Qwen2.5-7B-Instruct-GRPO / train_results.json
susumuota's picture
Model save
c6076ce verified
raw
history blame contribute delete
199 Bytes
{
"total_flos": 0.0,
"train_loss": 0.10446281573767292,
"train_runtime": 6557.3192,
"train_samples": 7500,
"train_samples_per_second": 1.144,
"train_steps_per_second": 0.009
}