llama3-8b-kasa-coding-11-v1 / train_results.json
chansung's picture
Model save
de1b350 verified
raw
history blame contribute delete
236 Bytes
{
"epoch": 1.0,
"total_flos": 8.089579620799611e+17,
"train_loss": 1.4743173070197557,
"train_runtime": 700.8271,
"train_samples": 116368,
"train_samples_per_second": 49.884,
"train_steps_per_second": 0.195
}