Qwen-2.5-7B-Simple-RL / all_results.json
zxnstc's picture
End of training
d2bd3c0 verified
{
"eval_loss": 0.01268028374761343,
"eval_runtime": 40196.1099,
"eval_samples": 5000,
"eval_samples_per_second": 0.124,
"eval_steps_per_second": 0.009,
"total_flos": 0.0,
"train_loss": 2.683533102224817,
"train_runtime": 211196.195,
"train_samples": 7500,
"train_samples_per_second": 0.036,
"train_steps_per_second": 0.002
}