Skip to content

Add cross attention type for Sana-Sprint training in diffusers. #11514

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

Merged
merged 12 commits into from
May 8, 2025
Merged

Add cross attention type for Sana-Sprint training in diffusers. #11514

Changes from 1 commit
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
5d9a5da
test permission
scxue May 7, 2025
1123ee7
Add cross attention type for Sana-Sprint.
scxue May 7, 2025
acefec8
Add Sana-Sprint training script in diffusers.
scxue May 7, 2025
9cb050b
make style && make quality;
lawrence-cj May 7, 2025
86bef58
modify the attention processor with `set_attn_processor` and change `…
lawrence-cj May 8, 2025
c190600
Merge branch 'main' into main
lawrence-cj May 8, 2025
6c3a398
Add import for SanaVanillaAttnProcessor
scxue May 8, 2025
04e1b02
Add README file.
scxue May 8, 2025
5951f8f
Apply suggestions from code review
sayakpaul May 8, 2025
93c3b4d
Merge branch 'main' into main
sayakpaul May 8, 2025
566aa64
style
sayakpaul May 8, 2025
740baa9
Update examples/research_projects/sana/README.md
lawrence-cj May 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Update examples/research_projects/sana/README.md
  • Loading branch information
@lawrence-cj
lawrence-cj authored May 8, 2025
commit 740baa9e1fbed0535d2723c1e70fe44b62fad680
2 changes: 1 addition & 1 deletion examples/research_projects/sana/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,7 +26,7 @@ huggingface-cli download Efficient-Large-Model/SANA_Sprint_1.6B_1024px_teacher_d

### 3. Acquire the dataset shards

The training script in this example uses specific `.parquet` shards from the `brivangl/midjourney-v6-llava` dataset instead of downloading the entire dataset automatically via `dataset_name`.
The training script in this example uses specific `.parquet` shards from a randomly selected `brivangl/midjourney-v6-llava` dataset instead of downloading the entire dataset automatically via `dataset_name`.

The script specifically uses these three files:
* `data/train_000.parquet`
Expand Down
Loading