Qwen-2.5-7B-GRA-WizardLM / train_results.json
GX-XinGao's picture
Upload initial model
f962083 verified
{
"epoch": 0.998849252013809,
"total_flos": 2.9080852357930025e+18,
"train_loss": 0.39522024055230454,
"train_runtime": 4203.0638,
"train_samples_per_second": 13.229,
"train_steps_per_second": 0.052
}