Fix deterministic issue when getting pipeline dtype and device #10696
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dimitribarbot
commented
Jan 31, 2025
| @@ -1619,10 +1619,12 @@ def components(self) -> Dict[str, Any]: | |||
| k: getattr(self, k) for k in self.config.keys() if not k.startswith("_") and k not in optional_parameters | |||
| } | |||
|
|
|||
| if set(components.keys()) != expected_modules: | |||
| actual = sorted(set(components.keys())) | |||
| expected = sorted(expected_modules) | |||
There was a problem hiding this comment.
I'm not sure if using sorted is necessary here, since expected_modules comes from the _get_signature_keys call which already calls sorted internally.
However, I leave it that way because it allows to be completely decoupled from the _get_signature_keys function. And I don't think it will add a big performance overhead (sorting of n elements, n=~10, elements already sorted).
dimitribarbot
commented
Jan 31, 2025
| @@ -585,3 +585,104 @@ def test_video_to_video(self): | |||
| with io.StringIO() as stderr, contextlib.redirect_stderr(stderr): | |||
| _ = pipe(**inputs) | |||
| self.assertTrue(stderr.getvalue() == "", "Progress bar should be disabled") | |||
|
|
|||
|
|
|||
| @require_torch_gpu | |||
There was a problem hiding this comment.
I need this to test multiple devices across pipeline components, so that testing with distinct pipeline devices fails without the code changes. Otherwise, I guess only CPU can be available?
|
Hi @sayakpaul, any news regarding this PR? |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
…ting-pipeline-dtype-and-device
…e-dtype-and-device
DN6
approved these changes
Mar 7, 2025
dimitribarbot
deleted the
fix-deterministic-issue-when-getting-pipeline-dtype-and-device
branch
sayakpaul
added a commit
that referenced
this pull request
Mar 20, 2025
Co-authored-by: SunMarc <[email protected]> condition better. support mapping. improvements. [Quantization] Add Quanto backend (#10756) * update * updaet * update * update * update * update * update * update * update * update * update * update * Update docs/source/en/quantization/quanto.md Co-authored-by: Sayak Paul <[email protected]> * update * update * update * update * update * update * update * update * update * update * update * update * update * update * update * update * update * update * Update src/diffusers/quantizers/quanto/utils.py Co-authored-by: Sayak Paul <[email protected]> * update * update --------- Co-authored-by: Sayak Paul <[email protected]> [Single File] Add single file loading for SANA Transformer (#10947) * added support for from_single_file * added diffusers mapping script * added testcase * bug fix * updated tests * corrected code quality * corrected code quality --------- Co-authored-by: Dhruv Nair <[email protected]> [LoRA] Improve warning messages when LoRA loading becomes a no-op (#10187) * updates * updates * updates * updates * notebooks revert * fix-copies. * seeing * fix * revert * fixes * fixes * fixes * remove print * fix * conflicts ii. * updates * fixes * better filtering of prefix. --------- Co-authored-by: hlky <[email protected]> [LoRA] CogView4 (#10981) * update * make fix-copies * update [Tests] improve quantization tests by additionally measuring the inference memory savings (#11021) * memory usage tests * fixes * gguf [`Research Project`] Add AnyText: Multilingual Visual Text Generation And Editing (#8998) * Add initial template * Second template * feat: Add TextEmbeddingModule to AnyTextPipeline * feat: Add AuxiliaryLatentModule template to AnyTextPipeline * Add bert tokenizer from the anytext repo for now * feat: Update AnyTextPipeline's modify_prompt method This commit adds improvements to the modify_prompt method in the AnyTextPipeline class. The method now handles special characters and replaces selected string prompts with a placeholder. Additionally, it includes a check for Chinese text and translation using the trans_pipe. * Fill in the `forward` pass of `AuxiliaryLatentModule` * `make style && make quality` * `chore: Update bert_tokenizer.py with a TODO comment suggesting the use of the transformers library` * Update error handling to raise and logging * Add `create_glyph_lines` function into `TextEmbeddingModule` * make style * Up * Up * Up * Up * Remove several comments * refactor: Remove ControlNetConditioningEmbedding and update code accordingly * Up * Up * up * refactor: Update AnyTextPipeline to include new optional parameters * up * feat: Add OCR model and its components * chore: Update `TextEmbeddingModule` to include OCR model components and dependencies * chore: Update `AuxiliaryLatentModule` to include VAE model and its dependencies for masked image in the editing task * `make style` * refactor: Update `AnyTextPipeline`'s docstring * Update `AuxiliaryLatentModule` to include info dictionary so that text processing is done once * simplify * `make style` * Converting `TextEmbeddingModule` to ordinary `encode_prompt()` function * Simplify for now * `make style` * Up * feat: Add scripts to convert AnyText controlnet to diffusers * `make style` * Fix: Move glyph rendering to `TextEmbeddingModule` from `AuxiliaryLatentModule` * make style * Up * Simplify * Up * feat: Add safetensors module for loading model file * Fix device issues * Up * Up * refactor: Simplify * refactor: Simplify code for loading models and handling data types * `make style` * refactor: Update to() method in FrozenCLIPEmbedderT3 and TextEmbeddingModule * refactor: Update dtype in embedding_manager.py to match proj.weight * Up * Add attribution and adaptation information to pipeline_anytext.py * Update usage example * Will refactor `controlnet_cond_embedding` initialization * Add `AnyTextControlNetConditioningEmbedding` template * Refactor organization * style * style * Move custom blocks from `AuxiliaryLatentModule` to `AnyTextControlNetConditioningEmbedding` * Follow one-file policy * style * [Docs] Update README and pipeline_anytext.py to use AnyTextControlNetModel * [Docs] Update import statement for AnyTextControlNetModel in pipeline_anytext.py * [Fix] Update import path for ControlNetModel, ControlNetOutput in anytext_controlnet.py * Refactor AnyTextControlNet to use configurable conditioning embedding channels * Complete control net conditioning embedding in AnyTextControlNetModel * up * [FIX] Ensure embeddings use correct device in AnyTextControlNetModel * up * up * style * [UPDATE] Revise README and example code for AnyTextPipeline integration with DiffusionPipeline * [UPDATE] Update example code in anytext.py to use correct font file and improve clarity * down * [UPDATE] Refactor BasicTokenizer usage to a new Checker class for text processing * update pillow * [UPDATE] Remove commented-out code and unnecessary docstring in anytext.py and anytext_controlnet.py for improved clarity * [REMOVE] Delete frozen_clip_embedder_t3.py as it is in the anytext.py file * [UPDATE] Replace edict with dict for configuration in anytext.py and RecModel.py for consistency * 🆙 * style * [UPDATE] Revise README.md for clarity, remove unused imports in anytext.py, and add author credits in anytext_controlnet.py * style * Update examples/research_projects/anytext/README.md Co-authored-by: Aryan <[email protected]> * Remove commented-out image preparation code in AnyTextPipeline * Remove unnecessary blank line in README.md [Quantization] Allow loading TorchAO serialized Tensor objects with torch>=2.6 (#11018) * update * update * update * update * update * update * update * update * update fix: mixture tiling sdxl pipeline - adjust gerating time_ids & embeddings (#11012) small fix on generating time_ids & embeddings [LoRA] support wan i2v loras from the world. (#11025) * support wan i2v loras from the world. * remove copied from. * upates * add lora. Fix SD3 IPAdapter feature extractor (#11027) chore: fix help messages in advanced diffusion examples (#10923) Fix missing **kwargs in lora_pipeline.py (#11011) * Update lora_pipeline.py * Apply style fixes * fix-copies --------- Co-authored-by: hlky <[email protected]> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Fix for multi-GPU WAN inference (#10997) Ensure that hidden_state and shift/scale are on the same device when running with multiple GPUs Co-authored-by: Jimmy <39@🇺🇸.com> [Refactor] Clean up import utils boilerplate (#11026) * update * update * update Use `output_size` in `repeat_interleave` (#11030) [hybrid inference 🍯🐝] Add VAE encode (#11017) * [hybrid inference 🍯🐝] Add VAE encode * _toctree: add vae encode * Add endpoints, tests * vae_encode docs * vae encode benchmarks * api reference * changelog * Update docs/source/en/hybrid_inference/overview.md Co-authored-by: Sayak Paul <[email protected]> * update --------- Co-authored-by: Sayak Paul <[email protected]> Wan Pipeline scaling fix, type hint warning, multi generator fix (#11007) * Wan Pipeline scaling fix, type hint warning, multi generator fix * Apply suggestions from code review [LoRA] change to warning from info when notifying the users about a LoRA no-op (#11044) * move to warning. * test related changes. Rename Lumina(2)Text2ImgPipeline -> Lumina(2)Pipeline (#10827) * Rename Lumina(2)Text2ImgPipeline -> Lumina(2)Pipeline --------- Co-authored-by: YiYi Xu <[email protected]> making ```formatted_images``` initialization compact (#10801) compact writing Co-authored-by: Sayak Paul <[email protected]> Co-authored-by: YiYi Xu <[email protected]> Fix aclnnRepeatInterleaveIntWithDim error on NPU for get_1d_rotary_pos_embed (#10820) * get_1d_rotary_pos_embed support npu * Update src/diffusers/models/embeddings.py --------- Co-authored-by: Kai zheng <[email protected]> Co-authored-by: hlky <[email protected]> Co-authored-by: YiYi Xu <[email protected]> [Tests] restrict memory tests for quanto for certain schemes. (#11052) * restrict memory tests for quanto for certain schemes. * Apply suggestions from code review Co-authored-by: Dhruv Nair <[email protected]> * fixes * style --------- Co-authored-by: Dhruv Nair <[email protected]> [LoRA] feat: support non-diffusers wan t2v loras. (#11059) feat: support non-diffusers wan t2v loras. [examples/controlnet/train_controlnet_sd3.py] Fixes #11050 - Cast prompt_embeds and pooled_prompt_embeds to weight_dtype to prevent dtype mismatch (#11051) Fix: dtype mismatch of prompt embeddings in sd3 controlnet training Co-authored-by: Andreas Jörg <[email protected]> Co-authored-by: Sayak Paul <[email protected]> reverts accidental change that removes attn_mask in attn. Improves fl… (#11065) reverts accidental change that removes attn_mask in attn. Improves flux ptxla by using flash block sizes. Moves encoding outside the for loop. Co-authored-by: Juan Acevedo <[email protected]> Fix deterministic issue when getting pipeline dtype and device (#10696) Co-authored-by: Dhruv Nair <[email protected]> [Tests] add requires peft decorator. (#11037) * add requires peft decorator. * install peft conditionally. * conditional deps. Co-authored-by: DN6 <[email protected]> --------- Co-authored-by: DN6 <[email protected]> CogView4 Control Block (#10809) * cogview4 control training --------- Co-authored-by: OleehyO <[email protected]> Co-authored-by: yiyixuxu <[email protected]> [CI] pin transformers version for benchmarking. (#11067) pin transformers version for benchmarking. updates Fix Wan I2V Quality (#11087) * fix_wan_i2v_quality * Update src/diffusers/pipelines/wan/pipeline_wan_i2v.py Co-authored-by: YiYi Xu <[email protected]> * Update src/diffusers/pipelines/wan/pipeline_wan_i2v.py Co-authored-by: YiYi Xu <[email protected]> * Update src/diffusers/pipelines/wan/pipeline_wan_i2v.py Co-authored-by: YiYi Xu <[email protected]> * Update pipeline_wan_i2v.py --------- Co-authored-by: YiYi Xu <[email protected]> Co-authored-by: hlky <[email protected]> LTX 0.9.5 (#10968) * update --------- Co-authored-by: YiYi Xu <[email protected]> Co-authored-by: hlky <[email protected]> make PR GPU tests conditioned on styling. (#11099) Group offloading improvements (#11094) update Fix pipeline_flux_controlnet.py (#11095) * Fix pipeline_flux_controlnet.py * Fix style update readme instructions. (#11096) Co-authored-by: Juan Acevedo <[email protected]> Resolve stride mismatch in UNet's ResNet to support Torch DDP (#11098) Modify UNet's ResNet implementation to resolve stride mismatch in Torch's DDP Fix Group offloading behaviour when using streams (#11097) * update * update Quality options in `export_to_video` (#11090) * Quality options in `export_to_video` * make style improve more. add placeholders for docstrings. formatting. smol fix. solidify validation and annotation
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR aims to fix a deterministic issue when getting the pipeline dtype or device across multiple runs of the same Python script.
Internally, the ordering issue comes from the use of
setin the_get_signature_keysfunction in thesrc/diffusers/pipelines/pipeline_utils.pyfile. As stated in the Python documentation,setdoes not guarantee that the elements will have the same order across multiple runs of a Python script. This is due to the hashing algorithm which is initialized with a random salt each time the python interpreter is started. More information can be found here (especially in the second note at the end of the section).Fixes #10671.
Before submitting
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@sayakpaul