VanillaNet
VanillaNet copied to clipboard
huawei-noah
Hi! Your work is fantastic! I have two questions regarding the hyperparameter λ: According to Formula 1, λ should start at 0 and end at 1 during the training process....
Dear team, I would like to express my heartfelt gratitude for this truly remarkable project! The effort and dedication you have put into it are truly commendable. I saw that...
I would like to ask how n^2 in Eq.8 was obtained, or could you please explain in detail the calculation process of Eq.8. Thank you very much
Question 1: Will the batch size during training affect the final performance of the model? Have you tried small batch size, such as 16 and 32? Question 2: Why do...
I came across your GitHub repository where you have shared a PyTorch implementation of a backbone network for a classification task. I'm interested in using this backbone for a segmentation...
An exciting work! It seems there is no "path" folder to support futher test or training. Will you upload is later?
Your work is exciting! but I have a question about the width of the model，Have you tried to reduce the width of the model, why is the width of the...
Hi, could you please provide any results or findings regarding the utilization of a batch size greater than 1?
Hi, Thanks for your amazing work! I found the [padding size](https://github.com/huawei-noah/VanillaNet/blob/f8385cf4ac0a5143431a05c84c5a40da4491b3d8/models/vanillanet.py#L29) used in activation funtion is actually same as `padding=self.act_num`, so can i ask why you use `padding=(self.act_num*2+1)//2` instead of...
