2336 lines
46 KiB
JSON
2336 lines
46 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 2.988235294117647,
|
|
"global_step": 381,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.01,
|
|
"learning_rate": 1.25e-06,
|
|
"loss": 0.9385,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 2.5e-06,
|
|
"loss": 0.9588,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.02,
|
|
"learning_rate": 3.7500000000000005e-06,
|
|
"loss": 0.8227,
|
|
"step": 3
|
|
},
|
|
{
|
|
"epoch": 0.03,
|
|
"learning_rate": 5e-06,
|
|
"loss": 0.7846,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"learning_rate": 6.25e-06,
|
|
"loss": 0.7867,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 7.500000000000001e-06,
|
|
"loss": 0.7457,
|
|
"step": 6
|
|
},
|
|
{
|
|
"epoch": 0.05,
|
|
"learning_rate": 8.750000000000001e-06,
|
|
"loss": 0.7486,
|
|
"step": 7
|
|
},
|
|
{
|
|
"epoch": 0.06,
|
|
"learning_rate": 1e-05,
|
|
"loss": 0.7079,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.07,
|
|
"learning_rate": 1.125e-05,
|
|
"loss": 0.6741,
|
|
"step": 9
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"learning_rate": 1.25e-05,
|
|
"loss": 0.6693,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.375e-05,
|
|
"loss": 0.6629,
|
|
"step": 11
|
|
},
|
|
{
|
|
"epoch": 0.09,
|
|
"learning_rate": 1.5000000000000002e-05,
|
|
"loss": 0.647,
|
|
"step": 12
|
|
},
|
|
{
|
|
"epoch": 0.1,
|
|
"learning_rate": 1.6250000000000002e-05,
|
|
"loss": 0.6316,
|
|
"step": 13
|
|
},
|
|
{
|
|
"epoch": 0.11,
|
|
"learning_rate": 1.7500000000000002e-05,
|
|
"loss": 0.6233,
|
|
"step": 14
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"learning_rate": 1.8750000000000002e-05,
|
|
"loss": 0.6264,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 2e-05,
|
|
"loss": 0.617,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.13,
|
|
"learning_rate": 1.9999629591162658e-05,
|
|
"loss": 0.6023,
|
|
"step": 17
|
|
},
|
|
{
|
|
"epoch": 0.14,
|
|
"learning_rate": 1.9998518392091163e-05,
|
|
"loss": 0.5996,
|
|
"step": 18
|
|
},
|
|
{
|
|
"epoch": 0.15,
|
|
"learning_rate": 1.9996666485105115e-05,
|
|
"loss": 0.5939,
|
|
"step": 19
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.999407400739705e-05,
|
|
"loss": 0.5945,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"learning_rate": 1.9990741151022302e-05,
|
|
"loss": 0.5871,
|
|
"step": 21
|
|
},
|
|
{
|
|
"epoch": 0.17,
|
|
"learning_rate": 1.9986668162884763e-05,
|
|
"loss": 0.5918,
|
|
"step": 22
|
|
},
|
|
{
|
|
"epoch": 0.18,
|
|
"learning_rate": 1.9981855344718587e-05,
|
|
"loss": 0.5829,
|
|
"step": 23
|
|
},
|
|
{
|
|
"epoch": 0.19,
|
|
"learning_rate": 1.997630305306586e-05,
|
|
"loss": 0.5715,
|
|
"step": 24
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.997001169925015e-05,
|
|
"loss": 0.5776,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"learning_rate": 1.996298174934608e-05,
|
|
"loss": 0.5663,
|
|
"step": 26
|
|
},
|
|
{
|
|
"epoch": 0.21,
|
|
"learning_rate": 1.9955213724144754e-05,
|
|
"loss": 0.5655,
|
|
"step": 27
|
|
},
|
|
{
|
|
"epoch": 0.22,
|
|
"learning_rate": 1.994670819911521e-05,
|
|
"loss": 0.5593,
|
|
"step": 28
|
|
},
|
|
{
|
|
"epoch": 0.23,
|
|
"learning_rate": 1.9937465804361783e-05,
|
|
"loss": 0.5597,
|
|
"step": 29
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.9927487224577402e-05,
|
|
"loss": 0.5573,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"learning_rate": 1.99167731989929e-05,
|
|
"loss": 0.5485,
|
|
"step": 31
|
|
},
|
|
{
|
|
"epoch": 0.25,
|
|
"learning_rate": 1.990532452132223e-05,
|
|
"loss": 0.5541,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.26,
|
|
"learning_rate": 1.9893142039703662e-05,
|
|
"loss": 0.5403,
|
|
"step": 33
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.9880226656636977e-05,
|
|
"loss": 0.5484,
|
|
"step": 34
|
|
},
|
|
{
|
|
"epoch": 0.27,
|
|
"learning_rate": 1.986657932891657e-05,
|
|
"loss": 0.5448,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"learning_rate": 1.9852201067560607e-05,
|
|
"loss": 0.5421,
|
|
"step": 36
|
|
},
|
|
{
|
|
"epoch": 0.29,
|
|
"learning_rate": 1.98370929377361e-05,
|
|
"loss": 0.542,
|
|
"step": 37
|
|
},
|
|
{
|
|
"epoch": 0.3,
|
|
"learning_rate": 1.982125605868001e-05,
|
|
"loss": 0.5368,
|
|
"step": 38
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9804691603616324e-05,
|
|
"loss": 0.5394,
|
|
"step": 39
|
|
},
|
|
{
|
|
"epoch": 0.31,
|
|
"learning_rate": 1.9787400799669155e-05,
|
|
"loss": 0.5217,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"learning_rate": 1.976938492777182e-05,
|
|
"loss": 0.5323,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 0.33,
|
|
"learning_rate": 1.9750645322571952e-05,
|
|
"loss": 0.534,
|
|
"step": 42
|
|
},
|
|
{
|
|
"epoch": 0.34,
|
|
"learning_rate": 1.973118337233262e-05,
|
|
"loss": 0.5273,
|
|
"step": 43
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9711000518829505e-05,
|
|
"loss": 0.5182,
|
|
"step": 44
|
|
},
|
|
{
|
|
"epoch": 0.35,
|
|
"learning_rate": 1.9690098257244063e-05,
|
|
"loss": 0.5278,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"learning_rate": 1.9668478136052776e-05,
|
|
"loss": 0.5209,
|
|
"step": 46
|
|
},
|
|
{
|
|
"epoch": 0.37,
|
|
"learning_rate": 1.9646141756912437e-05,
|
|
"loss": 0.5156,
|
|
"step": 47
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.962309077454149e-05,
|
|
"loss": 0.5252,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 0.38,
|
|
"learning_rate": 1.9599326896597448e-05,
|
|
"loss": 0.5261,
|
|
"step": 49
|
|
},
|
|
{
|
|
"epoch": 0.39,
|
|
"learning_rate": 1.9574851883550395e-05,
|
|
