Skip to content

Issues: tianweiy/DMD2

How about DMD2 on img2img task?
#17 opened May 31, 2024 by dawei03896
Open 2
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

21 Open 26 Closed
21 Open 26 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

The DMD loss increases during the first stage of training.
#49 opened Sep 13, 2024 by CS123n
2
About Loss
#48 opened Sep 5, 2024 by joelulu
Why use the intermediate estimated x_0|t as the distribution matching input， but not the final image of generator?
#45 opened Aug 30, 2024 by vv12kant
2
Support SD3？
#44 opened Aug 26, 2024 by mliand
1
DMD2 for Flux Dev
#43 opened Aug 24, 2024 by pandayummy
1
train a 4step SDXL got AssertionError: Invalid: `_post_backward_hook_state
#41 opened Aug 9, 2024 by joelulu
8
FSDP is too slow, why not use deepspeed?
#40 opened Aug 8, 2024 by sunhaha123
1
is this exepcted when you are training sdv1.5 in stage2: gen_cls_loss (g loss) keeps increasing while guidance_cls_loss (d loss) deacreases.
#39 opened Aug 6, 2024 by fire2323
9
DMD2 and Kandinsky, or Playground 2.5?
#37 opened Jul 21, 2024 by Quick-Eyed-Sky
muti-step model for sdv1.5
#36 opened Jul 20, 2024 by kmpartner
5
LoRA training
#35 opened Jul 19, 2024 by SYZhang0805
4
how long is the extra regression loss pre-training in 1-step SDXL distillation trained?
#34 opened Jul 17, 2024 by grovessss
1
When i applyed the dmd loss to velocity prediction , the loss varies a lot。
#33 opened Jul 16, 2024 by icelighting
1
After training, there are two pytorch_model files. Which one should I use?
#31 opened Jul 15, 2024 by koking0
5
is there plan for releasing code for grad accumulation?
#23 opened Jun 26, 2024 by grovessss
4
Convert 4-step unet into a checkpoint file
#20 opened Jun 11, 2024 by coastbluet
6
sdv15 train with torchrun successful, but sdxl train with fsdp failed.
#18 opened May 31, 2024 by conquer-pan
How about DMD2 on img2img task? img2img
#17 opened May 31, 2024 by dawei03896
2
Current CUDA Device does not support bfloat16. Please switch dtype to float16.
#14 opened May 30, 2024 by conquer-pan
Is it possible to integrate with HiDiffusion?
#10 opened May 27, 2024 by self-destruction
6
Possibility for 30fps image generation?
#6 opened May 26, 2024 by beyarkay
6
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.