DavidBert
574f8fd10a
parameter names match the standard diffusers conventions
2025-10-21 07:27:15 +00:00
DavidBert
6e05172682
Revert accidental .gitignore change
2025-10-21 07:27:15 +00:00
DavidBert
d0c029f15d
built-in RMSNorm
2025-10-21 07:27:15 +00:00
DavidBert
015774399e
Refactor PhotonAttention to match Flux pattern
2025-10-21 07:27:14 +00:00
DavidBert
5f99168def
utility function that determines the default resolution given the VAE
2025-10-21 07:27:14 +00:00
DavidBert
c329c8f667
add pipeline test + corresponding fixes
2025-10-21 07:27:14 +00:00
David Bertoin
bb36735379
make quality + style
2025-10-21 07:27:14 +00:00
David Bertoin
8ee17d20b3
Use _import_structure for lazy loading
2025-10-21 07:27:14 +00:00
David Bertoin
be1d14658e
add negative prompts
2025-10-21 07:27:14 +00:00
David Bertoin
c951adef45
move xattention conditionning out computation out of the denoising loop
2025-10-21 07:27:14 +00:00
David Bertoin
a74e0b726a
support prompt_embeds in call
2025-10-21 07:27:14 +00:00
David Bertoin
ffe3501c1c
rename vae_spatial_compression_ratio for vae_scale_factor
2025-10-21 07:27:14 +00:00
David Bertoin
2077252947
remove lora related code
2025-10-21 07:27:14 +00:00
David Bertoin
de1ceaf07a
renam LastLayer for FinalLayer
2025-10-21 07:27:14 +00:00
David Bertoin
3c60c9230e
put _attn_forward and _ffn_forward logic in PhotonBlock's forward
2025-10-21 07:27:14 +00:00
David Bertoin
af8882d7e6
remove modulation dataclass
2025-10-21 07:27:14 +00:00
David Bertoin
5f0bf0181f
Rename EmbedND for PhotoEmbedND
2025-10-21 07:27:14 +00:00
davidb
ae44d845b6
remove lora support from doc
2025-10-21 07:27:14 +00:00
davidb
12dbabe607
fix timestep shift
2025-10-21 07:27:14 +00:00
davidb
ec70e3fdc0
fix T5Gemma loading from hub
2025-10-21 07:27:14 +00:00
davidb
6634113ef6
update doc
2025-10-21 07:27:14 +00:00
davidb
b07d1c8799
update doc
2025-10-21 07:27:14 +00:00
davidb
a9e301366a
unify the structure of the forward block
2025-10-21 07:27:14 +00:00
davidb
5886925346
remove einops dependency and now inherits from AttentionMixin
2025-10-21 07:27:14 +00:00
davidb
25a0061d65
move PhotonAttnProcessor2_0 in transformer_photon
2025-10-21 07:27:14 +00:00
davidb
6284b9d062
remove enhance vae and use vae.config directly when possible
2025-10-21 07:27:14 +00:00
davidb
60d918d79b
conditioned CFG
2025-10-21 07:27:14 +00:00
David Briand
b327b36ad9
BF16 example
2025-10-21 07:27:14 +00:00
davidb
14903ee599
remove autocast for text encoder forwad
2025-10-21 07:27:14 +00:00
davidb
d71ddd0079
enhance_vae_properties if vae is provided only
2025-10-21 07:27:14 +00:00
davidb
6a66fbd2c4
just store the T5Gemma encoder
2025-10-21 07:27:13 +00:00
davidb
e487660e05
Add Photon model and pipeline support
...
This commit adds support for the Photon image generation model:
- PhotonTransformer2DModel: Core transformer architecture
- PhotonPipeline: Text-to-image generation pipeline
- Attention processor updates for Photon-specific attention mechanism
- Conversion script for loading Photon checkpoints
- Documentation and tests
2025-10-21 07:27:13 +00:00
Steven Liu
5b5fa49a89
[docs] Organize toctree by modality ( #12514 )
...
