Add multi-modal loss

[philippmwirth]

Add multi-modal loss

As discussed in #249, adding the loss function as described in CLIP would enable users to work with multi-modal datasets.

The pseudo-code (from the paper) is:

[aymuos15]

Hi @philippmwirth. I would like to work on this issue. Could you please guide me as to how I can proceed to do so?
Thank you.

[philippmwirth]

Hi @aymuos15 that's great! I will assign the issue to you.

To start off, I would recommend you fork the repo and read the guide on how to contribute.

Let me know if we can help you getting started :)

[aymuos15]

Thank you, I will get back to you soon!

[philippmwirth]

Hi @aymuos15, how is it going? Can we help you in any way?

[iamharsha1999]

Hi @philippmwirth Can you assign me this issue ? I would like to try this out.