124
},
{
"epoch": 0.64,
"learning_rate": 0.00018090169943749476,
"loss": 0.3765,
"step": 125
},
{
"epoch": 0.64,
"learning_rate": 0.00018057934885939734,
"loss": 0.4108,
"step": 126
},
{
"epoch": 0.65,
"learning_rate": 0.00018025459288449755,
"loss": 0.3517,
"step": 127
},
{
"epoch": 0.65,
"learning_rate": 0.00017992744120717735,
"loss": 0.3711,
"step": 128
},
{
"epoch": 0.66,
"learning_rate": 0.00017959790359333347,
"loss": 0.3513,
"step": 129
},
{
"epoch": 0.66,
"learning_rate": 0.00017926598988008582,
"loss": 0.323,
"step": 130
},
{
"epoch": 0.67,
"learning_rate": 0.00017893170997548407,
"loss": 0.3479,
"step": 131
},
{
"epoch": 0.68,
"learning_rate": 0.00017859507385821163,
"loss": 0.4064,
"step": 132
},
{
"epoch": 0.68,
"learning_rate": 0.00017825609157728786,
"loss": 0.3666,
"step": 133
},
{
"epoch": 0.69,
"learning_rate": 0.00017791477325176822,
"loss": 0.3472,
"step": 134
},
{
"epoch": 0.69,
"learning_rate": 0.000177571129070442,
"loss": 0.42,
"step": 135
},
{
"epoch": 0.7,
"learning_rate": 0.00017722516929152826,
"loss": 0.3668,
"step": 136
},
{
"epoch": 0.7,
"learning_rate": 0.00017687690424236967,
"loss": 0.3206,
"step": 137
},
{
"epoch": 0.71,
"learning_rate": 0.00017652634431912417,
"loss": 0.3175,
"step": 138
},
{
"epoch": 0.71,
"learning_rate": 0.00017617349998645456,
"loss": 0.3447,
"step": 139
},
{
"epoch": 0.72,
"learning_rate": 0.0001758183817772163,
"loss": 0.3392,
"step": 140
},
{
"epoch": 0.72,
"learning_rate": 0.00017546100029214285,
"loss": 0.3558,
"step": 141
},
{
"epoch": 0.73,
"learning_rate": 0.00017510136619952946,
"loss": 0.3608,
"step": 142
},
{
"epoch": 0.73,
"learning_rate": 0.00017473949023491455,
"loss": 0.3169,
"step": 143
},
{
"epoch": 0.74,
"learning_rate": 0.0001743753832007593,
"loss": 0.7764,
"step": 144
},
{
"epoch": 0.74,
"learning_rate": 0.0001740090559661252,
"loss": 0.372,
"step": 145
},
{
"epoch": 0.75,
"learning_rate": 0.00017364051946634952,
"loss": 0.3673,
"step": 146
},
{
"epoch": 0.75,
"learning_rate": 0.00017326978470271894,
"loss": 0.396,
"step": 147
},
{
"epoch": 0.76,
"learning_rate": 0.00017289686274214118,
"loss": 0.3247,
"step": 148
},
{
"epoch": 0.76,
"learning_rate": 0.00017252176471681453,
"loss": 0.3097,
"step": 149
},
{
"epoch": 0.77,
"learning_rate": 0.00017214450182389559,
"loss": 0.3513,
"step": 150
},
{
"epoch": 0.77,
"learning_rate": 0.0001717650853251651,
"loss": 0.3951,
"step": 151
},
{
"epoch": 0.78,
"learning_rate": 0.0001713835265466917,
"loss": 0.3385,
"step": 152
},
{
"epoch": 0.78,
"learning_rate": 0.00017099983687849372,
"loss": 0.3873,
"step": 153
},
{
"epoch": 0.79,
"learning_rate": 0.0001706140277741994,
"loss": 0.3257,
"step": 154
},
{
"epoch": 0.79,
"learning_rate": 0.00017022611075070474,
"loss": 0.3481,
"step": 155
},
{
"epoch": 0.8,
"learning_rate": 0.0001698360973878299,
"loss": 0.401,
"step": 156
},
{
"epoch": 0.8,
"learning_rate": 0.00016944399932797353,
"loss": 0.3221,
"step": 157
},
{
"epoch": 0.81,
