Skip to content

Issues: NVlabs/Sana

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

35 Open 76 Closed
35 Open 76 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

Even 80 GB is not sufficient for SANA 4K model VAE decoding - something feels wrong?
#139 opened Jan 9, 2025 by FurkanGozukara
10
No module named 'patch_conv'
#138 opened Jan 9, 2025 by nitinmukesh
1
The model cannot handle Chinese prompt correctly
#137 opened Jan 9, 2025 by zkyf
1
question about load sanapipeline
#134 opened Jan 8, 2025 by LearningHx
Error details: list index out of range
#133 opened Jan 7, 2025 by Shuaizhang7
[BUG] FSDP error: Could not find the transformer layer class SanaBlock in the model Answered Answered the question
#129 opened Jan 5, 2025 by Pevernow
7
Can inference be done at FP8? for both 1K and 2K models Answered Answered the question
#128 opened Jan 5, 2025 by FurkanGozukara
4
Full Tutorial, 1-click installers (Windows, RunPod, Massed Compute & free Kaggle account notebook) and extensive test examples of 2K model documentation Improvements or additions to documentation
#127 opened Jan 4, 2025 by FurkanGozukara
3
[Bug]Received signal 7 (SIGBUS) during training with multiple GPUs Answered Answered the question fixed fix a bug
#126 opened Jan 3, 2025 by Pevernow
4
Error(s) in loading state_dict for SanaMS: size mismatch for pos_embed: copying a param with shape torch.Size([1, 256, 1152]) from checkpoint, the shape in current model is torch.Size([1, 1024, 1152]). Answered Answered the question bug Something isn't working fixed fix a bug
#123 opened Dec 31, 2024 by VeteranXT
5
Question: How to construct wids-meta.json for SanaWebDataset? working working on this issue
#121 opened Dec 29, 2024 by Pevernow
6
clip documentation Improvements or additions to documentation
#116 opened Dec 25, 2024 by Muinez
4
Do not use Comfyui-Gemma nodes, use new project ComfyUI_ExtraModels Answered Answered the question documentation Improvements or additions to documentation
#113 opened Dec 24, 2024 by shiter007
3
Using tar shards to train sana working working on this issue
#111 opened Dec 23, 2024 by nicolas-dufour
3
Question: Flow-DPM-Solver Answered Answered the question
#110 opened Dec 22, 2024 by dxqbYD
6
[Request] Release BF16 version of 0.6B 1024x model working working on this issue
#109 opened Dec 22, 2024 by Bocchi-Chan2023
4
comfyui ExtraVAELoader error: No module named 'diffusers' bug Something isn't working
#101 opened Dec 20, 2024 by Ni-Deshmane
2
Fine tuning guidance Answered Answered the question
#99 opened Dec 18, 2024 by Otsutsukii
1
DCAE-Sana obtains the NAN value after decoding the latent which comes from encoding image bug Something isn't working
#96 opened Dec 18, 2024 by ZouYa99
6
Problem launching training with less GPUs Answered Answered the question
#92 opened Dec 15, 2024 by AfterHAL
10
Why is the flow-DPM Solver so much worse than Euler's method in my flow matching model. FID become very large when use it Answered Answered the question
#91 opened Dec 14, 2024 by lllzzz3
3
Sana finetuned example documentation Improvements or additions to documentation
#74 opened Dec 6, 2024 by recoilme
3
Controlnet Support Answered Answered the question enhancement New feature or request
#73 opened Dec 6, 2024 by Oguzhanercan
4
Wrong epoch count on training (Bug?)
#71 opened Dec 5, 2024 by AfterHAL
1
Sana checkpoint trained with SD-3 VAE
#70 opened Dec 4, 2024 by srikarym
3
ProTip! Mix and match filters to narrow down what you’re looking for.