{ "effective_tokens_per_sec": 2485.689782693279, "epoch": 3.0, "total_flos": 2.50959750918203e+18, "train_loss": 1.167248018547183, "train_runtime": 8368.8806, "train_samples_per_second": 3.212, "train_steps_per_second": 0.201 }