"learning_rate": 0.00016904982827576498,
"loss": 0.3908,
"step": 158
},
{
"epoch": 0.81,
"learning_rate": 0.0001686535959977152,
"loss": 0.3281,
"step": 159
},
{
"epoch": 0.82,
"learning_rate": 0.00016825531432186543,
"loss": 0.2504,
"step": 160
},
{
"epoch": 0.82,
"learning_rate": 0.00016785499513743382,
"loss": 0.3504,
"step": 161
},
{
"epoch": 0.83,
"learning_rate": 0.00016745265039446106,
"loss": 0.3395,
"step": 162
},
{
"epoch": 0.83,
"learning_rate": 0.00016704829210345315,
"loss": 0.3443,
"step": 163
},
{
"epoch": 0.84,
"learning_rate": 0.00016664193233502316,
"loss": 0.3599,
"step": 164
},
{
"epoch": 0.84,
"learning_rate": 0.00016623358321953078,
"loss": 0.3378,
"step": 165
},
{
"epoch": 0.85,
"learning_rate": 0.00016582325694672033,
"loss": 0.364,
"step": 166
},
{
"epoch": 0.85,
"learning_rate": 0.00016541096576535671,
"loss": 0.3096,
"step": 167
},
{
"epoch": 0.86,
"learning_rate": 0.00016499672198285996,
"loss": 0.2976,
"step": 168
},
{
"epoch": 0.86,
"learning_rate": 0.0001645805379649377,
"loss": 0.3313,
"step": 169
},
{
"epoch": 0.87,
"learning_rate": 0.0001641624261352161,
"loss": 0.35,
"step": 170
},
{
"epoch": 0.87,
"learning_rate": 0.000163742398974869,
"loss": 0.3341,
"step": 171
},
{
"epoch": 0.88,
"learning_rate": 0.00016332046902224517,
"loss": 0.3568,
"step": 172
},
{
"epoch": 0.88,
"learning_rate": 0.00016289664887249438,
"loss": 0.2995,
"step": 173
},
{
"epoch": 0.89,
"learning_rate": 0.00016247095117719106,
"loss": 0.3311,
"step": 174
},
{
"epoch": 0.9,
"learning_rate": 0.00016204338864395684,
"loss": 0.3401,
"step": 175
},
{
"epoch": 0.9,
"learning_rate": 0.0001616139740360811,
"loss": 0.411,
"step": 176
},
{
"epoch": 0.91,
"learning_rate": 0.00016118272017214005,
"loss": 0.3062,
"step": 177
},
{
"epoch": 0.91,
"learning_rate": 0.0001607496399256141,
"loss": 0.3165,
"step": 178
},
{
"epoch": 0.92,
"learning_rate": 0.00016031474622450344,
"loss": 0.3397,
"step": 179
},
{
"epoch": 0.92,
"learning_rate": 0.00015987805205094227,
"loss": 0.2843,
"step": 180
},
{
"epoch": 0.93,
"learning_rate": 0.00015943957044081114,
"loss": 0.3176,
"step": 181
},
{
"epoch": 0.93,
"learning_rate": 0.0001589993144833479,
"loss": 0.3202,
"step": 182
},
{
"epoch": 0.94,
"learning_rate": 0.00015855729732075694,
"loss": 0.295,
"step": 183
},
{
"epoch": 0.94,
"learning_rate": 0.0001581135321478169,
"loss": 0.3332,
"step": 184
},
{
"epoch": 0.95,
"learning_rate": 0.00015766803221148673,
"loss": 0.2974,
"step": 185
},
{
"epoch": 0.95,
"learning_rate": 0.0001572208108105103,
"loss": 0.3369,
"step": 186
},
{
"epoch": 0.96,
"learning_rate": 0.0001567718812950194,
"loss": 0.3598,
"step": 187
},
{
"epoch": 0.96,
"learning_rate": 0.00015632125706613532,
"loss": 0.3692,
"step": 188
},
{
"epoch": 0.97,
"learning_rate": 0.00015586895157556854,
"loss": 0.2853,
"step": 189
},
{
"epoch": 0.97,
"learning_rate": 0.0001554149783252175,
"loss": 0.3254,
"step": 190
},
{
"epoch": 0.98,
"learning_rate": 0.0001549593508667653,
"loss": 0.2956,
