I am looking into using CLIP-R precision to evaluate a model's performance. It seems that the compositional split generation is not present, is that right? If so, I'm wondering if it is a possibility for your team to include the Spacy parsing and the subsequent generation of data.pkl for a new set of captions?
Hi there,
I am looking into using CLIP-R precision to evaluate a model's performance. It seems that the compositional split generation is not present, is that right? If so, I'm wondering if it is a possibility for your team to include the Spacy parsing and the subsequent generation of data.pkl for a new set of captions?
Thanks.