qwen-3b-ot3-8k-qwq-r1-ilr / train_results.json
RZ412's picture
Add files using upload-large-folder tool
5d78027 verified
raw
history blame contribute delete
204 Bytes
{
"epoch": 5.0,
"total_flos": 578141297639424.0,
"train_loss": 0.7460250891673588,
"train_runtime": 98851.1844,
"train_samples_per_second": 0.421,
"train_steps_per_second": 0.053
}