Files
oh-dcft-v3.1-gpt-4o-mini-qwen/train_results.json

8 lines
205 B
JSON
Raw Permalink Normal View History

{
"epoch": 3.0,
"total_flos": 2759937528692736.0,
"train_loss": 0.5805289232719545,
"train_runtime": 71924.3439,
"train_samples_per_second": 9.369,
"train_steps_per_second": 0.073
}