"step": 191
},
{
"epoch": 0.98,
"learning_rate": 0.00015450208280127543,
"loss": 0.3405,
"step": 192
},
{
"epoch": 0.99,
"learning_rate": 0.00015404318777878544,
"loss": 0.3127,
"step": 193
},
{
"epoch": 0.99,
"learning_rate": 0.00015358267949789966,
"loss": 0.2814,
"step": 194
},
{
"epoch": 1.0,
"learning_rate": 0.00015312057170538035,
"loss": 0.3384,
"step": 195
},
{
"epoch": 1.0,
"learning_rate": 0.00015265687819573712,
"loss": 0.3027,
"step": 196
},
{
"epoch": 1.01,
"learning_rate": 0.00015219161281081536,
"loss": 0.3238,
"step": 197
},
{
"epoch": 1.01,
"learning_rate": 0.00015172478943938287,
"loss": 0.3162,
"step": 198
},
{
"epoch": 1.02,
"learning_rate": 0.0001512564220167154,
"loss": 0.3059,
"step": 199
},
{
"epoch": 1.0,
"learning_rate": 0.00015078652452418063,
"loss": 0.2869,
"step": 200
},
{
"epoch": 1.01,
"learning_rate": 0.0001503151109888207,
"loss": 0.2496,
"step": 201
},
{
"epoch": 1.01,
"learning_rate": 0.0001498421954829336,
"loss": 0.2783,
"step": 202
},
{
"epoch": 1.02,
"learning_rate": 0.00014936779212365318,
"loss": 0.2645,
"step": 203
},
{
"epoch": 1.02,
"learning_rate": 0.00014889191507252742,
"loss": 0.2581,
"step": 204
},
{
"epoch": 1.03,
"learning_rate": 0.00014841457853509606,
"loss": 0.3068,
"step": 205
},
{
"epoch": 1.03,
"learning_rate": 0.0001479357967604663,
"loss": 0.2443,
"step": 206
},
{
"epoch": 1.04,
"learning_rate": 0.00014745558404088751,
"loss": 0.2867,
"step": 207
},
{
"epoch": 1.04,
"learning_rate": 0.0001469739547113246,
"loss": 0.281,
"step": 208
},
{
"epoch": 1.05,
"learning_rate": 0.00014649092314903016,
"loss": 0.3124,
"step": 209
},
{
"epoch": 1.05,
"learning_rate": 0.00014600650377311522,
"loss": 0.3362,
"step": 210
},
{
"epoch": 1.06,
"learning_rate": 0.00014552071104411874,
"loss": 0.2616,
"step": 211
},
{
"epoch": 1.06,
"learning_rate": 0.0001450335594635761,
"loss": 0.2772,
"step": 212
},
{
"epoch": 1.07,
"learning_rate": 0.0001445450635735861,
"loss": 0.3061,
"step": 213
},
{
"epoch": 1.07,
"learning_rate": 0.000144055237956377,
"loss": 0.2847,
"step": 214
},
{
"epoch": 1.08,
"learning_rate": 0.0001435640972338709,
"loss": 0.2663,
"step": 215
},
{
"epoch": 1.08,
"learning_rate": 0.00014307165606724777,
"loss": 0.2415,
"step": 216
},
{
"epoch": 1.09,
"learning_rate": 0.00014257792915650728,
"loss": 0.2644,
"step": 217
},
{
"epoch": 1.09,
"learning_rate": 0.00014208293124003026,
"loss": 0.2536,
"step": 218
},
{
"epoch": 1.1,
"learning_rate": 0.00014158667709413876,
"loss": 0.3251,
"step": 219
},
{
"epoch": 1.1,
"learning_rate": 0.00014108918153265485,
"loss": 0.2704,
"step": 220
},
{
"epoch": 1.11,
"learning_rate": 0.00014059045940645833,
"loss": 0.2876,
"step": 221
},
{
"epoch": 1.12,
"learning_rate": 0.00014009052560304378,
"loss": 0.2934,
"step": 222
},
{
"epoch": 1.12,
"learning_rate": 0.00013958939504607565,
"loss": 0.308,
"step": 223
},
{
"epoch": 1.13,
"learning_rate": 0.00013908708269494318,
"loss": 0.2507,
"step": 224
},
{
"epoch": 1.13,