"loss": 0.5154,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"learning_rate": 1.9549667548552557e-05,
|
|
"loss": 0.5034,
|
|
"step": 51
|
|
},
|
|
{
|
|
"epoch": 0.41,
|
|
"learning_rate": 1.9523775757303975e-05,
|
|
"loss": 0.4981,
|
|
"step": 52
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.949717842791432e-05,
|
|
"loss": 0.5078,
|
|
"step": 53
|
|
},
|
|
{
|
|
"epoch": 0.42,
|
|
"learning_rate": 1.9469877530760753e-05,
|
|
"loss": 0.5134,
|
|
"step": 54
|
|
},
|
|
{
|
|
"epoch": 0.43,
|
|
"learning_rate": 1.9441875088342e-05,
|
|
"loss": 0.5017,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"learning_rate": 1.9413173175128472e-05,
|
|
"loss": 0.4926,
|
|
"step": 56
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.9383773917408644e-05,
|
|
"loss": 0.5033,
|
|
"step": 57
|
|
},
|
|
{
|
|
"epoch": 0.45,
|
|
"learning_rate": 1.9353679493131486e-05,
|
|
"loss": 0.4994,
|
|
"step": 58
|
|
},
|
|
{
|
|
"epoch": 0.46,
|
|
"learning_rate": 1.9322892131745135e-05,
|
|
"loss": 0.4944,
|
|
"step": 59
|
|
},
|
|
{
|
|
"epoch": 0.47,
|
|
"learning_rate": 1.9291414114031744e-05,
|
|
"loss": 0.4844,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"learning_rate": 1.92592477719385e-05,
|
|
"loss": 0.4975,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9226395488404875e-05,
|
|
"loss": 0.4911,
|
|
"step": 62
|
|
},
|
|
{
|
|
"epoch": 0.49,
|
|
"learning_rate": 1.9192859697186105e-05,
|
|
"loss": 0.4829,
|
|
"step": 63
|
|
},
|
|
{
|
|
"epoch": 0.5,
|
|
"learning_rate": 1.9158642882672873e-05,
|
|
"loss": 0.4931,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.51,
|
|
"learning_rate": 1.9123747579707275e-05,
|
|
"loss": 0.4882,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"learning_rate": 1.908817637339503e-05,
|
|
"loss": 0.4851,
|
|
"step": 66
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.9051931898913977e-05,
|
|
"loss": 0.4859,
|
|
"step": 67
|
|
},
|
|
{
|
|
"epoch": 0.53,
|
|
"learning_rate": 1.9015016841318843e-05,
|
|
"loss": 0.4737,
|
|
"step": 68
|
|
},
|
|
{
|
|
"epoch": 0.54,
|
|
"learning_rate": 1.8977433935342338e-05,
|
|
"loss": 0.4823,
|
|
"step": 69
|
|
},
|
|
{
|
|
"epoch": 0.55,
|
|
"learning_rate": 1.8939185965192572e-05,
|
|
"loss": 0.4893,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.890027576434677e-05,
|
|
"loss": 0.4842,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"learning_rate": 1.8860706215341383e-05,
|
|
"loss": 0.4703,
|
|
"step": 72
|
|
},
|
|
{
|
|
"epoch": 0.57,
|
|
"learning_rate": 1.8820480249558538e-05,
|
|
"loss": 0.4813,
|
|
"step": 73
|
|
},
|
|
{
|
|
"epoch": 0.58,
|
|
"learning_rate": 1.8779600847008884e-05,
|
|
"loss": 0.4773,
|
|
"step": 74
|
|
},
|
|
{
|
|
"epoch": 0.59,
|
|
"learning_rate": 1.873807103611081e-05,
|
|
"loss": 0.4703,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.869589389346611e-05,
|
|
"loss": 0.4689,
|
|
"step": 76
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"learning_rate": 1.8653072543632064e-05,
|
|
"loss": 0.4772,
|
|
"step": 77
|
|
},
|
|
{
|
|
"epoch": 0.61,
|
|
"learning_rate": 1.8609610158889943e-05,
|
|
"loss": 0.4702,
|
|
"step": 78
|
|
},
|
|
{
|
|
"epoch": 0.62,
|
|
"learning_rate": 1.8565509959010037e-05,
|
|
"loss": 0.4564,
|
|
"step": 79
|
|
},
|
|
{
|
|
"epoch": 0.63,
|
|
"learning_rate": 1.8520775211013094e-05,
|
|
"loss": 0.4746,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.8475409228928314e-05,
|
|
"loss": 0.464,
|
|
"step": 81
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"learning_rate": 1.842941537354783e-05,
|
|
"loss": 0.4626,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 0.65,
|
|
"learning_rate": 1.8382797052177746e-05,
|
|
"loss": 0.4579,
|
|
"step": 83
|
|
},
|
|
{
|
|
"epoch": 0.66,
|
|
"learning_rate": 1.8335557718385702e-05,
|
|
"loss": 0.4724,
|
|
"step": 84
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.8287700871745036e-05,
|
|
"loss": 0.461,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.67,
|
|
"learning_rate": 1.8239230057575542e-05,
|
|
"loss": 0.4595,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"learning_rate": 1.81901488666808e-05,
|
|
"loss": 0.4603,
|
|
"step": 87
|
|
},
|
|
{
|
|
"epoch": 0.69,
|
|
"learning_rate": 1.814046093508218e-05,
|
|
"loss": 0.4529,
|
|
"step": 88
|
|
},
|
|
{
|
|
"epoch": 0.7,
|
|
"learning_rate": 1.8090169943749477e-05,
|
|
"loss": 0.4513,
|
|
"step": 89
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.8039279618328215e-05,
|
|
"loss": 0.4514,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.71,
|
|
"learning_rate": 1.798779372886365e-05,
|
|
"loss": 0.4554,
|
|
"step": 91
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"learning_rate": 1.7935716089521474e-05,
|
|
"loss": 0.4661,
|
|
"step": 92
|
|
},
|
|
{
|
|
"epoch": 0.73,
|
|
"learning_rate": 1.7883050558305255e-05,
|
|
"loss": 0.4395,
|
|
"step": 93
|
|
},
|
|
{
|
|
"epoch": 0.74,
|
|
"learning_rate": 1.7829801036770628e-05,
|
|
"loss": 0.455,
|
|
"step": 94
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.777597146973627e-05,
|
|
"loss": 0.4512,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.75,
|
|
"learning_rate": 1.7721565844991643e-05,
|
|
"loss": 0.4639,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"learning_rate": 1.7666588193001595e-05,
|
|
"loss": 0.4463,
|
|
"step": 97
|
|
},
|
|
{
|
|
"epoch": 0.77,
|
|
"learning_rate": 1.7611042586607748e-05,
|
|
"loss": 0.451,
|
|
"step": 98
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.7554933140726803e-05,
|
|
"loss": 0.4411,
|
|
"step": 99
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"learning_rate": 1.7498264012045686e-05,
|
|
"loss": 0.4483,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.78,
|
|
