VLM

Poor Performance of Large Models on Specific Tasks