SCL is a vision-language model pre-trained on COCO, VG, CC3M, SBU. 

The code of SCL can be found at https://github.com/IIGROUP/SCL.

We have uploaded pre-trained model weights. The fine-tuned model weights will come soon.

*model explanation*
> GLSCL-100k: pre-training with MLM, CL, ITM, MGSC, MLTC
> MGSC-100k: pre-training with MLM, CL, ITM, MGSC