qwen2.5-0.5b-expo-DPO-ES-TRY3 / train_results.json
hZzy's picture
Training in progress, step 530
6a3238b verified
raw
history blame
219 Bytes
{
"epoch": 2.9976381672177608,
"total_flos": 0.0,
"train_loss": 0.0,
"train_runtime": 9.3184,
"train_samples": 50802,
"train_samples_per_second": 16355.433,
"train_steps_per_second": 56.662
}