Qwen-2.5-7B-Simple-RL-debug / all_results.json
alterego238's picture
End of training
0479954 verified
raw
history blame contribute delete
364 Bytes
{
"eval_loss": 0.00011165998876094818,
"eval_runtime": 77.0085,
"eval_samples": 8,
"eval_samples_per_second": 0.104,
"eval_steps_per_second": 0.026,
"total_flos": 0.0,
"train_loss": -0.062465423718094826,
"train_runtime": 165.9317,
"train_samples": 8,
"train_samples_per_second": 0.048,
"train_steps_per_second": 0.048
}