{ "effective_tokens_per_sec": 7368.548440514258, "epoch": 4.0, "total_flos": 3.5845266747435254e+18, "train_loss": 0.8258528457846233, "train_runtime": 9426.6588, "train_samples_per_second": 6.021, "train_steps_per_second": 0.201 }