"learning_rate": 0.00013858360354431355,
"loss": 0.2834,
"step": 225
},
{
"epoch": 1.14,
"learning_rate": 0.00013807897262368453,
"loss": 0.3131,
"step": 226
},
{
"epoch": 1.14,
"learning_rate": 0.0001375732049969356,
"loss": 0.2667,
"step": 227
},
{
"epoch": 1.15,
"learning_rate": 0.0001370663157618784,
"loss": 0.2756,
"step": 228
},
{
"epoch": 1.15,
"learning_rate": 0.00013655832004980608,
"loss": 0.3245,
"step": 229
},
{
"epoch": 1.16,
"learning_rate": 0.00013604923302504147,
"loss": 0.2593,
"step": 230
},
{
"epoch": 1.16,
"learning_rate": 0.00013553906988448452,
"loss": 0.2805,
"step": 231
},
{
"epoch": 1.17,
"learning_rate": 0.00013502784585715853,
"loss": 0.2656,
"step": 232
},
{
"epoch": 1.17,
"learning_rate": 0.00013451557620375578,
"loss": 0.3171,
"step": 233
},
{
"epoch": 1.18,
"learning_rate": 0.00013400227621618172,
"loss": 0.2946,
"step": 234
},
{
"epoch": 1.18,
"learning_rate": 0.00013348796121709862,
"loss": 0.2764,
"step": 235
},
{
"epoch": 1.19,
"learning_rate": 0.00013297264655946816,
"loss": 0.3237,
"step": 236
},
{
"epoch": 1.19,
"learning_rate": 0.00013245634762609307,
"loss": 0.2609,
"step": 237
},
{
"epoch": 1.2,
"learning_rate": 0.000131939079829158,
"loss": 0.2394,
"step": 238
},
{
"epoch": 1.2,
"learning_rate": 0.00013142085860976948,
"loss": 0.3038,
"step": 239
},
{
"epoch": 1.21,
"learning_rate": 0.00013090169943749476,
"loss": 0.2791,
"step": 240
},
{
"epoch": 1.21,
"learning_rate": 0.00013038161780990035,
"loss": 0.2882,
"step": 241
},
{
"epoch": 1.22,
"learning_rate": 0.00012986062925208914,
"loss": 0.3005,
"step": 242
},
{
"epoch": 1.22,
"learning_rate": 0.00012933874931623707,
"loss": 0.2622,
"step": 243
},
{
"epoch": 1.23,
"learning_rate": 0.0001288159935811289,
"loss": 0.2972,
"step": 244
},
{
"epoch": 1.23,
"learning_rate": 0.000128292377651693,
"loss": 0.2821,
"step": 245
},
{
"epoch": 1.24,
"learning_rate": 0.00012776791715853584,
"loss": 0.2537,
"step": 246
},
{
"epoch": 1.24,
"learning_rate": 0.00012724262775747498,
"loss": 0.2625,
"step": 247
},
{
"epoch": 1.25,
"learning_rate": 0.00012671652512907211,
"loss": 0.2502,
"step": 248
},
{
"epoch": 1.25,
"learning_rate": 0.0001261896249781647,
"loss": 0.243,
"step": 249
},
{
"epoch": 1.26,
"learning_rate": 0.00012566194303339739,
"loss": 0.2525,
"step": 250
},
{
"epoch": 1.26,
"learning_rate": 0.0001251334950467522,
"loss": 0.28,
"step": 251
},
{
"epoch": 1.27,
"learning_rate": 0.00012460429679307864,
"loss": 0.27,
"step": 252
},
{
"epoch": 1.27,
"learning_rate": 0.00012407436406962257,
"loss": 0.2837,
"step": 253
},
{
"epoch": 1.28,
"learning_rate": 0.00012354371269555477,
"loss": 0.2582,
"step": 254
},
{
"epoch": 1.28,
"learning_rate": 0.00012301235851149865,
"loss": 0.2729,
"step": 255
},
{
"epoch": 1.29,
"learning_rate": 0.0001224803173790573,
"loss": 0.2568,
"step": 256
},
{
"epoch": 1.29,
"learning_rate": 0.00012194760518034028,
"loss": 0.3049,
"step": 257
},
{
"epoch": 1.3,
"learning_rate": 0.00012141423781748913,
"loss": 0.278,
