clip-score
clip-score copied to clipboard
Quick scripts to calculate CLIP text-image similarity
Dear author, Thanks so much for the great contribution to the community, in recent SD benchmark models, they often mention the subject fidelity using CLIP-I and DINO, for prompt fidelity...
Thank you for your excellent tool. When using it for biomedical CLIP score calculation, I took the liberty of making slight modifications to some code in the `src/clip_score/clip_score.py` file by...
我的命令行输入的是`python -m clip-score ./target11 ./text11` 当我对应放4张图片对的时候,能跑出来结果 但是当我加一对,变成5张图片对的时候,就炸了。 下面是设置--num-workers 0的报错信息。 这可咋整啊,我可有足足几千张img-txt对,难道只能测4对吗。(但是当我将后面的./text11换成字符串时,对于我的全部几千张img都能一下测出)而且我测出来的clip-score是对的吗?我看好多issue里面说范围是[0,100]。