nam-drun

Zero-Shot Image Restoration model for final denoising steps in Image Generation task?

3

### Let's say an Image Generation model is meant to generate image from random noise in 50 steps: 1. What if I use Image Generation model for the initial 20...

How to start learning to manipulate tensor at low-level like flash-attention?

1

I'm keen to manipulate tensor at C++ and CUDA low level, and I can ask chatGPT to translate/explain C++ line-by-line. I just don't know where to start my learning journey

Do you think QLora finetuning will work fine with Mamba?

1

We're planning to use QLora finetuning Mamba in couple weeks soon

[Patch Diffusion](https://github.com/Zhendong-Wang/Patch-Diffusion/tree/main) can x2 training speed even on 256x256 ImageNet. If this works out between Mosaic Diffusion and Patch-Diffusion, that is potentially x10 cumulative boost. The issue is both have...

Difference between HRM and H-Net?

1

Dear authors, I'm curious to understand the crux of HRM. I feel like it's trying to solve the same problem as H-Net but with a different architectural and loss function...