Hephaestus: Improving Fundamental Agent Capabilities of Large Language Models through Continual Pre-Training

Open in new window