mamba_0_875_dpo_ep3 / train_results.json
Junxiong Wang
add models
7ca5856
raw
history blame contribute delete
219 Bytes
{
"epoch": 3.0,
"total_flos": 0.0,
"train_loss": 0.22836191894055405,
"train_runtime": 34822.6853,
"train_samples": 61134,
"train_samples_per_second": 5.267,
"train_steps_per_second": 0.165
}