obscura/diffusers
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
4,184 Commits 711 Branches 93 Tags
dbc8b516eee3407ef329d9e72fe48b3ebb4e4584
Commit Graph

4184 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sayak Paul dbc8b516ee Update README.md to update the MaPO project (#8470) 
Update README.md
 2024-12-23 13:02:12 +05:30
Jianqi Pan c2f2dc8013 🔧 chore: use modeling_outputs.Transformer2DModelOutput (#8436) 
* 🔧 chore: use modeling_outputs.Transformer2DModelOutput

* 🔧 chore: isort

* 🔧 chore: isort

* style

---------

Co-authored-by: sayakpaul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Luc Georges 1c9ac547a3 fix(ci): remove unnecessary permissions (#8457) 2024-12-23 13:02:12 +05:30
Luc Georges 14d7c986f2 feat(ci): add trufflehog secrets detection (#8430) 2024-12-23 13:02:12 +05:30
Lucain 4e7fa7715d Move away from cached_download (#8419) 
* Move away from

* unused constant

* Add custom error
 2024-12-23 13:02:12 +05:30
Sayak Paul 5b279af650 [Core] support saving and loading of sharded checkpoints (#7830) 
* feat: support saving a model in sharded checkpoints.

* feat: make loading of sharded checkpoints work.

* add tests

* cleanse the loading logic a bit more.

* more resilience while loading from the Hub.

* parallelize shard downloads by using snapshot_download()/

* default to a shard size.

* more fix

* Empty-Commit

* debug

* fix

* uality

* more debugging

* fix more

* initial comments from Benjamin

* move certain methods to loading_utils

* add test to check if the correct number of shards are present.

* add a test to check if loading of sharded checkpoints from the Hub is okay

* clarify the unit when passed as an int.

* use hf_hub for sharding.

* remove unnecessary code

* remove unnecessary function

* lucain's comments.

* fixes

* address high-level comments.

* fix test

* subfolder shenanigans./

* Update src/diffusers/utils/hub_utils.py

Co-authored-by: Lucain <lucainp@gmail.com>

* Apply suggestions from code review

Co-authored-by: Lucain <lucainp@gmail.com>

* remove _huggingface_hub_version as not needed.

* address more feedback.

* add a test for local_files_only=True/

* need hf hub to be at least 0.23.2

* style

* final comment.

* clean up subfolder.

* deal with suffixes in code.

* _add_variant default.

* use weights_name_pattern

* remove add_suffix_keyword

* clean up downloading of sharded ckpts.

* don't return something special when using index.json

* fix more

* don't use bare except

* remove comments and catch the errors better

* fix a couple of things when using is_file()

* empty

---------

Co-authored-by: Lucain <lucainp@gmail.com>
 2024-12-23 13:02:12 +05:30
Lucain 68892fb0a3 Final fix for mirror community pipeline (#8427) 2024-12-23 13:02:12 +05:30
Lucain 7d461e6b90 Fix mirror community pipeline (#8426) 2024-12-23 13:02:12 +05:30
Lucain 6130c5b827 Fix mirror_community_pipeline.yml name (#8425) 2024-12-23 13:02:12 +05:30
Lucain ad4b3eee9b Mirror ./examples/community folder on HF (#8417) 
* first draft

* secret

* tiktok

* capital matters

* dataset matter

* don't be a prick

* refact

* only on main or tag

* document with an example

* Update destination dataset

* link

* allow manual trigger

* better

* lin

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Tolga Cangöz 9ed762d207 Optimize test files by fixing CPU-offloading usage (#8409) 
* Refactor code to remove unnecessary calls to `to(torch_device)`

* Refactor code to remove unnecessary calls to `to("cuda")`

* Update pipeline_stable_diffusion_diffedit.py
 2024-12-23 13:02:12 +05:30
Steven Liu 072b14dea1 [docs] Single file usage (#8412) 
* single file usage

* edit
 2024-12-23 13:02:12 +05:30
Sayak Paul 9bafb24cd6 [Core] fix: legacy model mapping (#8416) 
* fix: legacy model mapping

* remove print
 2024-12-23 13:02:12 +05:30
Sayak Paul 751ed84019 [Hunyuan] add optimization related sections to the hunyuan dit docs. (#8402) 
* optimizations to the hunyuan dit docs.

* Apply suggestions from code review

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

* Update docs/source/en/api/pipelines/hunyuandit.md

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
 2024-12-23 13:02:12 +05:30
Tolga Cangöz d027cb4326 Errata (#8322) 
* Fix typos

* Trim trailing whitespaces

* Remove a trailing whitespace

* chore: Update MarigoldDepthPipeline checkpoint to prs-eth/marigold-lcm-v1-0

* Revert "chore: Update MarigoldDepthPipeline checkpoint to prs-eth/marigold-lcm-v1-0"

This reverts commit fd742b30b4.

