deepmath_32B / train_results.json
marianna13's picture
Upload folder using huggingface_hub
e821056 verified
raw
history blame
210 Bytes
{
"epoch": 5.0,
"total_flos": 5.689611896487936e+16,
"train_loss": 0.22220699680913794,
"train_runtime": 41938.1616,
"train_samples_per_second": 12.314,
"train_steps_per_second": 0.024
}