Qwen2.5-0.5B-Instruct-Simple-RL / train_results.json
bryanlincoln's picture
Model save
772a288 verified
raw
history blame
200 Bytes
{
"total_flos": 0.0,
"train_loss": 0.7647797629284884,
"train_runtime": 290057.3567,
"train_samples": 7500,
"train_samples_per_second": 0.026,
"train_steps_per_second": 0.003
}