Select alternative CLIP implementation

Explore alternative implementations to fine-tune CLIP more easily
* CLIP-italian [github](https://github.com/clip-italian/clip-italian/)
* VisionTextDualEncoder [huggingface](https://huggingface.co/docs/transformers/model_doc/vision-text-dual-encoder)
* OpenCLIP [github](https://github.com/mlfoundations/open_clip) | [huggingface](https://huggingface.co/docs/hub/open_clip)