{ "effective_tokens_per_sec": 2197.929095468877, "epoch": 4.0, "total_flos": 3.61934117404672e+18, "train_loss": 0.7229323635552207, "train_runtime": 32622.9814, "train_samples_per_second": 1.707, "train_steps_per_second": 0.028 }