* pokemon -> naruto

* `DPMSolverMultistep` -> `DPMSolverMultistepScheduler`

* Improve Markdown stylization

* Improve style

* Improve style

* Refactor pipeline variable names for consistency

* up style
 2024-12-23 13:02:12 +05:30
Guillaume LEGENDRE a4dd48662e Update tailscale action to main (#8403) 2024-12-23 13:02:12 +05:30
Hzzone cce0e5af64 Gligen training (#7906) 
* add training code of gligen

* fix code quality tests.

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Sayak Paul 49b9fbb51c [Scheduler] fix: EDM schedulers when using the exp sigma schedule. (#8385) 
* fix: euledm when using the exp sigma schedule.

* fix-copies

* remove print.

* reduce friction

* yiyi's suggestioms
 2024-12-23 13:02:12 +05:30
Sayak Paul 9873a35ae6 [Hunyuan] allow Hunyuan DiT to run under 6GB for GPU VRAM (#8399) 
* allow hunyuan dit to run under 6GB for GPU VRAM

* add section in the docs/
 2024-12-23 13:02:12 +05:30
Sayak Paul 385abf623f [LoRA] Remove legacy LoRA code and related adjustments (#8316) 
* remove legacy code from load_attn_procs.

* finish first draft

* fix more.

* fix more

* add test

* add serialization support.

* fix-copies

* require peft backend for lora tests

* style

* fix test

* fix loading.

* empty

* address benjamin's feedback.
 2024-12-23 13:02:12 +05:30
Sayak Paul 3e841ebca6 [Hunyuan] feat: support chunked ff. (#8397) 
feat: support chunked ff.
 2024-12-23 13:02:12 +05:30
Sayak Paul 8ade2c94d0 [Hunyuan DiT] feat: enable fusing qkv projections when doing attention (#8396) 
* feat: introduce qkv fusion for Hunyuan

* fix copies
 2024-12-23 13:02:12 +05:30
leaps 942bc38be9 Update code example in pipeline_stable_unclip_img2img.py EXAMPLE_DOC_STRING (#8401) 
Update code example in pipeline_stable_unclip_img2img.py

Previous code caused an error when run
 2024-12-23 13:02:12 +05:30
Sayak Paul 4f77f52853 [Transformer2DModel] Handle norm_type safely while remapping (#8370) 
* handle norm_type of transformer2d_model safely.

* log an info when old model class is being returned.

* Apply suggestions from code review

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

* remove extra stuff

---------

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
 2024-12-23 13:02:12 +05:30
Sayak Paul 9b1c118692 [HunyuanDiT] minor docs changes in hunyuandit (#8395) 
minor docs changes in hunyuandit
 2024-12-23 13:02:12 +05:30
townwish4git c0a81199d3 Fix AsymmetricAutoencoderKL forward (#8378) 2024-12-23 13:02:12 +05:30
Marçal Comajoan Cara 26e7b4bd9a Update transformer2d.md title (#8375) 
* Update transformer2d.md title

For the other classes (e.g., UNet2DModel) the title of the documentation coincides with the name of the class, but that was not the case for Transformer2DModel.

* Update model docs titles for consistency with class names
 2024-12-23 13:02:12 +05:30
Dhruv Nair 92b915aca7 Update slow test actions (#8381) 
* update

* update

* update

* update
 2024-12-23 13:02:12 +05:30
XCL d6c7e17867 Tencent Hunyuan Team - Updated Doc for HunyuanDiT (#8383) 
* add hunyuandit doc

* update hunyuandit doc

* update hunyuandit 2d model

* update toctree.yml for hunyuandit
 2024-12-23 13:02:12 +05:30
XCL 43510582cd Tencent Hunyuan Team: add HunyuanDiT related updates (#8240) 
* Hunyuan Team: add HunyuanDiT related updates


---------

Co-authored-by: XCLiu <liuxc1996@gmail.com>
Co-authored-by: yiyixuxu <yixu310@gmail.com>
 2024-12-23 13:02:12 +05:30
39th president of the United States, probably 7f58fad44d Fix DREAM training (#8302) 
Co-authored-by: Jimmy <39@🇺🇸.com>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Co-authored-by: YiYi Xu <yixu310@gmail.com>
 2024-12-23 13:02:12 +05:30
Anton Obukhov a495ed3e8b Fix marigold documentation (#8372) 
* rename prs-eth/marigold-lcm-v1-0 into prs-eth/marigold-depth-lcm-v1-0

* update image paths in https://huggingface.co/datasets/huggingface/documentation-images to use main branch

* fix relative paths to other diffusers pages

* Update docs/source/en/using-diffusers/marigold_usage.md

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
 2024-12-23 13:02:12 +05:30
Sayak Paul b9e55fcbf7 [Core] Introduce class variants for Transformer2DModel (#7647) 
* init for patches

* finish patched model.