"eval_loss": 0.4502714276313782,
|
|
"eval_runtime": 173.0431,
|
|
"eval_samples_per_second": 11.546,
|
|
"eval_steps_per_second": 0.064,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.79,
|
|
"learning_rate": 1.744103939871361e-05,
|
|
"loss": 0.453,
|
|
"step": 101
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"learning_rate": 1.738326354003107e-05,
|
|
"loss": 0.4494,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 0.81,
|
|
"learning_rate": 1.732494071613579e-05,
|
|
"loss": 0.4485,
|
|
"step": 103
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.7266075247685656e-05,
|
|
"loss": 0.4333,
|
|
"step": 104
|
|
},
|
|
{
|
|
"epoch": 0.82,
|
|
"learning_rate": 1.720667149553861e-05,
|
|
"loss": 0.4324,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.83,
|
|
"learning_rate": 1.7146733860429614e-05,
|
|
"loss": 0.442,
|
|
"step": 106
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"learning_rate": 1.70862667826446e-05,
|
|
"loss": 0.4334,
|
|
"step": 107
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.702527474169157e-05,
|
|
"loss": 0.4356,
|
|
"step": 108
|
|
},
|
|
{
|
|
"epoch": 0.85,
|
|
"learning_rate": 1.6963762255968723e-05,
|
|
"loss": 0.4404,
|
|
"step": 109
|
|
},
|
|
{
|
|
"epoch": 0.86,
|
|
"learning_rate": 1.690173388242972e-05,
|
|
"loss": 0.4221,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.87,
|
|
"learning_rate": 1.683919421624611e-05,
|
|
"loss": 0.4175,
|
|
"step": 111
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"learning_rate": 1.677614789046689e-05,
|
|
"loss": 0.4317,
|
|
"step": 112
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.6712599575675318e-05,
|
|
"loss": 0.43,
|
|
"step": 113
|
|
},
|
|
{
|
|
"epoch": 0.89,
|
|
"learning_rate": 1.6648553979642867e-05,
|
|
"loss": 0.4288,
|
|
"step": 114
|
|
},
|
|
{
|
|
"epoch": 0.9,
|
|
"learning_rate": 1.658401584698049e-05,
|
|
"loss": 0.4369,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.91,
|
|
"learning_rate": 1.6518989958787126e-05,
|
|
"loss": 0.4262,
|
|
"step": 116
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"learning_rate": 1.6453481132295507e-05,
|
|
"loss": 0.4361,
|
|
"step": 117
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.6387494220515276e-05,
|
|
"loss": 0.4326,
|
|
"step": 118
|
|
},
|
|
{
|
|
"epoch": 0.93,
|
|
"learning_rate": 1.6321034111873487e-05,
|
|
"loss": 0.4189,
|
|
"step": 119
|
|
},
|
|
{
|
|
"epoch": 0.94,
|
|
"learning_rate": 1.6254105729852466e-05,
|
|
"loss": 0.425,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.95,
|
|
"learning_rate": 1.6186714032625036e-05,
|
|
"loss": 0.4279,
|
|
"step": 121
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.6118864012687246e-05,
|
|
"loss": 0.4239,
|
|
"step": 122
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"learning_rate": 1.6050560696488493e-05,
|
|
"loss": 0.4196,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 0.97,
|
|
"learning_rate": 1.5981809144059167e-05,
|
|
"loss": 0.4206,
|
|
"step": 124
|
|
},
|
|
{
|
|
"epoch": 0.98,
|
|
"learning_rate": 1.5912614448635784e-05,
|
|
"loss": 0.4293,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.99,
|
|
"learning_rate": 1.5842981736283686e-05,
|
|
"loss": 0.4258,
|
|
"step": 126
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"learning_rate": 1.5772916165517275e-05,
|
|
"loss": 0.422,
|
|
"step": 127
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"learning_rate": 1.5702422926917872e-05,
|
|
"loss": 0.4023,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 1.01,
|
|
"learning_rate": 1.5631507242749187e-05,
|
|
"loss": 0.3869,
|
|
"step": 129
|
|
},
|
|
{
|
|
"epoch": 1.02,
|
|
"learning_rate": 1.5560174366570448e-05,
|
|
"loss": 0.4038,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 1.03,
|
|
"learning_rate": 1.5488429582847194e-05,
|
|
"loss": 0.3991,
|
|
"step": 131
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.5416278206559816e-05,
|
|
"loss": 0.402,
|
|
"step": 132
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"learning_rate": 1.5343725582809793e-05,
|
|
"loss": 0.3949,
|
|
"step": 133
|
|
},
|
|
{
|
|
"epoch": 1.05,
|
|
"learning_rate": 1.5270777086423724e-05,
|
|
"loss": 0.3973,
|
|
"step": 134
|
|
},
|
|
{
|
|
"epoch": 1.06,
|
|
"learning_rate": 1.5197438121555159e-05,
|
|
"loss": 0.389,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.512371412128424e-05,
|
|
"loss": 0.4054,
|
|
"step": 136
|
|
},
|
|
{
|
|
"epoch": 1.07,
|
|
"learning_rate": 1.5049610547215205e-05,
|
|
"loss": 0.4026,
|
|
"step": 137
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"learning_rate": 1.4975132889071808e-05,
|
|
"loss": 0.3877,
|
|
"step": 138
|
|
},
|
|
{
|
|
"epoch": 1.09,
|
|
"learning_rate": 1.4900286664290593e-05,
|
|
"loss": 0.3954,
|
|
"step": 139
|
|
},
|
|
{
|
|
"epoch": 1.1,
|
|
"learning_rate": 1.4825077417612187e-05,
|
|
"loss": 0.3879,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.4749510720670506e-05,
|
|
"loss": 0.3897,
|
|
"step": 141
|
|
},
|
|
{
|
|
"epoch": 1.11,
|
|
"learning_rate": 1.4673592171580025e-05,
|
|
"loss": 0.3914,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"learning_rate": 1.4597327394521044e-05,
|
|
"loss": 0.4045,
|
|
"step": 143
|
|
},
|
|
{
|
|
"epoch": 1.13,
|
|
"learning_rate": 1.4520722039323045e-05,
|
|
"loss": 0.3794,
|
|
"step": 144
|
|
},
|
|
{
|
|
"epoch": 1.14,
|
|
"learning_rate": 1.4443781781046135e-05,
|
|
"loss": 0.3857,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.4366512319560642e-05,
|
|
"loss": 0.39,
|
|
"step": 146
|
|
},
|
|
{
|
|
"epoch": 1.15,
|
|
"learning_rate": 1.4288919379124837e-05,
|
|
"loss": 0.3761,
|
|
"step": 147
