In paper, in additional details chapter A.2. Fine-tuning Details, it is mentioned that after fine tuning, text embedding can be stored in the classification layer. Which part of the code is doing this. How we can do it if in code base it is not mentioned.
In paper, in additional details chapter A.2. Fine-tuning Details, it is mentioned that after fine tuning, text embedding can be stored in the classification layer. Which part of the code is doing this. How we can do it if in code base it is not mentioned.
Thanks