"step": 258
},
{
"epoch": 1.3,
"learning_rate": 0.00012088023121220306,
"loss": 0.2725,
"step": 259
},
{
"epoch": 1.31,
"learning_rate": 0.0001203456013052634,
"loss": 0.2921,
"step": 260
},
{
"epoch": 1.31,
"learning_rate": 0.00011981036405605781,
"loss": 0.6775,
"step": 261
},
{
"epoch": 1.32,
"learning_rate": 0.00011927453544210397,
"loss": 0.2876,
"step": 262
},
{
"epoch": 1.32,
"learning_rate": 0.00011873813145857249,
"loss": 0.2447,
"step": 263
},
{
"epoch": 1.33,
"learning_rate": 0.0001182011681178095,
"loss": 0.258,
"step": 264
},
{
"epoch": 1.34,
"learning_rate": 0.00011766366144885877,
"loss": 0.3096,
"step": 265
},
{
"epoch": 1.34,
"learning_rate": 0.00011712562749698291,
"loss": 0.2821,
"step": 266
},
{
"epoch": 1.35,
"learning_rate": 0.00011658708232318482,
"loss": 0.2885,
"step": 267
},
{
"epoch": 1.35,
"learning_rate": 0.00011604804200372786,
"loss": 0.2688,
"step": 268
},
{
"epoch": 1.36,
"learning_rate": 0.00011550852262965622,
"loss": 0.3084,
"step": 269
},
{
"epoch": 1.36,
"learning_rate": 0.00011496854030631443,
"loss": 0.3622,
"step": 270
},
{
"epoch": 1.37,
"learning_rate": 0.00011442811115286669,
"loss": 0.3177,
"step": 271
},
{
"epoch": 1.37,
"learning_rate": 0.00011388725130181565,
"loss": 0.2365,
"step": 272
},
{
"epoch": 1.38,
"learning_rate": 0.00011334597689852076,
"loss": 0.315,
"step": 273
},
{
"epoch": 1.38,
"learning_rate": 0.00011280430410071652,
"loss": 0.2996,
"step": 274
},
{
"epoch": 1.39,
"learning_rate": 0.00011226224907802985,
"loss": 0.2397,
"step": 275
},
{
"epoch": 1.39,
"learning_rate": 0.00011171982801149774,
"loss": 0.2619,
"step": 276
},
{
"epoch": 1.4,
"learning_rate": 0.00011117705709308395,
"loss": 0.2595,
"step": 277
},
{
"epoch": 1.4,
"learning_rate": 0.0001106339525251958,
"loss": 0.3081,
"step": 278
},
{
"epoch": 1.41,
"learning_rate": 0.00011009053052020047,
"loss": 0.2777,
"step": 279
},
{
"epoch": 1.41,
"learning_rate": 0.00010954680729994102,
"loss": 0.2983,
"step": 280
},
{
"epoch": 1.42,
"learning_rate": 0.00010900279909525226,
"loss": 0.2704,
"step": 281
},
{
"epoch": 1.42,
"learning_rate": 0.00010845852214547601,
"loss": 0.3239,
"step": 282
},
{
"epoch": 1.43,
"learning_rate": 0.00010791399269797659,
"loss": 0.2763,
"step": 283
},
{
"epoch": 1.43,
"learning_rate": 0.00010736922700765571,
"loss": 0.2625,
"step": 284
},
{
"epoch": 1.44,
"learning_rate": 0.0001068242413364671,
"loss": 0.2829,
"step": 285
},
{
"epoch": 1.44,
"learning_rate": 0.00010627905195293135,
"loss": 0.3096,
"step": 286
},
{
"epoch": 1.45,
"learning_rate": 0.00010573367513165002,
"loss": 0.3141,
"step": 287
},
{
"epoch": 1.45,
"learning_rate": 0.00010518812715282001,
"loss": 0.2886,
"step": 288
},
{
"epoch": 1.46,
"learning_rate": 0.00010464242430174737,
"loss": 0.3128,
"step": 289
},
{
"epoch": 1.46,
"learning_rate": 0.00010409658286836143,
"loss": 0.2515,
"step": 290
},
{
"epoch": 1.47,
"learning_rate": 0.0001035506191467283,
"loss": 0.2786,
"step": 291
},
{
"epoch": 1.47,
