FIX: Small fixes to hotswapping #2366
Merged
Conversation
This file contains hidden or 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
A couple of smaller issues that surfaced when working on the diffusers integration are not fixed. - Better detection if model is compiled in prepare_model_for_compiled_hotswap - Fix handling of models that are compiled but where compilation is not detected (from "inside" the model) - Handle device of swapped in adapter weights. - Wrong adapter name in compiled diffusion model test - Add hotswap test for different alphas and ranks but model not being compiled (linear and conv2d)
|
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. |
BenjaminBossan
added a commit
to BenjaminBossan/diffusers
that referenced
this pull request
Feb 7, 2025
Requires huggingface/peft#2366 More changes to make hotswapping work. Together with the mentioned PEFT PR, the tests pass for me locally. List of changes: - docstring for hotswap - remove code copied from PEFT, import from PEFT now - adjustments to PeftAdapterMixin.load_lora_adapter (unfortunately, some state dict renaming was necessary, LMK if there is a better solution) - adjustments to UNet2DConditionLoadersMixin._process_lora: LMK if this is even necessary or not, I'm unsure what the overall relationship is between this and PeftAdapterMixin.load_lora_adapter - also in UNet2DConditionLoadersMixin._process_lora, I saw that there is no LoRA unloading when loading the adapter fails, so I added it there (in line with what happens in PeftAdapterMixin.load_lora_adapter) - rewritten tests to avoid shelling out, make the test more precise by making sure that the outputs align, parametrize it - also checked the pipeline code mentioned in this comment: huggingface#9453 (comment); when running this inside the with torch._dynamo.config.patch(error_on_recompile=True) context, there is no error, so I think hotswapping is now working with pipelines.
6 tasks
It's used directly in diffusers.
Use local imports instead. This is because the bnb nightly tests use this test file but diffusers is not installed, so there is an error right now. Additionally, I added a skipif for when diffusers is not installed. This is not strictly necessary, since the bnb tests run with: pytest -m "single_gpu_tests and bitsandbytes" So this test should be skipped anyway, but it won't hurt to have this decorator.
It was actually easier than I thought as for Conv2d, the pad dimensions are still either 0 or 1, same as linear, so no extra code was required.
githubnemo
approved these changes
Feb 12, 2025
BenjaminBossan
added a commit
to BenjaminBossan/peft
that referenced
this pull request
Feb 18, 2025
A couple of smaller issues that surfaced when working on the diffusers integration are not fixed. - Better detection if model is compiled in prepare_model_for_compiled_hotswap - Fix handling of models that are compiled but where compilation is not detected (from "inside" the model) - Handle device of swapped in adapter weights. - Wrong adapter name in compiled diffusion model test - Add hotswap test for different alphas and ranks but model not being compiled (linear and conv2d) - Make _check_hotswap_configs_compatible "public" - Don't import diffusers in test root - Add support for compiled Conv2d
jonluca
added a commit
to weights-ai/diffusers
that referenced
this pull request
Mar 20, 2025
* [WIP][LoRA] Implement hot-swapping of LoRA This PR adds the possibility to hot-swap LoRA adapters. It is WIP. Description As of now, users can already load multiple LoRA adapters. They can offload existing adapters or they can unload them (i.e. delete them). However, they cannot "hotswap" adapters yet, i.e. substitute the weights from one LoRA adapter with the weights of another, without the need to create a separate LoRA adapter. Generally, hot-swapping may not appear not super useful but when the model is compiled, it is necessary to prevent recompilation. See huggingface#9279 for more context. Caveats To hot-swap a LoRA adapter for another, these two adapters should target exactly the same layers and the "hyper-parameters" of the two adapters should be identical. For instance, the LoRA alpha has to be the same: Given that we keep the alpha from the first adapter, the LoRA scaling would be incorrect for the second adapter otherwise. Theoretically, we could override the scaling dict with the alpha values derived from the second adapter's config, but changing the dict will trigger a guard for recompilation, defeating the main purpose of the feature. I also found that compilation flags can have an impact on whether this works or not. E.g. when passing "reduce-overhead", there will be errors of the type: > input name: arg861_1. data pointer changed from 139647332027392 to 139647331054592 I don't know enough about compilation to determine whether this is problematic or not. Current state This is obviously WIP right now to collect feedback and discuss which direction to take this. If this PR turns out to be useful, the hot-swapping functions will be added to PEFT itself and can be imported here (or there is a separate copy in diffusers to avoid the need for a min PEFT version to use this feature). Moreover, more