[docs] Pipeline-level quantization #11604
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
|
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. |
sayakpaul
approved these changes
May 23, 2025
| Initialize [`~quantizers.PipelineQuantizationConfig`] with the following parameters. | ||
|
|
||
| </Tip> | ||
| - `quant_backend` specifies which quantization backend to use. Currently supported backends include: `bitsandbytes_4bit`, `bitsandbytes_8bit`, `gguf`, `quanto`, and `torchao`. |
There was a problem hiding this comment.
Easy to miss when adding a new backend. Additionally, with the granular option, it is possible to use a quantization backend for transformers that may not exist in diffusers and vice-versa. How do we make that info clear? 👀
| ``` | ||
|
|
||
| Then pass it to [`~DiffusionPipeline.from_pretrained`] and run inference: | ||
| There is a separate bitsandbytes backend in [Transformers](https://huggingface.co/docs/transformers/main_classes/quantization#transformers.BitsAndBytesConfig). You need to import and use [`transformers.BitsAndBytesConfig`] for components that come from Transformers. For example, `text_encoder_2` in [`FluxPipeline`] is a [`~transformers.T5EncoderModel`] from Transformers so you need to use [`transformers.BitsAndBytesConfig`] instead of [`diffusers.BitsAndBytesConfig`]. |
There was a problem hiding this comment.
Suggested change
| There is a separate bitsandbytes backend in [Transformers](https://huggingface.co/docs/transformers/main_classes/quantization#transformers.BitsAndBytesConfig). You need to import and use [`transformers.BitsAndBytesConfig`] for components that come from Transformers. For example, `text_encoder_2` in [`FluxPipeline`] is a [`~transformers.T5EncoderModel`] from Transformers so you need to use [`transformers.BitsAndBytesConfig`] instead of [`diffusers.BitsAndBytesConfig`]. | |
| There is a separate `bitsandbytes` backend in [Transformers](https://huggingface.co/docs/transformers/main_classes/quantization#transformers.BitsAndBytesConfig). You need to import and use [`transformers.BitsAndBytesConfig`] for components that come from Transformers. For example, `text_encoder_2` in [`FluxPipeline`] is a [`~transformers.T5EncoderModel`] from Transformers so you need to use [`transformers.BitsAndBytesConfig`] instead of [`diffusers.BitsAndBytesConfig`]. |
| recommended to quantize the text encoders that are memory-intensive. Some examples include T5, | ||
| Llama, Gemma, etc. In the above example, you quantized the T5 model of [`FluxPipeline`] through | ||
| `text_encoder_2` while keeping the CLIP model intact (accessible through `text_encoder`). | ||
| - Read the [Exploring Quantization Backends in Diffusers](https://huggingface.co/blog/diffusers-quantization) blog post for a brief introduction to each quantization backend, how to choose a backend, and combining quantization with other memory optimizations. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
Lightly refactors the pipeline-level quantization doc to emphasize the two ways of using
PipelineQuantizationConfig, the difference betweendiffusers.BitsAndBytesConfigandtransformers.BitsAndBytesConfig, and create a new Resource section.