"learning_rate": 0.00010300454943456457,
"loss": 0.2659,
"step": 292
},
{
"epoch": 1.48,
"learning_rate": 0.00010245839003275075,
"loss": 0.3111,
"step": 293
},
{
"epoch": 1.48,
"learning_rate": 0.00010191215724484476,
"loss": 0.2724,
"step": 294
},
{
"epoch": 1.49,
"learning_rate": 0.0001013658673765951,
"loss": 0.2643,
"step": 295
},
{
"epoch": 1.49,
"learning_rate": 0.00010081953673545432,
"loss": 0.2959,
"step": 296
},
{
"epoch": 1.5,
"learning_rate": 0.00010027318163009201,
"loss": 0.32,
"step": 297
},
{
"epoch": 1.5,
"learning_rate": 9.9726818369908e-05,
"loss": 0.2284,
"step": 298
},
{
"epoch": 1.51,
"learning_rate": 9.918046326454568e-05,
"loss": 0.2774,
"step": 299
},
{
"epoch": 1.51,
"learning_rate": 9.863413262340491e-05,
"loss": 0.2739,
"step": 300
},
{
"epoch": 1.52,
"learning_rate": 9.808784275515525e-05,
"loss": 0.2802,
"step": 301
},
{
"epoch": 1.52,
"learning_rate": 9.754160996724927e-05,
"loss": 0.2807,
"step": 302
},
{
"epoch": 1.53,
"learning_rate": 9.699545056543547e-05,
"loss": 0.3228,
"step": 303
},
{
"epoch": 1.53,
"learning_rate": 9.644938085327173e-05,
"loss": 0.2623,
"step": 304
},
{
"epoch": 1.54,
"learning_rate": 9.590341713163858e-05,
"loss": 0.2146,
"step": 305
},
{
"epoch": 1.54,
"learning_rate": 9.535757569825265e-05,
"loss": 0.3186,
"step": 306
},
{
"epoch": 1.55,
"learning_rate": 9.481187284718006e-05,
"loss": 0.2684,
"step": 307
},
{
"epoch": 1.55,
"learning_rate": 9.426632486834998e-05,
"loss": 0.2485,
"step": 308
},
{
"epoch": 1.56,
"learning_rate": 9.372094804706867e-05,
"loss": 0.2952,
"step": 309
},
{
"epoch": 1.57,
"learning_rate": 9.317575866353292e-05,
"loss": 0.2922,
"step": 310
},
{
"epoch": 1.57,
"learning_rate": 9.263077299234433e-05,
"loss": 0.2877,
"step": 311
},
{
"epoch": 1.58,
"learning_rate": 9.208600730202339e-05,
"loss": 0.3022,
"step": 312
},
{
"epoch": 1.58,
"learning_rate": 9.1541477854524e-05,
"loss": 0.3008,
"step": 313
},
{
"epoch": 1.59,
"learning_rate": 9.099720090474779e-05,
"loss": 0.2642,
"step": 314
},
{
"epoch": 1.59,
"learning_rate": 9.0453192700059e-05,
"loss": 0.2634,
"step": 315
},
{
"epoch": 1.6,
"learning_rate": 8.990946947979954e-05,
"loss": 0.2458,
"step": 316
},
{
"epoch": 1.6,
"learning_rate": 8.936604747480422e-05,
"loss": 0.2905,
"step": 317
},
{
"epoch": 1.61,
"learning_rate": 8.882294290691609e-05,
"loss": 0.3353,
"step": 318
},
{
"epoch": 1.61,
"learning_rate": 8.828017198850227e-05,
"loss": 0.235,
"step": 319
},
{
"epoch": 1.62,
"learning_rate": 8.773775092197017e-05,
"loss": 0.2419,
"step": 320
},
{
"epoch": 1.62,
"learning_rate": 8.719569589928353e-05,
"loss": 0.2427,
"step": 321
},
{
"epoch": 1.63,
"learning_rate": 8.665402310147924e-05,
"loss": 0.3272,
"step": 322
},
{
"epoch": 1.63,
"learning_rate": 8.611274869818437e-05,
"loss": 0.2641,
"step": 323
},
{
"epoch": 1.64,
"learning_rate": 8.557188884713333e-05,
"loss": 0.323,
"step": 324
},
{
"epoch": 1.64,
"learning_rate": 8.503145969368562e-05,
"loss": 0.3062,
"step": 325
