Running 128 Open FinLLM Leaderboard π₯ 128 Compare performance of large language models on financial tasks