227 lines
5.9 KiB
JSON
227 lines
5.9 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.4979031179724063,
|
|
"eval_steps": 99999999,
|
|
"global_step": 2048,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.000243116756822464,
|
|
"grad_norm": 245.63973999023438,
|
|
"learning_rate": 5e-05,
|
|
"loss": 175.7636,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.000243116756822464,
|
|
"eval_accuracy": 0.03065178878801227,
|
|
"eval_loss": 10.476305961608887,
|
|
"eval_runtime": 122.5372,
|
|
"eval_samples_per_second": 51.397,
|
|
"eval_steps_per_second": 6.431,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.000486233513644928,
|
|
"grad_norm": 113.21363067626953,
|
|
"learning_rate": 4.998784637822071e-05,
|
|
"loss": 161.0339,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.000486233513644928,
|
|
"eval_accuracy": 0.03065178878801227,
|
|
"eval_loss": 10.297815322875977,
|
|
"eval_runtime": 120.5427,
|
|
"eval_samples_per_second": 52.247,
|
|
"eval_steps_per_second": 6.537,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.000972467027289856,
|
|
"grad_norm": 65.4986801147461,
|
|
"learning_rate": 4.996353913466213e-05,
|
|
"loss": 156.102,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.000972467027289856,
|
|
"eval_accuracy": 0.03066870675635363,
|
|
"eval_loss": 10.12234878540039,
|
|
"eval_runtime": 134.3118,
|
|
"eval_samples_per_second": 46.891,
|
|
"eval_steps_per_second": 5.867,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.001944934054579712,
|
|
"grad_norm": 38.307884216308594,
|
|
"learning_rate": 4.991492464754497e-05,
|
|
"loss": 152.2924,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.001944934054579712,
|
|
"eval_accuracy": 0.030665912963416524,
|
|
"eval_loss": 10.031266212463379,
|
|
"eval_runtime": 123.2321,
|
|
"eval_samples_per_second": 51.107,
|
|
"eval_steps_per_second": 6.394,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.003889868109159424,
|
|
"grad_norm": 33.05560302734375,
|
|
"learning_rate": 4.981769567331065e-05,
|
|
"loss": 148.7256,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.003889868109159424,
|
|
"eval_accuracy": 0.03471380850784451,
|
|
"eval_loss": 9.771815299987793,
|
|
"eval_runtime": 121.9704,
|
|
"eval_samples_per_second": 51.635,
|
|
"eval_steps_per_second": 6.461,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.007779736218318848,
|
|
"grad_norm": 33.15456771850586,
|
|
"learning_rate": 4.9623237724842005e-05,
|
|
"loss": 141.2284,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.007779736218318848,
|
|
"eval_accuracy": 0.04097904438002165,
|
|
"eval_loss": 9.279535293579102,
|
|
"eval_runtime": 124.7723,
|
|
"eval_samples_per_second": 50.476,
|
|
"eval_steps_per_second": 6.316,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.015559472436637696,
|
|
"grad_norm": 20.02243423461914,
|
|
"learning_rate": 4.9234321827904715e-05,
|
|
"loss": 127.002,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.015559472436637696,
|
|
"eval_accuracy": 0.06789103090028115,
|
|
"eval_loss": 8.393986701965332,
|
|
"eval_runtime": 137.4464,
|
|
"eval_samples_per_second": 45.821,
|
|
"eval_steps_per_second": 5.733,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 0.031118944873275392,
|
|
"grad_norm": 28.861949920654297,
|
|
"learning_rate": 4.8456490034030144e-05,
|
|
"loss": 116.9628,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 0.031118944873275392,
|
|
"eval_accuracy": 0.099492243654753,
|
|
"eval_loss": 7.541510105133057,
|
|
"eval_runtime": 137.5865,
|
|
"eval_samples_per_second": 45.775,
|
|
"eval_steps_per_second": 5.727,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 0.062237889746550784,
|
|
"grad_norm": 32.826515197753906,
|
|
"learning_rate": 4.6900826446280993e-05,
|
|
"loss": 103.7502,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 0.062237889746550784,
|
|
"eval_accuracy": 0.11699364908781108,
|
|
"eval_loss": 6.44133186340332,
|
|
"eval_runtime": 127.4576,
|
|
"eval_samples_per_second": 49.413,
|
|
"eval_steps_per_second": 6.182,
|
|
"step": 256
|
|
},
|
|
{
|
|
"epoch": 0.12447577949310157,
|
|
"grad_norm": 24.268686294555664,
|
|
"learning_rate": 4.37894992707827e-05,
|
|
"loss": 99.1234,
|
|
"step": 512
|
|
},
|
|
{
|
|
"epoch": 0.12447577949310157,
|
|
"eval_accuracy": 0.16933101386435265,
|
|
"eval_loss": 5.398009300231934,
|
|
"eval_runtime": 135.7602,
|
|
"eval_samples_per_second": 46.391,
|
|
"eval_steps_per_second": 5.804,
|
|
"step": 512
|
|
},
|
|
{
|
|
"epoch": 0.24895155898620314,
|
|
"grad_norm": 8.205699920654297,
|
|
"learning_rate": 3.75668449197861e-05,
|
|
"loss": 84.3663,
|
|
"step": 1024
|
|
},
|
|
{
|
|
"epoch": 0.24895155898620314,
|
|
"eval_accuracy": 0.21648232289603334,
|
|
"eval_loss": 4.741060733795166,
|
|
"eval_runtime": 116.5089,
|
|
"eval_samples_per_second": 54.056,
|
|
"eval_steps_per_second": 6.763,
|
|
"step": 1024
|
|
},
|
|
{
|
|
"epoch": 0.4979031179724063,
|
|
"grad_norm": 10.254305839538574,
|
|
"learning_rate": 2.51215362177929e-05,
|
|
"loss": 69.5639,
|
|
"step": 2048
|
|
},
|
|
{
|
|
"epoch": 0.4979031179724063,
|
|
"eval_accuracy": 0.20927154332536482,
|
|
"eval_loss": 4.547865390777588,
|
|
"eval_runtime": 135.1009,
|
|
"eval_samples_per_second": 46.617,
|
|
"eval_steps_per_second": 5.833,
|
|
"step": 2048
|
|
}
|
|
],
|
|
"logging_steps": 99999999,
|
|
"max_steps": 4114,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 99999999,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.36992276873216e+17,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|