Files
HT-phase_scale-Phi-140k-phase2/all_results.json

8 lines
206 B
JSON
Raw Normal View History

{
"epoch": 3.0,
"total_flos": 2307529338208256.0,
"train_loss": 0.3301271531536556,
"train_runtime": 285378.9111,
"train_samples_per_second": 1.141,
"train_steps_per_second": 0.024
}