{ "effective_tokens_per_sec": 4731.714930074944, "epoch": 4.0, "total_flos": 3.800554784105169e+18, "train_loss": 0.9838281438538903, "train_runtime": 15715.7329, "train_samples_per_second": 4.062, "train_steps_per_second": 0.034 }