* continuous transformer

* vectorized transformer2d.

* style.

* inits.

* fix-copies.

* introduce DiTTransformer2DModel.

* fixes

* use REMAPPING as suggested by @DN6

* better logging.

* add pixart transformer model.

* inits.

* caption_channels.

* attention masking.

* fix use_additional_conditions.

* remove print.

* debug

* flatten

* fix: assertion for sigma

* handle remapping for modeling_utils

* add tests for dit transformer2d

* quality

* placeholder for pixart tests

* pixart tests

* add _no_split_modules

* add docs.

* check

* check

* check

* check

* fix tests

* fix tests

* move Transformer output to modeling_output

* move errors better and bring back use_additional_conditions attribute.

* add unnecessary things from DiT.

* clean up pixart

* fix remapping

* fix device_map things in pixart2d.

* replace Transformer2DModel with appropriate classes in dit, pixart tests

* empty

* legacy mixin classes./

* use a remapping dict for fetching class names.

* change to specifc model types in the pipeline implementations.

* move _fetch_remapped_cls_from_config to modeling_loading_utils.py

* fix dependency problems.

* add deprecation note.
 2024-12-23 13:02:12 +05:30
Dhruv Nair f6553106f7 Change checkpoint key used to identify CLIP models in single file checkpoints (#8319) 
update
 2024-12-23 13:02:12 +05:30
Jonah 6f0153747c Fix depth pipeline "input/weight type should be the same" error at fp16 (#8321) 
Fix "input/weight type should be the same"

Co-authored-by: YiYi Xu <yixu310@gmail.com>
 2024-12-23 13:02:12 +05:30
satani99 760e5993de Modularize train_text_to_image_lora_sdxl inferencing during and after training in example (#8335) 
* Modularized the train_lora_sdxl file

* Modularized the train_lora_sdxl file

* Modularized the train_lora_sdxl file

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Genius Patrick a1cfb0accc fix(training): lr scheduler doesn't work properly in distributed scenarios (#8312) 2024-12-23 13:02:12 +05:30
Dhruv Nair 4be03cfa82 Fix StableDiffusionPipeline when text_encoder=None (#8297) 
* update

* update

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Tolga Cangöz b2abd9e371 Fix Copying Mechanism typo/bug (#8232) 
* Fix copying mechanism typos

* fix copying mecha

* Revert, since they are in TODO

* Fix copying mechanism
 2024-12-23 13:02:12 +05:30
Steven Liu c62a927ba1 [docs] Files and formats (#7874) 
* files and formats

* fix callout

* feedback

* code sample

* feedback
 2024-12-23 13:02:12 +05:30
Steven Liu 8ccc3f225c [docs] DeepFloyd training (#8224) 
deepfloyd training

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Tolga Cangöz 64e3d998b3 Simplify platform_info assignment in diffusers-cli env (#8298) 
chore: Simplify `platform_info` assignment
 2024-12-23 13:02:12 +05:30
satani99 630a71361f Modularize train_text_to_image_lora SD inferencing during and after training in example (#8283) 
* Modularized the train_lora file

* Modularized the train_lora file

* Modularized the train_lora file

* Modularized the train_lora file

* Modularized the train_lora file

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2024-12-23 13:02:12 +05:30
Sayak Paul e135b4dadf post release v0.28.0 (#8286) 
* post release v0.28.0

* style
 2024-12-23 13:02:12 +05:30
Sayak Paul 93c3abef71 [Core] Refactor IPAdapterPlusImageProjection a bit (#7994) 
* use IPAdapterPlusImageProjectionBlock in IPAdapterPlusImageProjection

* reposition IPAdapterPlusImageProjection

* refactor complete?

* fix heads param retrieval.

* update test dict creation method.
 2024-12-23 13:02:12 +05:30
Sayak Paul 3a932ec0bd move vqmodel to models.autoencoders. (#8292) 
move vqmodel to models.autoencoders.
 2024-12-23 13:02:12 +05:30
Sayak Paul 84f7039fb3 [Post release 0.28.0] remove deprecated blocks. (#8291) 
* remove deprecated blocks.

* update the location paths.
 2024-12-23 13:02:12 +05:30
Vladimir Mandic 57843c9fac fix pixart-sigma negative prompt handling (#8299) 
* fix negative prompt

* fix

---------

Co-authored-by: yiyixuxu <yixu310@gmail,com>
Co-authored-by: YiYi Xu <yixu310@gmail.com>
 2024-12-23 13:02:12 +05:30
Steven Liu e73f80f05a [docs] Outpaint (#7964) 
* first draft

* edits
 2024-12-23 13:02:12 +05:30
Steven Liu 6e0c2947e7 [docs] Scheduler features (#7990) 
* noise schedule

* sigmas and zero snr

* feedback

* feedback
 2024-12-23 13:02:12 +05:30