llama-3-8B-lora-sft / all_results.json
ytcheng's picture
End of training
215ed2c verified
raw
history blame contribute delete
365 Bytes
{
"epoch": 2.9986300118732303,
"eval_loss": 1.9779435396194458,
"eval_runtime": 105.0149,
"eval_samples_per_second": 11.589,
"eval_steps_per_second": 11.589,
"total_flos": 1.3798092749438976e+17,
"train_loss": 2.123500373163651,
"train_runtime": 11771.2816,
"train_samples_per_second": 2.79,
"train_steps_per_second": 0.349
}