Qwen2.5-3B-GRPO-Countdown / train_results.json
JeffP111's picture
Model save
6c6d28f verified
raw
history blame contribute delete
200 Bytes
{
"total_flos": 0.0,
"train_loss": 253112.25296123006,
"train_runtime": 92260.7677,
"train_samples": 45000,
"train_samples_per_second": 0.488,
"train_steps_per_second": 0.004
}