Short survey of the most prominent contrastive learning techniques
In the first formula, when Y=1, the similar loss is zeroed, so I think it's the opposite (Y=1 means dissimilar)
Thank you, you're right, ill fix soon
Thank *you*, I enjoy your writing
In the first formula, when Y=1, the similar loss is zeroed, so I think it's the opposite (Y=1 means dissimilar)
Thank you, you're right, ill fix soon
Thank *you*, I enjoy your writing