},
{
"epoch": 1.65,
"learning_rate": 8.449147737034379e-05,
"loss": 0.2871,
"step": 326
},
{
"epoch": 1.65,
"learning_rate": 8.395195799627216e-05,
"loss": 0.2646,
"step": 327
},
{
"epoch": 1.66,
"learning_rate": 8.341291767681523e-05,
"loss": 0.2825,
"step": 328
},
{
"epoch": 1.66,
"learning_rate": 8.287437250301709e-05,
"loss": 0.2766,
"step": 329
},
{
"epoch": 1.67,
"learning_rate": 8.233633855114127e-05,
"loss": 0.2694,
"step": 330
},
{
"epoch": 1.67,
"learning_rate": 8.179883188219053e-05,
"loss": 0.2568,
"step": 331
},
{
"epoch": 1.68,
"learning_rate": 8.126186854142752e-05,
"loss": 0.2722,
"step": 332
},
{
"epoch": 1.68,
"learning_rate": 8.072546455789604e-05,
"loss": 0.2528,
"step": 333
},
{
"epoch": 1.69,
"learning_rate": 8.018963594394221e-05,
"loss": 0.2779,
"step": 334
},
{
"epoch": 1.69,
"learning_rate": 7.965439869473664e-05,
"loss": 0.3015,
"step": 335
},
{
"epoch": 1.7,
"learning_rate": 7.911976878779696e-05,
"loss": 0.3106,
"step": 336
},
{
"epoch": 1.7,
"learning_rate": 7.858576218251089e-05,
"loss": 0.2919,
"step": 337
},
{
"epoch": 1.71,
"learning_rate": 7.805239481965976e-05,
"loss": 0.2134,
"step": 338
},
{
"epoch": 1.71,
"learning_rate": 7.75196826209427e-05,
"loss": 0.2825,
"step": 339
},
{
"epoch": 1.72,
"learning_rate": 7.698764148850137e-05,
"loss": 0.6593,
"step": 340
},
{
"epoch": 1.72,
"learning_rate": 7.645628730444524e-05,
"loss": 0.2553,
"step": 341
},
{
"epoch": 1.73,
"learning_rate": 7.592563593037746e-05,
"loss": 0.2711,
"step": 342
},
{
"epoch": 1.73,
"learning_rate": 7.539570320692137e-05,
"loss": 0.2259,
"step": 343
},
{
"epoch": 1.74,
"learning_rate": 7.486650495324783e-05,
"loss": 0.298,
"step": 344
},
{
"epoch": 1.74,
"learning_rate": 7.433805696660266e-05,
"loss": 0.2845,
"step": 345
},
{
"epoch": 1.75,
"learning_rate": 7.381037502183528e-05,
"loss": 0.2418,
"step": 346
},
{
"epoch": 1.75,
"learning_rate": 7.328347487092791e-05,
"loss": 0.2922,
"step": 347
},
{
"epoch": 1.76,
"learning_rate": 7.275737224252503e-05,
"loss": 0.2081,
"step": 348
},
{
"epoch": 1.76,
"learning_rate": 7.223208284146421e-05,
"loss": 0.617,
"step": 349
},
{
"epoch": 1.77,
"learning_rate": 7.170762234830699e-05,
"loss": 0.2832,
"step": 350
},
{
"epoch": 1.77,
"learning_rate": 7.118400641887115e-05,
"loss": 0.3025,
"step": 351
},
{
"epoch": 1.78,
"learning_rate": 7.066125068376297e-05,
"loss": 0.2848,
"step": 352
},
{
"epoch": 1.79,
"learning_rate": 7.013937074791089e-05,
"loss": 0.2231,
"step": 353
},
{
"epoch": 1.79,
"learning_rate": 6.961838219009967e-05,
"loss": 0.3314,
"step": 354
},
{
"epoch": 1.8,
"learning_rate": 6.909830056250527e-05,
"loss": 0.6465,
"step": 355
},
{
"epoch": 1.8,
"learning_rate": 6.857914139023057e-05,
"loss": 0.2569,
"step": 356
},
{
"epoch": 1.81,
"learning_rate": 6.8060920170842e-05,
"loss": 0.2786,
"step": 357
},
{
"epoch": 1.81,
"learning_rate": 6.754365237390696e-05,
"loss": 0.2755,
"step": 358
},
{
"epoch": 1.82,
"learning_rate": 6.702735344053187e-05,