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"learning_rate": 1.4211008707960897e-05,
|
|
"loss": 0.3761,
|
|
"step": 148
|
|
},
|
|
{
|
|
"epoch": 1.17,
|
|
"learning_rate": 1.4132786077829044e-05,
|
|
"loss": 0.3796,
|
|
"step": 149
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.4054257283599974e-05,
|
|
"loss": 0.3833,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 1.18,
|
|
"learning_rate": 1.3975428142825562e-05,
|
|
"loss": 0.3831,
|
|
"step": 151
|
|
},
|
|
{
|
|
"epoch": 1.19,
|
|
"learning_rate": 1.3896304495307881e-05,
|
|
"loss": 0.3806,
|
|
"step": 152
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"learning_rate": 1.3816892202666591e-05,
|
|
"loss": 0.3761,
|
|
"step": 153
|
|
},
|
|
{
|
|
"epoch": 1.21,
|
|
"learning_rate": 1.373719714790469e-05,
|
|
"loss": 0.377,
|
|
"step": 154
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.3657225234972695e-05,
|
|
"loss": 0.3749,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 1.22,
|
|
"learning_rate": 1.3576982388331258e-05,
|
|
"loss": 0.3817,
|
|
"step": 156
|
|
},
|
|
{
|
|
"epoch": 1.23,
|
|
"learning_rate": 1.3496474552512286e-05,
|
|
"loss": 0.3797,
|
|
"step": 157
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"learning_rate": 1.3415707691678557e-05,
|
|
"loss": 0.3795,
|
|
"step": 158
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.333468778918187e-05,
|
|
"loss": 0.3825,
|
|
"step": 159
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"learning_rate": 1.3253420847119804e-05,
|
|
"loss": 0.3745,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 1.26,
|
|
"learning_rate": 1.3171912885891063e-05,
|
|
"loss": 0.3838,
|
|
"step": 161
|
|
},
|
|
{
|
|
"epoch": 1.27,
|
|
"learning_rate": 1.3090169943749475e-05,
|
|
"loss": 0.3807,
|
|
"step": 162
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"learning_rate": 1.3008198076356678e-05,
|
|
"loss": 0.3673,
|
|
"step": 163
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.2926003356333487e-05,
|
|
"loss": 0.3663,
|
|
"step": 164
|
|
},
|
|
{
|
|
"epoch": 1.29,
|
|
"learning_rate": 1.2843591872810039e-05,
|
|
"loss": 0.3625,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 1.3,
|
|
"learning_rate": 1.2760969730974692e-05,
|
|
"loss": 0.3774,
|
|
"step": 166
|
|
},
|
|
{
|
|
"epoch": 1.31,
|
|
"learning_rate": 1.2678143051621743e-05,
|
|
"loss": 0.3662,
|
|
"step": 167
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"learning_rate": 1.2595117970697998e-05,
|
|
"loss": 0.3597,
|
|
"step": 168
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 1.2511900638848196e-05,
|
|
"loss": 0.3736,
|
|
"step": 169
|
|
},
|
|
{
|
|
"epoch": 1.33,
|
|
"learning_rate": 1.2428497220959359e-05,
|
|
"loss": 0.3515,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 1.34,
|
|
"learning_rate": 1.2344913895704099e-05,
|
|
"loss": 0.3729,
|
|
"step": 171
|
|
},
|
|
{
|
|
"epoch": 1.35,
|
|
"learning_rate": 1.2261156855082882e-05,
|
|
"loss": 0.3582,
|
|
"step": 172
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 1.217723230396532e-05,
|
|
"loss": 0.3607,
|
|
"step": 173
|
|
},
|
|
{
|
|
"epoch": 1.36,
|
|
"learning_rate": 1.2093146459630488e-05,
|
|
"loss": 0.3631,
|
|
"step": 174
|
|
},
|
|
{
|
|
"epoch": 1.37,
|
|
"learning_rate": 1.2008905551306356e-05,
|
|
"loss": 0.3594,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 1.38,
|
|
"learning_rate": 1.19245158197083e-05,
|
|
"loss": 0.3595,
|
|
"step": 176
|
|
},
|
|
{
|
|
"epoch": 1.39,
|
|
"learning_rate": 1.1839983516576802e-05,
|
|
"loss": 0.3513,
|
|
"step": 177
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"learning_rate": 1.1755314904214284e-05,
|
|
"loss": 0.3519,
|
|
"step": 178
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"learning_rate": 1.1670516255021193e-05,
|
|
"loss": 0.3613,
|
|
"step": 179
|
|
},
|
|
{
|
|
"epoch": 1.41,
|
|
"learning_rate": 1.1585593851031346e-05,
|
|
"loss": 0.3538,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 1.42,
|
|
"learning_rate": 1.1500553983446527e-05,
|
|
"loss": 0.348,
|
|
"step": 181
|
|
},
|
|
{
|
|
"epoch": 1.43,
|
|
"learning_rate": 1.1415402952170434e-05,
|
|
"loss": 0.3465,
|
|
"step": 182
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.1330147065341962e-05,
|
|
"loss": 0.3646,
|
|
"step": 183
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"learning_rate": 1.1244792638867895e-05,
|
|
"loss": 0.3539,
|
|
"step": 184
|
|
},
|
|
{
|
|
"epoch": 1.45,
|
|
"learning_rate": 1.1159345995955007e-05,
|
|
"loss": 0.3463,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 1.46,
|
|
"learning_rate": 1.1073813466641633e-05,
|
|
"loss": 0.3653,
|
|
"step": 186
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 1.0988201387328716e-05,
|
|
"loss": 0.3669,
|
|
"step": 187
|
|
},
|
|
{
|
|
"epoch": 1.47,
|
|
"learning_rate": 1.0902516100310412e-05,
|
|
"loss": 0.3471,
|
|
"step": 188
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"learning_rate": 1.0816763953304228e-05,
|
|
"loss": 0.3513,
|
|
"step": 189
|
|
},
|
|
{
|
|
"epoch": 1.49,
|
|
"learning_rate": 1.0730951298980776e-05,
|
|
"loss": 0.3679,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 1.5,
|
|
"learning_rate": 1.0645084494493166e-05,
|
|
"loss": 0.3435,
|
|
"step": 191
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"learning_rate": 1.0559169901006035e-05,
|
|
"loss": 0.3441,
|
|
"step": 192
|
|
},
|
|
{
|
|
"epoch": 1.51,
|
|
"learning_rate": 1.0473213883224321e-05,
|
|
"loss": 0.3436,
|
|
"step": 193
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"learning_rate": 1.0387222808921746e-05,
|
|
"loss": 0.3503,
|
|
"step": 194
|
|
},
|
|
{
|
|
"epoch": 1.53,
|
|
"learning_rate": 1.0301203048469084e-05,
