qwen2-0.5b-sft / all_results.json
hZzy's picture
End of training
911cc23 verified
raw
history blame contribute delete
417 Bytes
{
"epoch": 0.9993049349617714,
"eval_loss": 1.5326974391937256,
"eval_runtime": 239.605,
"eval_samples": 23109,
"eval_samples_per_second": 111.617,
"eval_steps_per_second": 4.653,
"total_flos": 106140763422720.0,
"train_loss": 1.5477893754295022,
"train_runtime": 7899.7649,
"train_samples": 207864,
"train_samples_per_second": 30.594,
"train_steps_per_second": 0.159
}