"loss": 0.2787,
"step": 359
},
{
"epoch": 1.82,
"learning_rate": 6.651203878290139e-05,
"loss": 0.2526,
"step": 360
},
{
"epoch": 1.83,
"learning_rate": 6.59977237838183e-05,
"loss": 0.2558,
"step": 361
},
{
"epoch": 1.83,
"learning_rate": 6.548442379624424e-05,
"loss": 0.2445,
"step": 362
},
{
"epoch": 1.84,
"learning_rate": 6.497215414284146e-05,
"loss": 0.2696,
"step": 363
},
{
"epoch": 1.84,
"learning_rate": 6.446093011551551e-05,
"loss": 0.5883,
"step": 364
},
{
"epoch": 1.85,
"learning_rate": 6.395076697495854e-05,
"loss": 0.2489,
"step": 365
},
{
"epoch": 1.85,
"learning_rate": 6.344167995019395e-05,
"loss": 0.2775,
"step": 366
},
{
"epoch": 1.86,
"learning_rate": 6.29336842381216e-05,
"loss": 0.2961,
"step": 367
},
{
"epoch": 1.86,
"learning_rate": 6.242679500306443e-05,
"loss": 0.2515,
"step": 368
},
{
"epoch": 1.87,
"learning_rate": 6.192102737631551e-05,
"loss": 0.2506,
"step": 369
},
{
"epoch": 1.87,
"learning_rate": 6.141639645568646e-05,
"loss": 0.2633,
"step": 370
},
{
"epoch": 1.88,
"learning_rate": 6.091291730505684e-05,
"loss": 0.2475,
"step": 371
},
{
"epoch": 1.88,
"learning_rate": 6.041060495392436e-05,
"loss": 0.2636,
"step": 372
},
{
"epoch": 1.89,
"learning_rate": 5.9909474396956246e-05,
"loss": 0.3151,
"step": 373
},
{
"epoch": 1.89,
"learning_rate": 5.9409540593541645e-05,
"loss": 0.3158,
"step": 374
},
{
"epoch": 1.9,
"learning_rate": 5.8910818467345185e-05,
"loss": 0.2886,
"step": 375
},
{
"epoch": 1.9,
"learning_rate": 5.8413322905861254e-05,
"loss": 0.2976,
"step": 376
},
{
"epoch": 1.91,
"learning_rate": 5.791706875996974e-05,
"loss": 0.2651,
"step": 377
},
{
"epoch": 1.91,
"learning_rate": 5.7422070843492734e-05,
"loss": 0.3056,
"step": 378
},
{
"epoch": 1.92,
"learning_rate": 5.692834393275226e-05,
"loss": 0.2226,
"step": 379
},
{
"epoch": 1.92,
"learning_rate": 5.643590276612909e-05,
"loss": 0.2415,
"step": 380
},
{
"epoch": 1.93,
"learning_rate": 5.594476204362303e-05,
"loss": 0.3688,
"step": 381
},
{
"epoch": 1.93,
"learning_rate": 5.5454936426413884e-05,
"loss": 0.28,
"step": 382
},
{
"epoch": 1.94,
"learning_rate": 5.496644053642394e-05,
"loss": 0.3079,
"step": 383
},
{
"epoch": 1.94,
"learning_rate": 5.447928895588128e-05,
"loss": 0.259,
"step": 384
},
{
"epoch": 1.95,
"learning_rate": 5.399349622688479e-05,
"loss": 0.2649,
"step": 385
},
{
"epoch": 1.95,
"learning_rate": 5.350907685096983e-05,
"loss": 0.2557,
"step": 386
},
{
"epoch": 1.96,
"learning_rate": 5.302604528867544e-05,
"loss": 0.2479,
"step": 387
},
{
"epoch": 1.96,
"learning_rate": 5.254441595911255e-05,
"loss": 0.277,
"step": 388
},
{
"epoch": 1.97,
"learning_rate": 5.206420323953374e-05,
"loss": 0.2504,
"step": 389
},
{
"epoch": 1.97,
"learning_rate": 5.1585421464903994e-05,
"loss": 0.2799,
"step": 390
}
],
"logging_steps": 1,
"max_steps": 585,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 195,
"total_flos": 3.0058143620708434e+18,
"train_batch_size": 10,
"trial_name": null,
"trial_params": null
}