|
|
"loss": 0.3479,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 1.54,
|
|
"learning_rate": 1.0215160974362224e-05,
|
|
"loss": 0.3321,
|
|
"step": 196
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 1.0129102960750092e-05,
|
|
"loss": 0.3433,
|
|
"step": 197
|
|
},
|
|
{
|
|
"epoch": 1.55,
|
|
"learning_rate": 1.0043035382962443e-05,
|
|
"loss": 0.3505,
|
|
"step": 198
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"learning_rate": 9.956964617037559e-06,
|
|
"loss": 0.3436,
|
|
"step": 199
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"learning_rate": 9.87089703924991e-06,
|
|
"loss": 0.3421,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 1.57,
|
|
"eval_loss": 0.3613594174385071,
|
|
"eval_runtime": 172.6903,
|
|
"eval_samples_per_second": 11.57,
|
|
"eval_steps_per_second": 0.064,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 9.78483902563778e-06,
|
|
"loss": 0.3399,
|
|
"step": 201
|
|
},
|
|
{
|
|
"epoch": 1.58,
|
|
"learning_rate": 9.69879695153092e-06,
|
|
"loss": 0.335,
|
|
"step": 202
|
|
},
|
|
{
|
|
"epoch": 1.59,
|
|
"learning_rate": 9.612777191078257e-06,
|
|
"loss": 0.3449,
|
|
"step": 203
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"learning_rate": 9.526786116775682e-06,
|
|
"loss": 0.3385,
|
|
"step": 204
|
|
},
|
|
{
|
|
"epoch": 1.61,
|
|
"learning_rate": 9.440830098993969e-06,
|
|
"loss": 0.3503,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"learning_rate": 9.354915505506839e-06,
|
|
"loss": 0.3442,
|
|
"step": 206
|
|
},
|
|
{
|
|
"epoch": 1.62,
|
|
"learning_rate": 9.269048701019226e-06,
|
|
"loss": 0.3364,
|
|
"step": 207
|
|
},
|
|
{
|
|
"epoch": 1.63,
|
|
"learning_rate": 9.183236046695777e-06,
|
|
"loss": 0.3349,
|
|
"step": 208
|
|
},
|
|
{
|
|
"epoch": 1.64,
|
|
"learning_rate": 9.09748389968959e-06,
|
|
"loss": 0.3384,
|
|
"step": 209
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 9.011798612671286e-06,
|
|
"loss": 0.3438,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 1.65,
|
|
"learning_rate": 8.92618653335837e-06,
|
|
"loss": 0.3363,
|
|
"step": 211
|
|
},
|
|
{
|
|
"epoch": 1.66,
|
|
"learning_rate": 8.840654004044996e-06,
|
|
"loss": 0.3344,
|
|
"step": 212
|
|
},
|
|
{
|
|
"epoch": 1.67,
|
|
"learning_rate": 8.755207361132109e-06,
|
|
"loss": 0.3265,
|
|
"step": 213
|
|
},
|
|
{
|
|
"epoch": 1.68,
|
|
"learning_rate": 8.669852934658042e-06,
|
|
"loss": 0.3449,
|
|
"step": 214
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 8.58459704782957e-06,
|
|
"loss": 0.3308,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 1.69,
|
|
"learning_rate": 8.499446016553475e-06,
|
|
"loss": 0.337,
|
|
"step": 216
|
|
},
|
|
{
|
|
"epoch": 1.7,
|
|
"learning_rate": 8.414406148968657e-06,
|
|
"loss": 0.3272,
|
|
"step": 217
|
|
},
|
|
{
|
|
"epoch": 1.71,
|
|
"learning_rate": 8.32948374497881e-06,
|
|
"loss": 0.3278,
|
|
"step": 218
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"learning_rate": 8.24468509578572e-06,
|
|
"loss": 0.3415,
|
|
"step": 219
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"learning_rate": 8.1600164834232e-06,
|
|
"loss": 0.3221,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 1.73,
|
|
"learning_rate": 8.075484180291702e-06,
|
|
"loss": 0.3195,
|
|
"step": 221
|
|
},
|
|
{
|
|
"epoch": 1.74,
|
|
"learning_rate": 7.991094448693648e-06,
|
|
"loss": 0.324,
|
|
"step": 222
|
|
},
|
|
{
|
|
"epoch": 1.75,
|
|
"learning_rate": 7.906853540369514e-06,
|
|
"loss": 0.3293,
|
|
"step": 223
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 7.822767696034683e-06,
|
|
"loss": 0.3268,
|
|
"step": 224
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"learning_rate": 7.738843144917119e-06,
|
|
"loss": 0.3254,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 1.77,
|
|
"learning_rate": 7.655086104295904e-06,
|
|
"loss": 0.3395,
|
|
"step": 226
|
|
},
|
|
{
|
|
"epoch": 1.78,
|
|
"learning_rate": 7.571502779040646e-06,
|
|
"loss": 0.3177,
|
|
"step": 227
|
|
},
|
|
{
|
|
"epoch": 1.79,
|
|
"learning_rate": 7.4880993611518095e-06,
|
|
"loss": 0.3227,
|
|
"step": 228
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"learning_rate": 7.404882029302003e-06,
|
|
"loss": 0.3237,
|
|
"step": 229
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"learning_rate": 7.321856948378259e-06,
|
|
"loss": 0.3121,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 1.81,
|
|
"learning_rate": 7.239030269025311e-06,
|
|
"loss": 0.3341,
|
|
"step": 231
|
|
},
|
|
{
|
|
"epoch": 1.82,
|
|
"learning_rate": 7.156408127189964e-06,
|
|
"loss": 0.3199,
|
|
"step": 232
|
|
},
|
|
{
|
|
"epoch": 1.83,
|
|
"learning_rate": 7.073996643666516e-06,
|
|
"loss": 0.3185,
|
|
"step": 233
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"learning_rate": 6.991801923643324e-06,
|
|
"loss": 0.3198,
|
|
"step": 234
|
|
},
|
|
{
|
|
"epoch": 1.84,
|
|
"learning_rate": 6.909830056250527e-06,
|
|
"loss": 0.319,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 1.85,
|
|
"learning_rate": 6.8280871141089415e-06,
|
|
"loss": 0.3154,
|
|
"step": 236
|
|
},
|
|
{
|
|
"epoch": 1.86,
|
|
"learning_rate": 6.746579152880201e-06,
|
|
"loss": 0.3231,
|
|
"step": 237
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"learning_rate": 6.665312210818133e-06,
|
|
"loss": 0.3192,
|
|
"step": 238
|
|
},
|
|
{
|
|
"epoch": 1.87,
|
|
"learning_rate": 6.584292308321445e-06,
|
|
"loss": 0.33,
|
|
"step": 239
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"learning_rate": 6.503525447487717e-06,
|
|
"loss": 0.3121,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 1.89,
|
|
"learning_rate": 6.423017611668745e-06,
|
|
"loss": 0.3103,
|
|
"step": 241
|
|
},
|
|
