search

OFA-Sys / Chinese-CLIP

Chinese version of CLIP which achieves Chinese cross-modal retrieval and representation generation.
MIT License
4.04k stars 422 forks source link

evaluation datasets #17

Open rom1504 opened 1 year ago

rom1504 commented 1 year ago

could you please provide the muge, flickr 30 cn and coco cn you used for eval ?

thanks

yangapku commented 1 year ago

Hi, for the raw data of MUGE-Retrieval, you can refer to this link to download the dataset (need to apply first, typically very soon). The raw data of Flickr30K-CN can be found and downloaded at this repo. We have also provided download links of our preprocessed data for these two datasets in the readme (we have provided an English version of readme :) ), which can be directly used by Chinese-CLIP code. Downloading the raw data of COCO-CN needs to get permission from its original author, you need to refer to this repo and fill out its Google Form first, which may take some time. We do not provide the preprocessed version of it due to this reason.