tests need to be added to better cover this feature, although we don't necessarily need tests for the hot-swapping functionality itself, since those tests will be added to PEFT. Furthermore, as of now, this is only implemented for the unet. Other pipeline components have yet to implement this feature. Finally, it should be properly documented. I would like to collect feedback on the current state of the PR before putting more time into finalizing it. * Reviewer feedback * Reviewer feedback, adjust test * Fix, doc * Make fix * Fix for possible g++ error * Add test for recompilation w/o hotswapping * Make hotswap work Requires huggingface/peft#2366 More changes to make hotswapping work. Together with the mentioned PEFT PR, the tests pass for me locally. List of changes: - docstring for hotswap - remove code copied from PEFT, import from PEFT now - adjustments to PeftAdapterMixin.load_lora_adapter (unfortunately, some state dict renaming was necessary, LMK if there is a better solution) - adjustments to UNet2DConditionLoadersMixin._process_lora: LMK if this is even necessary or not, I'm unsure what the overall relationship is between this and PeftAdapterMixin.load_lora_adapter - also in UNet2DConditionLoadersMixin._process_lora, I saw that there is no LoRA unloading when loading the adapter fails, so I added it there (in line with what happens in PeftAdapterMixin.load_lora_adapter) - rewritten tests to avoid shelling out, make the test more precise by making sure that the outputs align, parametrize it - also checked the pipeline code mentioned in this comment: huggingface#9453 (comment); when running this inside the with torch._dynamo.config.patch(error_on_recompile=True) context, there is no error, so I think hotswapping is now working with pipelines. * Address reviewer feedback: - Revert deprecated method - Fix PEFT doc link to main - Don't use private function - Clarify magic numbers - Add pipeline test Moreover: - Extend docstrings - Extend existing test for outputs != 0 - Extend existing test for wrong adapter name * Change order of test decorators parameterized.expand seems to ignore skip decorators if added in last place (i.e. innermost decorator). * Split model and pipeline tests Also increase test coverage by also targeting conv2d layers (support of which was added recently on the PEFT PR). * Reviewer feedback: Move decorator to test classes ... instead of having them on each test method. * Apply suggestions from code review Co-authored-by: hlky <[email protected]> * Reviewer feedback: version check, TODO comment * Add enable_lora_hotswap method * Reviewer feedback: check _lora_loadable_modules * Revert changes in unet.py * Add possibility to ignore enabled at wrong time * Fix docstrings * Log possible PEFT error, test * Raise helpful error if hotswap not supported I.e. for the text encoder * Formatting * More linter * More ruff * Doc-builder complaint * Update docstring: - mention no text encoder support yet - make it clear that LoRA is meant - mention that same adapter name should be passed * Fix error in docstring * Update more methods with hotswap argument - SDXL - SD3 - Flux No changes were made to load_lora_into_transformer. * Add hotswap argument to load_lora_into_transformer For SD3 and Flux. Use shorter docstring for brevity. * Extend docstrings * Add version guards to tests * Formatting * Fix LoRA loading call to add prefix=None See: huggingface#10187 (comment) * Run make fix-copies * Add hot swap documentation to the docs * Apply suggestions from code review Co-authored-by: Steven Liu <[email protected]> --------- Co-authored-by: Benjamin Bossan <[email protected]> Co-authored-by: Sayak Paul <[email protected]> Co-authored-by: Benjamin Bossan <[email protected]> Co-authored-by: hlky <[email protected]> Co-authored-by: YiYi Xu <[email protected]> Co-authored-by: Steven Liu <[email protected]>
sayakpaul
added a commit
to huggingface/diffusers
that referenced
this pull request
Apr 8, 2025
* [WIP][LoRA] Implement hot-swapping of LoRA This PR adds the possibility to hot-swap LoRA adapters. It is WIP. Description As of now, users can already load multiple LoRA adapters. They can offload existing adapters or they can unload them (i.e. delete them). However, they cannot "hotswap" adapters yet, i.e. substitute the weights from one LoRA adapter with the weights of another, without the need to create a separate LoRA adapter. Generally, hot-swapping may not appear not super useful but when the model is compiled, it is necessary to prevent recompilation. See #9279 for more context. Caveats To hot-swap a LoRA adapter for another, these two adapters should target exactly the same layers and the "hyper-parameters" of the two adapters should be identical. For instance, the LoRA alpha has to be the same: Given that we keep the alpha from the first adapter, the LoRA scaling would be incorrect for the second adapter otherwise. Theoretically, we could override the scaling dict with the alpha values derived from the second adapter's config, but changing the dict will trigger a guard for recompilation, defeating the main purpose of the feature. I also found that compilation flags can have an impact on whether this works or not. E.g. when