bytedance / DEADiff

[CVPR 2024] Official implementation of "DEADiff: An Efficient Stylization Diffusion Model with Disentangled Representations"
Apache License 2.0
214 stars 4 forks source link

Concern about Code Release #1

Closed haofanwang closed 6 months ago

haofanwang commented 6 months ago

Thanks for your good work. I have a few concerns.

  1. When will you release inference/training code?
  2. Why not train a SDXL version?
  3. Is there a plan to release your dataset?
Corleone-Huang commented 6 months ago

The same concerns.

Tianhao-Qi commented 6 months ago

Thanks for your good work. I have a few concerns.

  1. When will you release inference/training code?
  2. Why not train a SDXL version?
  3. Is there a plan to release your dataset?

Thanks for your interest in our work. We've included our replies to your concerns:

  1. We'll release our inference code within this month.
  2. DEADiff adapted to SDXL is in progress and will also be released.
  3. We will release the detailed procedure of the dataset construction and are considering the form of releasing our dataset.
loboere commented 6 months ago

Another question, how much vram ram do you need?

Tianhao-Qi commented 6 months ago

Another question, how much vram ram do you need?

You can refer to Table 5 in our paper:

image
akk-123 commented 6 months ago

@Tianhao-Qi April is here, when will you release the code?