Hello, I would like to know how to adjust the embedding of category conditions when fine-tuning DiT on the target dataset, such as CUB200, which only has 200 categories. However, DiT pre trains on data with 1000 categories and their number of categories is inconsistent. Looking forward to the author's reply.
Hello, I would like to know how to adjust the embedding of category conditions when fine-tuning DiT on the target dataset, such as CUB200, which only has 200 categories. However, DiT pre trains on data with 1000 categories and their number of categories is inconsistent. Looking forward to the author's reply.