{
|
|
"epoch": 1.9,
|
|
"learning_rate": 6.342774765027309e-06,
|
|
"loss": 0.3078,
|
|
"step": 242
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"learning_rate": 6.262802852095311e-06,
|
|
"loss": 0.3078,
|
|
"step": 243
|
|
},
|
|
{
|
|
"epoch": 1.91,
|
|
"learning_rate": 6.183107797333411e-06,
|
|
"loss": 0.3262,
|
|
"step": 244
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"learning_rate": 6.103695504692122e-06,
|
|
"loss": 0.3225,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 1.93,
|
|
"learning_rate": 6.024571857174443e-06,
|
|
"loss": 0.3076,
|
|
"step": 246
|
|
},
|
|
{
|
|
"epoch": 1.94,
|
|
"learning_rate": 5.94574271640003e-06,
|
|
"loss": 0.3214,
|
|
"step": 247
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 5.867213922170958e-06,
|
|
"loss": 0.3161,
|
|
"step": 248
|
|
},
|
|
{
|
|
"epoch": 1.95,
|
|
"learning_rate": 5.788991292039103e-06,
|
|
"loss": 0.3081,
|
|
"step": 249
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"learning_rate": 5.711080620875165e-06,
|
|
"loss": 0.3137,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 1.97,
|
|
"learning_rate": 5.633487680439362e-06,
|
|
"loss": 0.3104,
|
|
"step": 251
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"learning_rate": 5.5562182189538684e-06,
|
|
"loss": 0.3129,
|
|
"step": 252
|
|
},
|
|
{
|
|
"epoch": 1.98,
|
|
"learning_rate": 5.479277960676959e-06,
|
|
"loss": 0.3049,
|
|
"step": 253
|
|
},
|
|
{
|
|
"epoch": 1.99,
|
|
"learning_rate": 5.40267260547896e-06,
|
|
"loss": 0.3036,
|
|
"step": 254
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"learning_rate": 5.32640782841998e-06,
|
|
"loss": 0.3141,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 2.01,
|
|
"learning_rate": 5.250489279329501e-06,
|
|
"loss": 0.2912,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"learning_rate": 5.174922582387819e-06,
|
|
"loss": 0.3003,
|
|
"step": 257
|
|
},
|
|
{
|
|
"epoch": 2.02,
|
|
"learning_rate": 5.099713335709409e-06,
|
|
"loss": 0.3078,
|
|
"step": 258
|
|
},
|
|
{
|
|
"epoch": 2.03,
|
|
"learning_rate": 5.024867110928193e-06,
|
|
"loss": 0.2958,
|
|
"step": 259
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"learning_rate": 4.950389452784796e-06,
|
|
"loss": 0.2929,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"learning_rate": 4.876285878715764e-06,
|
|
"loss": 0.2984,
|
|
"step": 261
|
|
},
|
|
{
|
|
"epoch": 2.05,
|
|
"learning_rate": 4.802561878444846e-06,
|
|
"loss": 0.3017,
|
|
"step": 262
|
|
},
|
|
{
|
|
"epoch": 2.06,
|
|
"learning_rate": 4.729222913576279e-06,
|
|
"loss": 0.2982,
|
|
"step": 263
|
|
},
|
|
{
|
|
"epoch": 2.07,
|
|
"learning_rate": 4.6562744171902144e-06,
|
|
"loss": 0.2889,
|
|
"step": 264
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"learning_rate": 4.583721793440188e-06,
|
|
"loss": 0.3052,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"learning_rate": 4.5115704171528105e-06,
|
|
"loss": 0.298,
|
|
"step": 266
|
|
},
|
|
{
|
|
"epoch": 2.09,
|
|
"learning_rate": 4.439825633429558e-06,
|
|
"loss": 0.2943,
|
|
"step": 267
|
|
},
|
|
{
|
|
"epoch": 2.1,
|
|
"learning_rate": 4.368492757250814e-06,
|
|
"loss": 0.2875,
|
|
"step": 268
|
|
},
|
|
{
|
|
"epoch": 2.11,
|
|
"learning_rate": 4.297577073082129e-06,
|
|
"loss": 0.2952,
|
|
"step": 269
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"learning_rate": 4.2270838344827285e-06,
|
|
"loss": 0.2846,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"learning_rate": 4.1570182637163155e-06,
|
|
"loss": 0.2952,
|
|
"step": 271
|
|
},
|
|
{
|
|
"epoch": 2.13,
|
|
"learning_rate": 4.087385551364219e-06,
|
|
"loss": 0.2963,
|
|
"step": 272
|
|
},
|
|
{
|
|
"epoch": 2.14,
|
|
"learning_rate": 4.018190855940837e-06,
|
|
"loss": 0.3008,
|
|
"step": 273
|
|
},
|
|
{
|
|
"epoch": 2.15,
|
|
"learning_rate": 3.949439303511512e-06,
|
|
"loss": 0.2894,
|
|
"step": 274
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"learning_rate": 3.881135987312758e-06,
|
|
"loss": 0.3,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"learning_rate": 3.8132859673749688e-06,
|
|
"loss": 0.2929,
|
|
"step": 276
|
|
},
|
|
{
|
|
"epoch": 2.17,
|
|
"learning_rate": 3.7458942701475385e-06,
|
|
"loss": 0.3014,
|
|
"step": 277
|
|
},
|
|
{
|
|
"epoch": 2.18,
|
|
"learning_rate": 3.6789658881265135e-06,
|
|
"loss": 0.2823,
|
|
"step": 278
|
|
},
|
|
{
|
|
"epoch": 2.19,
|
|
"learning_rate": 3.6125057794847286e-06,
|
|
"loss": 0.2895,
|
|
"step": 279
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"learning_rate": 3.546518867704499e-06,
|
|
"loss": 0.2927,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"learning_rate": 3.4810100412128743e-06,
|
|
"loss": 0.2958,
|
|
"step": 281
|
|
},
|
|
{
|
|
"epoch": 2.21,
|
|
"learning_rate": 3.4159841530195127e-06,
|
|
"loss": 0.2908,
|
|
"step": 282
|
|
},
|
|
{
|
|
"epoch": 2.22,
|
|
"learning_rate": 3.3514460203571365e-06,
|
|
"loss": 0.2925,
|
|
"step": 283
|
|
},
|
|
{
|
|
"epoch": 2.23,
|
|
"learning_rate": 3.287400424324687e-06,
|
|
"loss": 0.295,
|
|
"step": 284
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"learning_rate": 3.223852109533112e-06,
|
|
"loss": 0.2844,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"learning_rate": 3.1608057837538976e-06,
|
|
"loss": 0.2962,
|
|
"step": 286
|
|
},
|
|
{
|
|
"epoch": 2.25,
|
|
"learning_rate": 3.098266117570282e-06,
|
|
"loss": 0.2939,
|
|
"step": 287
|
|
},
|
|
{
|
|
"epoch": 2.26,
|
|
"learning_rate": 3.0362377440312783e-06,
|
|
"loss": 0.2857,
|
|
"step": 288
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 2.9747252583084297e-06,
|
|
"loss": 0.2855,
|
|
"step": 289
|
|
},
|
|
{
|
|
"epoch": 2.27,
|
|
"learning_rate": 2.9137332173554043e-06,
