This PR contains a description on how to use finetuner and tutorials to fine-tune SBERT and CLIP models in an unsupervised way:
The overall description is placed in the 13th Kata and the code is placed in a new finetuner/ directory together with a README which describes how to execute it.
I did not include evaluation results. I also have some scripts on my machine to evaluate the models, but especially for CLIP they are a bit slow without a GPU machine
This PR contains a description on how to use finetuner and tutorials to fine-tune SBERT and CLIP models in an unsupervised way: The overall description is placed in the 13th Kata and the code is placed in a new
finetuner/
directory together with a README which describes how to execute it.