Cathy

Results 3 issues of Cathy

I have a question because there is a difference in the value of the author's top-k recall and my top-k recall. When validating with the checkpoint posted by the author,...

In model/OSCAR/OSCAR_CIRPLANT.py part of “def comput_soft_triplet_loss” in 211 1. I wonder why you're reshape the size to 160. 2. In vvv=1, I wonder why you get rid of -1 from...

While digging through the predictors I noticed a small but interesting difference in the way the mask tokens are initialized: ``` # I‑JEPA (vision_transformer_predictor.py) self.mask_token = nn.Parameter(torch.zeros(1, 1, predictor_embed_dim)) trunc_normal_(self.mask_token,...