|
|
"loss": 0.2886,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 2.28,
|
|
"learning_rate": 2.853266139570391e-06,
|
|
"loss": 0.2844,
|
|
"step": 291
|
|
},
|
|
{
|
|
"epoch": 2.29,
|
|
"learning_rate": 2.793328504461391e-06,
|
|
"loss": 0.2914,
|
|
"step": 292
|
|
},
|
|
{
|
|
"epoch": 2.3,
|
|
"learning_rate": 2.733924752314345e-06,
|
|
"loss": 0.2811,
|
|
"step": 293
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"learning_rate": 2.6750592838642144e-06,
|
|
"loss": 0.2914,
|
|
"step": 294
|
|
},
|
|
{
|
|
"epoch": 2.31,
|
|
"learning_rate": 2.616736459968936e-06,
|
|
"loss": 0.2904,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"learning_rate": 2.5589606012863968e-06,
|
|
"loss": 0.2838,
|
|
"step": 296
|
|
},
|
|
{
|
|
"epoch": 2.33,
|
|
"learning_rate": 2.5017359879543168e-06,
|
|
"loss": 0.284,
|
|
"step": 297
|
|
},
|
|
{
|
|
"epoch": 2.34,
|
|
"learning_rate": 2.4450668592731976e-06,
|
|
"loss": 0.2859,
|
|
"step": 298
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"learning_rate": 2.3889574133922532e-06,
|
|
"loss": 0.2966,
|
|
"step": 299
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"learning_rate": 2.3334118069984102e-06,
|
|
"loss": 0.2881,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 2.35,
|
|
"eval_loss": 0.31201496720314026,
|
|
"eval_runtime": 172.7826,
|
|
"eval_samples_per_second": 11.564,
|
|
"eval_steps_per_second": 0.064,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"learning_rate": 2.2784341550083577e-06,
|
|
"loss": 0.2876,
|
|
"step": 301
|
|
},
|
|
{
|
|
"epoch": 2.37,
|
|
"learning_rate": 2.224028530263733e-06,
|
|
"loss": 0.28,
|
|
"step": 302
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"learning_rate": 2.170198963229372e-06,
|
|
"loss": 0.2797,
|
|
"step": 303
|
|
},
|
|
{
|
|
"epoch": 2.38,
|
|
"learning_rate": 2.116949441694748e-06,
|
|
"loss": 0.2896,
|
|
"step": 304
|
|
},
|
|
{
|
|
"epoch": 2.39,
|
|
"learning_rate": 2.064283910478527e-06,
|
|
"loss": 0.281,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"learning_rate": 2.012206271136353e-06,
|
|
"loss": 0.2774,
|
|
"step": 306
|
|
},
|
|
{
|
|
"epoch": 2.41,
|
|
"learning_rate": 1.960720381671789e-06,
|
|
"loss": 0.282,
|
|
"step": 307
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"learning_rate": 1.9098300562505266e-06,
|
|
"loss": 0.2847,
|
|
"step": 308
|
|
},
|
|
{
|
|
"epoch": 2.42,
|
|
"learning_rate": 1.8595390649178214e-06,
|
|
"loss": 0.2832,
|
|
"step": 309
|
|
},
|
|
{
|
|
"epoch": 2.43,
|
|
"learning_rate": 1.8098511333192026e-06,
|
|
"loss": 0.2945,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"learning_rate": 1.7607699424244583e-06,
|
|
"loss": 0.2903,
|
|
"step": 311
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"learning_rate": 1.712299128254965e-06,
|
|
"loss": 0.2907,
|
|
"step": 312
|
|
},
|
|
{
|
|
"epoch": 2.45,
|
|
"learning_rate": 1.6644422816143024e-06,
|
|
"loss": 0.2742,
|
|
"step": 313
|
|
},
|
|
{
|
|
"epoch": 2.46,
|
|
"learning_rate": 1.6172029478222595e-06,
|
|
"loss": 0.2776,
|
|
"step": 314
|
|
},
|
|
{
|
|
"epoch": 2.47,
|
|
"learning_rate": 1.5705846264521728e-06,
|
|
"loss": 0.2909,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"learning_rate": 1.5245907710716912e-06,
|
|
"loss": 0.2878,
|
|
"step": 316
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"learning_rate": 1.4792247889869072e-06,
|
|
"loss": 0.2877,
|
|
"step": 317
|
|
},
|
|
{
|
|
"epoch": 2.49,
|
|
"learning_rate": 1.4344900409899643e-06,
|
|
"loss": 0.2868,
|
|
"step": 318
|
|
},
|
|
{
|
|
"epoch": 2.5,
|
|
"learning_rate": 1.3903898411100568e-06,
|
|
"loss": 0.2807,
|
|
"step": 319
|
|
},
|
|
{
|
|
"epoch": 2.51,
|
|
"learning_rate": 1.3469274563679401e-06,
|
|
"loss": 0.2733,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"learning_rate": 1.30410610653389e-06,
|
|
"loss": 0.2957,
|
|
"step": 321
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.261928963889194e-06,
|
|
"loss": 0.2811,
|
|
"step": 322
|
|
},
|
|
{
|
|
"epoch": 2.53,
|
|
"learning_rate": 1.2203991529911197e-06,
|
|
"loss": 0.2748,
|
|
"step": 323
|
|
},
|
|
{
|
|
"epoch": 2.54,
|
|
"learning_rate": 1.1795197504414657e-06,
|
|
"loss": 0.2782,
|
|
"step": 324
|
|
},
|
|
{
|
|
"epoch": 2.55,
|
|
"learning_rate": 1.1392937846586216e-06,
|
|
"loss": 0.2914,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"learning_rate": 1.0997242356532335e-06,
|
|
"loss": 0.2871,
|
|
"step": 326
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"learning_rate": 1.0608140348074292e-06,
|
|
"loss": 0.2839,
|
|
"step": 327
|
|
},
|
|
{
|
|
"epoch": 2.57,
|
|
"learning_rate": 1.022566064657663e-06,
|
|
"loss": 0.2869,
|
|
"step": 328
|
|
},
|
|
{
|
|
"epoch": 2.58,
|
|
"learning_rate": 9.849831586811597e-07,
|
|
"loss": 0.2968,
|
|
"step": 329
|
|
},
|
|
{
|
|
"epoch": 2.59,
|
|
"learning_rate": 9.48068101086026e-07,
|
|
"loss": 0.276,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"learning_rate": 9.118236266049707e-07,
|
|
"loss": 0.2933,
|
|
"step": 331
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"learning_rate": 8.762524202927281e-07,
|
|
"loss": 0.2796,
|
|
"step": 332
|
|
},
|
|
{
|
|
"epoch": 2.61,
|
|
"learning_rate": 8.413571173271295e-07,
|
|
"loss": 0.2657,
|
|
"step": 333
|
|
},
|
|
{
|
|
"epoch": 2.62,
|
|
"learning_rate": 8.071403028138969e-07,
|
|
"loss": 0.2775,
|
|
"step": 334
|
|
},
|
|
{
|
|
"epoch": 2.63,
|
|
"learning_rate": 7.736045115951252e-07,
|
|
"loss": 0.2795,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"learning_rate": 7.40752228061502e-07,
|
|
"loss": 0.2739,
|
|
"step": 336
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"learning_rate": 7.08585885968257e-07,
