Add support of zero-shot classification/retrieval evaluation of captioning models using log-likelihood scoring, e.g. LLaVA, and others - Githubissues

LAION-AI / CLIP_benchmark

CLIP-like model evaluation

MIT License

623 stars 80 forks source link

Add support of zero-shot classification/retrieval evaluation of captioning models using log-likelihood scoring, e.g. LLaVA, and others #115

Open mehdidc opened 10 months ago