mamba_0_75_dpo_ep3 / train_results.json
Junxiong Wang
add models
7005c37
raw
history blame contribute delete
219 Bytes
{
"epoch": 3.0,
"total_flos": 0.0,
"train_loss": 0.22357279954321543,
"train_runtime": 32339.7878,
"train_samples": 61134,
"train_samples_per_second": 5.671,
"train_steps_per_second": 0.177
}