Explore >> Select a destination
|
You are here 
|
distill.pub | ||
| | | | |
blog.evjang.com
|
|
| | | | | This is a tutorial on common practices in training generative models that optimize likelihood directly, such as autoregressive models and ... | |
| | | | |
colinraffel.com
|
|
| | | | | ||
| | | | |
sander.ai
|
|
| | | | | Thoughts on the tension between iterative refinement as the thing that makes diffusion models work, and our continual attempts to make it _less_ iterative. | |
| | | | |
www.paepper.com
|
|
| | | Today's paper: Rethinking 'Batch' in BatchNorm by Wu & Johnson BatchNorm is a critical building block in modern convolutional neural networks. Its unique property of operating on "batches" instead of individual samples introduces significantly different behaviors from most other operations in deep learning. As a result, it leads to many hidden caveats that can negatively impact model's performance in subtle ways. This is a citation from the paper's abstract and the emphasis is mine which caught my attention. Let's explore these subtle ways which can negatively impact your model's performance! The paper of Wu & Johnson can be found on arxiv. | ||