* reorganize
* fix
---------
Co-authored-by: Álvaro Somoza <asomoza@users.noreply.github.com >
2025-10-21 10:18:54 +05:30
Fei Xie
decfa3c9e1
Fix: Use incorrect temporary variable key when replacing adapter name… ( #12502 )
...
Fix: Use incorrect temporary variable key when replacing adapter name in state dict within load_lora_adapter function
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com >
2025-10-20 15:45:37 -10:00
Dhruv Nair
48305755bf
Raise warning instead of error when imports are missing for custom code ( #12513 )
...
update
2025-10-20 07:02:23 -10:00
dg845
7853bfbed7
Remove Qwen Image Redundant RoPE Cache ( #12452 )
...
Refactor QwenEmbedRope to only use the LRU cache for RoPE caching
2025-10-19 18:41:58 -07:00
Lev Novitskiy
23ebbb4bc8
Kandinsky 5 is finally in Diffusers! ( #12478 )
...
* add kandinsky5 transformer pipeline first version
---------
Co-authored-by: Álvaro Somoza <asomoza@users.noreply.github.com >
Co-authored-by: YiYi Xu <yixu310@gmail.com >
Co-authored-by: Charles <charles@huggingface.co >
2025-10-17 18:34:30 -10:00
Ali Imran
1b456bd5d5
docs: cleanup of runway model ( #12503 )
...
* cleanup of runway model
* quality fixes
2025-10-17 14:10:50 -07:00
Sayak Paul
af769881d3
[tests] introduce VAETesterMixin to consolidate tests for slicing and tiling ( #12374 )
...
* up
* up
* up
* up
* up
* u[
* up
* up
* up
2025-10-17 12:02:29 +05:30
Sayak Paul
4715c5c769
[ci] xfail more incorrect transformer imports. ( #12455 )
...
* xfail more incorrect transformer imports.
* xfail more.
* up
* up
* up
2025-10-17 10:35:19 +05:30
Steven Liu
dbe413668d
[CI] Check links ( #12491 )
...
* check links
* update
* feedback
* remove
2025-10-16 10:38:16 -07:00
Steven Liu
26475082cb
[docs] Attention checks ( #12486 )
...
* checks
* feedback
---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com >
2025-10-16 09:19:30 -07:00
YiYi Xu
f072c64bf2
ltx0.9.8 (without IC lora, autoregressive sampling) ( #12493 )
...
update
Co-authored-by: Aryan <aryan@huggingface.co >
2025-10-15 07:41:17 -10:00
Sayak Paul
aed636f5f0
[tests] fix clapconfig for text backbone in audioldm2 ( #12490 )
...
fix clapconfig for text backbone in audioldm2
2025-10-15 10:57:09 +05:30
Sayak Paul
53a10518b9
remove unneeded checkpoint imports. ( #12488 )
2025-10-15 09:51:18 +05:30
Steven Liu
b4e6dc3037
[docs] Fix broken links ( #12487 )
...
fix broken links
2025-10-15 06:42:10 +05:30
Steven Liu
3eb40786ca
[docs] Prompting ( #12312 )
...
* init
* fix
* batch inf
* feedback
* update
2025-10-14 13:53:56 -07:00
Meatfucker
a4bc845478
Fix missing load_video documentation and load_video import in WanVideoToVideoPipeline example code ( #12472 )
...
* Update utilities.md
Update missing load_video documentation
* Update pipeline_wan_video2video.py
Fix missing load_video import in example code
2025-10-14 10:43:21 -07:00
Manith Ratnayake
fa468c5d57
docs: api-pipelines-qwenimage typo fix ( #12461 )
2025-10-13 08:57:46 -07:00
Steven Liu
8abc7aeb71
[docs] Fix syntax ( #12464 )
...
* fix syntax
* fix
* style
* fix
2025-10-11 08:13:30 +05:30