passing "reduce-overhead", there will be errors of the type: > input name: arg861_1. data pointer changed from 139647332027392 to 139647331054592 I don't know enough about compilation to determine whether this is problematic or not. Current state This is obviously WIP right now to collect feedback and discuss which direction to take this. If this PR turns out to be useful, the hot-swapping functions will be added to PEFT itself and can be imported here (or there is a separate copy in diffusers to avoid the need for a min PEFT version to use this feature). Moreover, more tests need to be added to better cover this feature, although we don't necessarily need tests for the hot-swapping functionality itself, since those tests will be added to PEFT. Furthermore, as of now, this is only implemented for the unet. Other pipeline components have yet to implement this feature. Finally, it should be properly documented. I would like to collect feedback on the current state of the PR before putting more time into finalizing it. * Reviewer feedback * Reviewer feedback, adjust test * Fix, doc * Make fix * Fix for possible g++ error * Add test for recompilation w/o hotswapping * Make hotswap work Requires huggingface/peft#2366 More changes to make hotswapping work. Together with the mentioned PEFT PR, the tests pass for me locally. List of changes: - docstring for hotswap - remove code copied from PEFT, import from PEFT now - adjustments to PeftAdapterMixin.load_lora_adapter (unfortunately, some state dict renaming was necessary, LMK if there is a better solution) - adjustments to UNet2DConditionLoadersMixin._process_lora: LMK if this is even necessary or not, I'm unsure what the overall relationship is between this and PeftAdapterMixin.load_lora_adapter - also in UNet2DConditionLoadersMixin._process_lora, I saw that there is no LoRA unloading when loading the adapter fails, so I added it there (in line with what happens in PeftAdapterMixin.load_lora_adapter) - rewritten tests to avoid shelling out, make the test more precise by making sure that the outputs align, parametrize it - also checked the pipeline code mentioned in this comment: #9453 (comment); when running this inside the with torch._dynamo.config.patch(error_on_recompile=True) context, there is no error, so I think hotswapping is now working with pipelines. * Address reviewer feedback: - Revert deprecated method - Fix PEFT doc link to main - Don't use private function - Clarify magic numbers - Add pipeline test Moreover: - Extend docstrings - Extend existing test for outputs != 0 - Extend existing test for wrong adapter name * Change order of test decorators parameterized.expand seems to ignore skip decorators if added in last place (i.e. innermost decorator). * Split model and pipeline tests Also increase test coverage by also targeting conv2d layers (support of which was added recently on the PEFT PR). * Reviewer feedback: Move decorator to test classes ... instead of having them on each test method. * Apply suggestions from code review Co-authored-by: hlky <[email protected]> * Reviewer feedback: version check, TODO comment * Add enable_lora_hotswap method * Reviewer feedback: check _lora_loadable_modules * Revert changes in unet.py * Add possibility to ignore enabled at wrong time * Fix docstrings * Log possible PEFT error, test * Raise helpful error if hotswap not supported I.e. for the text encoder * Formatting * More linter * More ruff * Doc-builder complaint * Update docstring: - mention no text encoder support yet - make it clear that LoRA is meant - mention that same adapter name should be passed * Fix error in docstring * Update more methods with hotswap argument - SDXL - SD3 - Flux No changes were made to load_lora_into_transformer. * Add hotswap argument to load_lora_into_transformer For SD3 and Flux. Use shorter docstring for brevity. * Extend docstrings * Add version guards to tests * Formatting * Fix LoRA loading call to add prefix=None See: #10187 (comment) * Run make fix-copies * Add hot swap documentation to the docs * Apply suggestions from code review Co-authored-by: Steven Liu <[email protected]> --------- Co-authored-by: Sayak Paul <[email protected]> Co-authored-by: hlky <[email protected]> Co-authored-by: YiYi Xu <[email protected]> Co-authored-by: Steven Liu <[email protected]>
Guy-Bilitski
pushed a commit
to Guy-Bilitski/peft
that referenced
this pull request
May 13, 2025
A couple of smaller issues that surfaced when working on the diffusers integration are not fixed. - Better detection if model is compiled in prepare_model_for_compiled_hotswap - Fix handling of models that are compiled but where compilation is not detected (from "inside" the model) - Handle device of swapped in adapter weights. - Wrong adapter name in compiled diffusion model test - Add hotswap test for different alphas and ranks but model not being compiled (linear and conv2d) - Make _check_hotswap_configs_compatible "public" - Don't import diffusers in test root - Add support for compiled Conv2d
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
A couple of smaller issues that surfaced when working on the diffusers integration are not fixed.
prepare_model_for_compiled_hotswap