|
|
"loss": 0.2754,
|
|
"step": 337
|
|
},
|
|
{
|
|
"epoch": 2.65,
|
|
"learning_rate": 6.77107868254867e-07,
|
|
"loss": 0.2868,
|
|
"step": 338
|
|
},
|
|
{
|
|
"epoch": 2.66,
|
|
"learning_rate": 6.463205068685174e-07,
|
|
"loss": 0.2704,
|
|
"step": 339
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"learning_rate": 6.162260825913591e-07,
|
|
"loss": 0.2795,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 2.67,
|
|
"learning_rate": 5.868268248715292e-07,
|
|
"loss": 0.2829,
|
|
"step": 341
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"learning_rate": 5.581249116580068e-07,
|
|
"loss": 0.2733,
|
|
"step": 342
|
|
},
|
|
{
|
|
"epoch": 2.69,
|
|
"learning_rate": 5.301224692392482e-07,
|
|
"loss": 0.2784,
|
|
"step": 343
|
|
},
|
|
{
|
|
"epoch": 2.7,
|
|
"learning_rate": 5.028215720856821e-07,
|
|
"loss": 0.2755,
|
|
"step": 344
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"learning_rate": 4.762242426960262e-07,
|
|
"loss": 0.2708,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 2.71,
|
|
"learning_rate": 4.503324514474483e-07,
|
|
"loss": 0.2732,
|
|
"step": 346
|
|
},
|
|
{
|
|
"epoch": 2.72,
|
|
"learning_rate": 4.2514811644960743e-07,
|
|
"loss": 0.2825,
|
|
"step": 347
|
|
},
|
|
{
|
|
"epoch": 2.73,
|
|
"learning_rate": 4.006731034025546e-07,
|
|
"loss": 0.2674,
|
|
"step": 348
|
|
},
|
|
{
|
|
"epoch": 2.74,
|
|
"learning_rate": 3.769092254585138e-07,
|
|
"loss": 0.2798,
|
|
"step": 349
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"learning_rate": 3.538582430875659e-07,
|
|
"loss": 0.2896,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 2.75,
|
|
"learning_rate": 3.3152186394722506e-07,
|
|
"loss": 0.2852,
|
|
"step": 351
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"learning_rate": 3.099017427559392e-07,
|
|
"loss": 0.2837,
|
|
"step": 352
|
|
},
|
|
{
|
|
"epoch": 2.77,
|
|
"learning_rate": 2.889994811704966e-07,
|
|
"loss": 0.2707,
|
|
"step": 353
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 2.688166276673809e-07,
|
|
"loss": 0.2897,
|
|
"step": 354
|
|
},
|
|
{
|
|
"epoch": 2.78,
|
|
"learning_rate": 2.493546774280531e-07,
|
|
"loss": 0.2806,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 2.79,
|
|
"learning_rate": 2.3061507222818303e-07,
|
|
"loss": 0.277,
|
|
"step": 356
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"learning_rate": 2.1259920033084746e-07,
|
|
"loss": 0.2662,
|
|
"step": 357
|
|
},
|
|
{
|
|
"epoch": 2.81,
|
|
"learning_rate": 1.9530839638367993e-07,
|
|
"loss": 0.2699,
|
|
"step": 358
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"learning_rate": 1.7874394131999428e-07,
|
|
"loss": 0.2788,
|
|
"step": 359
|
|
},
|
|
{
|
|
"epoch": 2.82,
|
|
"learning_rate": 1.6290706226390286e-07,
|
|
"loss": 0.2828,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 2.83,
|
|
"learning_rate": 1.4779893243939358e-07,
|
|
"loss": 0.2792,
|
|
"step": 361
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"learning_rate": 1.334206710834296e-07,
|
|
"loss": 0.2796,
|
|
"step": 362
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.1977334336302437e-07,
|
|
"loss": 0.2859,
|
|
"step": 363
|
|
},
|
|
{
|
|
"epoch": 2.85,
|
|
"learning_rate": 1.068579602963371e-07,
|
|
"loss": 0.2765,
|
|
"step": 364
|
|
},
|
|
{
|
|
"epoch": 2.86,
|
|
"learning_rate": 9.467547867777261e-08,
|
|
"loss": 0.2852,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 2.87,
|
|
"learning_rate": 8.322680100710023e-08,
|
|
"loss": 0.2692,
|
|
"step": 366
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"learning_rate": 7.251277542259849e-08,
|
|
"loss": 0.282,
|
|
"step": 367
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"learning_rate": 6.253419563821971e-08,
|
|
"loss": 0.2721,
|
|
"step": 368
|
|
},
|
|
{
|
|
"epoch": 2.89,
|
|
"learning_rate": 5.3291800884789356e-08,
|
|
"loss": 0.2697,
|
|
"step": 369
|
|
},
|
|
{
|
|
"epoch": 2.9,
|
|
"learning_rate": 4.478627585524753e-08,
|
|
"loss": 0.2789,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 2.91,
|
|
"learning_rate": 3.701825065392184e-08,
|
|
"loss": 0.2872,
|
|
"step": 371
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"learning_rate": 2.998830074984915e-08,
|
|
"loss": 0.2853,
|
|
"step": 372
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"learning_rate": 2.369694693414304e-08,
|
|
"loss": 0.2803,
|
|
"step": 373
|
|
},
|
|
{
|
|
"epoch": 2.93,
|
|
"learning_rate": 1.8144655281413515e-08,
|
|
"loss": 0.2799,
|
|
"step": 374
|
|
},
|
|
{
|
|
"epoch": 2.94,
|
|
"learning_rate": 1.3331837115241331e-08,
|
|
"loss": 0.2872,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 2.95,
|
|
"learning_rate": 9.25884897770013e-09,
|
|
"loss": 0.2708,
|
|
"step": 376
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"learning_rate": 5.925992602952013e-09,
|
|
"loss": 0.2847,
|
|
"step": 377
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"learning_rate": 3.333514894887646e-09,
|
|
"loss": 0.2752,
|
|
"step": 378
|
|
},
|
|
{
|
|
"epoch": 2.97,
|
|
"learning_rate": 1.4816079088375567e-09,
|
|
"loss": 0.2837,
|
|
"step": 379
|
|
},
|
|
{
|
|
"epoch": 2.98,
|
|
"learning_rate": 3.7040883734462685e-10,
|
|
"loss": 0.2769,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.28,
|
|
"step": 381
|
|
},
|
|
{
|
|
"epoch": 2.99,
|
|
"step": 381,
|
|
"total_flos": 5.847367827681444e+18,
|
|
"train_loss": 0.3848528464635213,
|
|
"train_runtime": 103573.3214,
|
|
"train_samples_per_second": 2.834,
|
|
"train_steps_per_second": 0.004
|
|
}
|
|
],
|
|
"max_steps": 381,
|
|
"num_train_epochs": 3,
|
|
"total_flos": 5.847367827681444e+18,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|